Terrace Rugby: An Improvement Case Study
A note on reading this case study
The Terrace Rugby Case Study was designed and typeset as a PDF monograph, with a layout intended for sustained reading.
The full text is reproduced below for convenience, but I recommend reading the PDF edition for the best experience, particularly when reviewing the tables, data and historical comparisons.
Executive Summary
This case study examines the transformation of the rugby program at St Joseph's College, Gregory Terrace, and the competitive results that followed.
Beginning in 2012, Terrace undertook an ambitious improvement program that fundamentally altered its traditional approach to rugby. A school that had historically resisted sporting scholarships and recruitment began actively recruiting talented players. Substantial investment in facilities, specialist staffing and athletic development accompanied a broader institutional commitment to improving the program. Changes to the competition calendar and the age profile of players also removed disadvantages under which Terrace had previously operated.
These changes created considerably more favourable conditions for competitive performance. The central question of this study is whether they produced the improvement they were intended to achieve.
The evidence indicates that they did not.
Across the eight assessed seasons preceding the intervention (2003–2010), Terrace recorded a winning percentage of 42.9%. Across thirteen assessed seasons following its commencement (2012–2019 and 2022–2026), that figure declined to 35.6%. Expressed in practical terms, Terrace averaged approximately 0.58 fewer win-equivalents per eight-match season after the changes than before them.
The deterioration was not confined to winning percentage. Terrace fell from fifth to eighth among the nine GPS schools in overall winning-percentage ranking, while its average and typical finishing positions also worsened. The five most recent assessed seasons produced a winning percentage of just 32.5%.
Perhaps more surprisingly, representative-player production also declined. Across comparable nine-year periods, Terrace's annual production of Australian Schoolboys representatives fell by approximately 45%, despite the introduction of recruitment practices that included attracting players already identified within representative development pathways. In the extended representative-production comparison, the annual rate fell by approximately 53% relative to the historical benchmark.
The findings are not dependent on a small number of exceptional seasons. Even when Terrace's undefeated 2004 premiership is removed from the historical baseline while the post-intervention program retains its exceptional 2016 season, the resulting advantage amounts to just 0.06 additional win-equivalents per season. Projected across thirteen seasons, that represents less than one additional win-equivalent in total. This is a deliberately favourable comparison, not an actual count of additional victories.
The magnitude of this result deserves emphasis. Terrace undertook extensive changes to recruitment, resources, facilities, preparation and organisational arrangements, yet failed to achieve even modest, sustained competitive improvement against its own historical record.
This study does not suggest that Terrace was obliged to become the dominant rugby school in Queensland, nor does it equate the educational value of school sport exclusively with winning. The relevant benchmark is the competitive performance of Terrace's previous program, which operated under substantially less favourable conditions.
Nor do the results establish that the interventions themselves caused the deterioration. Changes in opposition strength, player cohorts and other circumstances complicate causal interpretation. Benefits in participation, wellbeing or other educational outcomes may also exist beyond the measures examined here.
What the evidence does establish is a substantial discrepancy between the scale of the changes undertaken and the competitive outcomes that followed.
The purpose of this report is to document that discrepancy, establish the historical context and provide a foundation for further investigation into the intervention's design, implementation and underlying assumptions.
Its broader significance lies in a fundamental problem of organisational improvement: changing the components, resources and operating conditions of a system does not necessarily improve the performance of the system itself.
Prologue: An Amateur Team in a Professional Competition
An illustrative thought experiment
Imagine it is 1996. Rugby has entered the professional era, and Super Rugby is about to begin.
But in this version of events, one franchise makes an extraordinary decision. The Queensland Reds decide they will remain amateur.
They will continue competing against professional opponents, but they will not adopt the professional model themselves.
The disadvantages are obvious.
While their opponents can employ full-time players and staff, the Reds must organise their rugby around the other commitments of amateur athletes. They cannot dedicate comparable time and resources to training, preparation and development.
Recruitment presents another problem. While their competitors can acquire players from further afield, the Reds remain dependent on Queensland's local club competition.
It is a strong competition, but talent is not distributed evenly. Some years produce outstanding players in particular positions. Others do not. If the Reds find themselves short of quality in a critical position, they cannot simply recruit someone to address the problem.
Yet something surprising happens.
Despite these disadvantages, the Reds remain reasonably competitive. Across their first decade, they win approximately four of every ten matches. They occasionally produce exceptional teams and develop an impressive number of Wallabies.
They are not dominant, but neither are they hopelessly outclassed.
Then comes their eleventh season.
A combination of injuries, an unusually weak playing group, poor decisions and unfortunate circumstances produces a disaster. The Reds lose every match, often by substantial margins.
The result provokes a fundamental reassessment.
The organisation has spent a decade operating under disadvantages that its competitors do not face. Surely those disadvantages explain what has happened. Surely the solution is to remove them.
And so the Reds finally embrace professionalism.
They recruit talented players from beyond their traditional catchment. They employ specialist staff, invest in facilities, expand their preparation and devote substantially more institutional resources to rugby.
The conditions under which the program operates are transformed.
Now consider what you would reasonably expect to happen.
Not necessarily a premiership. Not necessarily immediate dominance. But surely an improvement in competitive performance.
Instead, imagine that over the following decade, the Reds win only 35% of their matches, compared with approximately 40% before the transformation. Imagine that their production of Wallabies also falls by almost half.
Despite acquiring many of the advantages they previously lacked, they become less successful.
It sounds implausible. An organisation operates at a substantial competitive disadvantage, commits to an extensive improvement program, removes many of those disadvantages, and subsequently achieves worse results.
The Queensland Reds scenario is fictional.
The underlying pattern is not.
This is the story of what happened to the rugby program at one of Australia's most prominent schools.
And it raises a question that extends far beyond schoolboy rugby:
How can an organisation change so much, acquire so many advantages, and improve so little?
Part One: Introduction and Context
1. Why This Case Study?
For more than a decade, the rugby program at St Joseph's College, Gregory Terrace (referred to throughout this document as Terrace) has been one of my most frequently used examples when discussing coaching, organisational performance and improvement.
Whether consulting with a rugby organisation, working with coaches, or simply talking shop over a coffee or a beer, there is a good chance I have drawn on something from Terrace to illustrate a particular problem or principle. More often than not, it has been an example of what not to do.
For readers unfamiliar with the background, Terrace is a school in Brisbane, Australia, competing in the Queensland Great Public Schools (GPS) rugby competition. In 2012, it embarked on an overhaul of its rugby program unprecedented in the school's history, fundamentally departing from a long-established sporting philosophy and institutional approach that had distinguished it from many of its competitors. What followed was not merely an increase in resources or a change in coaching personnel, but a far-reaching transformation of the program's approach to recruitment, staffing, preparation and organisation, all undertaken with the intention of improving competitive performance. More than a decade later, there is now a substantial record of results against which to examine those changes.
Despite the extraordinary scale of the intervention, the competitive improvement it was intended to produce has not materialised. Across the completed seasons examined here, Terrace has failed to demonstrate meaningful improvement against its own historical performance. Indeed, its winning percentage declined from 42.9% to approximately 35.6%, the equivalent of about 0.58 fewer win-equivalents per eight-match season. Representative-player production also plummeted against the relevant historical benchmarks.
It is on this basis that I regard the improvement program as a failure. Not because Terrace has failed to become the best rugby school in Queensland, but because a substantial intervention intended to improve an existing program has not delivered the competitive improvement it was established to achieve. The recorded decline is an additional finding; it does not, on its own, establish that the intervention caused it.
That history helps explain why I keep returning to Terrace. The scale of the intervention, the contrast between its ambitions and subsequent outcomes, and the persistence of some of its underlying approaches make it an unusually useful example of organisational improvement. But its value to me has never rested on one particular statistic or one particular mistake.
The case is remarkably versatile. I have used it to discuss specific coaching decisions, player selection, recruitment and day-to-day operational practices. I have also used it to explore much larger questions: how sporting organisations understand their own capabilities; how the parts of a program interact; why apparently sensible changes can fail to produce improvement; and how an organisation decides whether the path it has chosen is actually taking it towards its intended destination.
Sometimes the conversation has not even been about rugby. An example from Terrace can be a useful way to make an otherwise abstract problem in systems thinking or organisational decision-making concrete.
One theme recurs throughout those conversations. When organisations want to improve, they often begin by identifying individual things they believe are missing or inadequate. Recruit better players. Employ more qualified coaches. Increase training time. Introduce specialist programs. Invest in facilities and resources.
None of those changes is necessarily a mistake. Each might be valuable in the right circumstances. But changing or improving an individual component does not guarantee improvement in the system of which it is a part. The outcome also depends on the relationships, constraints and interactions that determine how the whole program operates.
That difference between making changes and producing improvement sits near the centre of my work through Rugbyology, and Terrace has repeatedly given me a practical setting in which to examine it.
Until now, most of that examination has happened privately: in consulting engagements, coaching discussions and informal conversations. As I publish more of my ideas through Rugbyology, I want to make the examples and historical context behind those discussions accessible to other people.
Ideas about improvement are useful, but they become more revealing when we can test them against actual decisions, actual organisations and actual outcomes. An example gives us something to investigate rather than simply another abstract proposition to agree or disagree with.
Although the subject of this document is a particular school rugby program, the larger purpose is not to produce an exhaustive commentary on that school's fortunes. Terrace is a vehicle through which to examine problems of improvement that extend well beyond one school, one competition and, in many instances, rugby itself.
2. Purpose and Scope of This Document
This document provides the background for a wider, ongoing case study of Terrace Rugby. Its immediate purpose is straightforward: to establish what happened, what was changed and what outcomes followed.
It introduces the setting in which the improvement program emerged, describes the circumstances and changes involved, and sets out the subsequent results using the available historical record. It is intended to give readers an accessible point of reference, particularly those who have little or no familiarity with Terrace, Queensland schoolboy rugby or the scale of the intervention.
It is not intended to provide a complete explanation of why the program produced those outcomes. That is a different undertaking.
The possible effects of recruiting players, the interaction of coaching and selection decisions, the development of existing students, the organisation of resources, and the persistence of particular approaches all warrant their own examination. So do the larger questions of systems thinking, organisational learning and the assumptions behind common improvement strategies.
Those subjects will be explored separately through Rugbyology essays, articles, videos, podcasts and other publications. Some may be substantial investigations of Terrace itself. Others may be discussions of a general coaching or organisational idea that use Terrace only briefly as an illustration.
This document allows those discussions to begin with the relevant question rather than repeatedly reconstructing the history of the program. A reader encountering Terrace in a future article should be able to come here, understand the circumstances and results, and return to that article with the necessary context.
The relationship should work in the other direction too. Someone encountering Terrace through this document should be able to explore related analyses and see how particular aspects of the case illuminate broader improvement problems.
There is a wider reason to make that material available. An American college coach, a community club volunteer or a professional director of rugby need not have any particular interest in one Australian school to find value in studying what it did. The specific circumstances may differ, but problems of diagnosis, capability, resource allocation, unintended consequences and adaptation arise across sport.
Whether Terrace has learned from the mistakes of its improvement program is open to debate. Its subsequent results provide little evidence that any such learning has translated into meaningful improvement.
Fortunately, an organisation doesn't have to learn from its own mistakes for others to learn from them.
Whether Terrace itself engages with these questions does not determine their value to anyone else. If this material helps other organisations think more carefully about their own systems and make better decisions about improvement, it will have served a worthwhile purpose.
The document is also intended to remain useful over time. It will be published in an initial version and updated as further information, results and relevant measurements become available. The historical account should become more complete without requiring every subsequent interpretation to be incorporated into it.
3. Introducing Terrace and the Queensland GPS Competition
To understand the case, it helps to understand the environment in which Terrace competes. For an international reader, the phrase schoolboy rugby may suggest a fairly ordinary school sporting competition. Queensland GPS rugby is something quite different.
The GPS is a long-established association of Brisbane and Queensland schools competing across a range of sports. Its rugby competition involves nine schools playing an eight-match, single round-robin season. Each school plays every other school once, with home and away fixtures alternating between seasons. There is no finals series: the premiership is decided by the competition standings.
The structure is simple, but unforgiving. Eight matches provide limited room for a poor performance, and each opponent represents a substantial competitive challenge. Winning the GPS First XV premiership carries considerable prestige within the schools and their wider communities.
The competition is also an important part of the Australian rugby development pathway. Its schools have a long history of producing Wallabies, Australian Schoolboys and professional players. About one-third of Wallabies have come through this pathway, both over the past decade and across the history of the national team. At the next Rugby World Cup, half of the competing nations will have had a player from this small, nine-team competition represent them during the preceding quadrennial period.
The significance of that pathway extends beyond rugby union. GPS rugby is also a route into professional rugby league, and students can progress directly from school into elite rugby union and rugby league programs. Unlike development systems in which school is followed by several years of university competition, Queensland schoolboy rugby can be the final stage before a player enters a professional environment. This does not mean every player leaves school ready for a professional first team, but the pathway can be immediate.
It remains school sport, yet it operates at times as a highly resourced performance environment. Some schools commit extraordinary resources to recruitment, scholarships, specialist coaching and preparation. Established rivalries, substantial community attention and the direct connection to professional pathways give the competition a significance that extends well beyond the school calendar.
Terrace occupies a distinctive place within that environment. It has a long and distinguished rugby history, including a strong run of premierships in the 1970s, and has historically been one of Australia's leading school producers of Wallabies. Its rugby identity is closely bound to its rivalry with Nudgee College.
Nudgee emerged from Terrace's early institutional history, and the schools' traditional Battle of the Colours is associated with the story of how Nudgee retained blue and white while Terrace adopted red and black. The fixture remains one of the great occasions in Australian schoolboy rugby, with an importance that extends well beyond its position in any given season.
Yet historical significance and contemporary competitiveness are different things. Terrace's victories over Nudgee since 1996 have been rare, occurring in 1996, 2004 and 2016. The rivalry has retained its cultural stature even when results on the field have been heavily one-sided.
A major change in the wider GPS environment occurred around the late 1990s and early 2000s, as the recruitment of sporting talent and the use of scholarships expanded. Schools increasingly sought competitive advantages through attracting players as well as developing those already enrolled. Over time, approaches that had once distinguished a small number of schools became much more widespread.
Terrace was unusual in resisting that shift. Its traditional approach placed academic achievement and a broad education ahead of sporting recruitment. Rugby was an important co-curricular activity, but not a reason to abandon the school's wider educational philosophy. For much of the period before the intervention, the program largely worked with the students already at the school rather than recruiting players to fill gaps in a particular team.
I have often described Terrace during that period as an amateur team in a professional competition. That is not a claim that its players or coaches lacked commitment. It describes the contrast between Terrace's traditional approach and the increasingly recruitment- and resource-driven practices of many competitors.
This difference had obvious implications. A school recruiting for a particular age group or playing position could address a perceived shortage by bringing in a player. Terrace could not ordinarily respond in the same way. It had to accept the natural variation in its student cohorts and make the most of the players available.
There were certainly stronger and weaker years. Some cohorts contained players already identified as exceptional talents throughout their junior rugby years. Others had fewer established representative players and greater developmental demands. But across the pre-intervention period, Terrace was generally a competitive, roughly middle-of-the-table program rather than either a perennial premiership contender or a perennial last-place team. It also continued to produce high-level representative players.
The 2010 First XV is worth noting here because of its timing. Immediately before the 2011 collapse that helped precipitate the improvement program, a Terrace team expected to finish last without a victory instead finished fourth and produced an exceptional representative cohort. Its implications for the capabilities of the existing system are examined more fully in Section 4.4.
In other words, before examining what was changed, we need to recognise the kind of program that already existed: one with a different philosophy from many of its competitors, recognisable constraints, a substantial history of player development and a relatively stable range of competitive results.
That background matters because the improvement program did not begin with a blank sheet of paper. It intervened in an existing organisation, with established priorities, practices and capabilities.
That relative stability was dramatically interrupted in 2011, a season that would become the catalyst for Terrace's subsequent rugby improvement program.
4. Why Terrace Is a Particularly Valuable Case Study
Sport tends to study success. A team wins a championship, a coach establishes a reputation, or an organisation enjoys an exceptional season, and attention turns to identifying the decisions and practices that supposedly made it happen. Much of that analysis takes place after the result is known. It is not always easy to distinguish the causes of success from the explanations constructed around it.
There is at least as much to learn from an organisation that makes a serious, deliberate effort to improve and does not achieve the intended improvement. Terrace is especially useful in this respect. It is not simply a program that has endured some disappointing rugby seasons. It is a program that underwent substantial change, in a well-established competition, with a long history against which those changes can be examined.
Several features make the case more revealing than an ordinary account of a struggling team. Some concern the intervention; others concern the evidence, the decisions made over time and the relevance of the lessons. The reasons overlap in places, but each brings a different question into view.
4.1. Questions raised before the outcomes were known
One of the distinctive features of the Terrace case is that important questions about the improvement program's assumptions and design were raised before its competitive outcomes became apparent. The concerns that now warrant examination were not simply invented after the results were known.
This matters because retrospective sporting analysis is especially susceptible to convenient explanations. Once a team succeeds or fails, it is relatively easy to construct a plausible account of why. A more demanding exercise is to examine the assumptions behind decisions before their consequences were known, and compare those assumptions with the outcomes that followed.
Early concerns centred on the diagnosis of the problem. The disastrous 2011 season occurred amid an exceptional combination of circumstances, including the disruption caused by the Brisbane floods, an unusually weak playing group and concerns about the effectiveness of the coaching arrangements. It was therefore reasonable to question whether one extraordinary season provided a sound basis for judging the capabilities of the existing program or redesigning it so extensively.
There were also questions about the competitive strengths Terrace already possessed, how those strengths had been developed, and whether the proposed changes recognised and preserved them. As recruitment and other reforms were introduced, the concerns became more specific: could the receiving coaching and development systems actually exploit the additional players, resources and specialist support? Recruitment and investment were not necessarily misguided. Their value depended in part on the quality of the program into which they were introduced. More resources would not automatically correct weaknesses in how players were coached, prepared, selected or organised.
These concerns were raised in repeated discussions with Terrace's Director of Rugby, Director of Sport and other decision-makers between 2011 and 2014. Their significance lies in the fact that important questions about the intervention's assumptions and design were being raised before its competitive outcomes became apparent.
4.2. Persistence despite accumulating contrary evidence
Imagine setting out on a bushwalk to see a particular waterfall. Somewhere along the way, the signage becomes unclear and you arrive at a fork in the trail. Most of the other hikers seem to be taking one particular path, and a member of your group, who has previously seen a photograph of the waterfall, confidently assures everyone that this is the correct direction.
Just as you're about to set off, a passing park ranger, someone intimately familiar with the terrain and the various trails, warns that you're heading in the wrong direction. Not only will the path fail to take you to the waterfall, but it will actually take you further away from your intended destination.
Nevertheless, you decide to follow the crowd and the advice of your companion.
Before long, the evidence begins to accumulate. The terrain doesn't match your expectations, the signs increasingly suggest you're heading in the wrong direction, and the waterfall appears no closer. Yet the group continues along the same path.
Perhaps the initial decision was a mistake. It may even have been a poor decision given the warning you received. But people make imperfect decisions under uncertainty all the time. That's hardly unusual.
What makes this particular bushwalk interesting is something else entirely.
At regular intervals along the incorrect path, there are travelators that will take you directly back to the correct trail. They're free, clearly signposted, air-conditioned, sheltered and comfortable. There's no need to retrace your steps through difficult terrain or expend considerable effort correcting the original mistake. You simply have to step onto one and let it carry you back.
Every few hundred metres, you encounter another.
And every time, you walk straight past.
You continue putting in the effort, covering more ground and moving further from the waterfall, despite increasingly obvious evidence that the path is wrong and repeated opportunities to correct it.
Eventually, the most interesting question is no longer why you chose the wrong path in the first place.
It's why you continue walking past every opportunity to get back on the right one.
This is one of the most intriguing aspects of the Terrace Rugby improvement program.
Organisations must make decisions without knowing the future. A decision that subsequently proves unsuccessful is not necessarily evidence of poor judgment. Even the decision to proceed despite contemporaneous warnings might be defensible if the alternative evidence and reasoning were sufficiently compelling.
But the decision to initiate an improvement program is only one part of the story. Equally important is what happens as the results of that program become apparent.
In Terrace's case, the subsequent performance record repeatedly raised questions about the effectiveness of the intervention. Yet the broad direction appears to have persisted across an extended period, despite numerous opportunities to reconsider its underlying assumptions.
That does not mean no adjustments were made. Coaches, personnel, practices and other elements of the program have changed over time. The more significant question is whether those adjustments represented fundamental reconsiderations of the original theory of improvement, or merely modifications within an approach that remained substantially unchanged.
The travelator is particularly relevant because many of the constraints and bottlenecks that appear to limit Terrace's performance may be relatively straightforward and inexpensive to address. Unlike organisations that become structurally trapped by the consequences of their decisions, Terrace may have retained accessible opportunities to change direction throughout much of the intervention.
That proposition will require closer examination, but it adds another dimension to the question of organisational persistence.
4.3. The magnitude of the intervention and the subsequent outcomes
The scale of the changes is another reason this case deserves attention. Terrace did not merely change one coach or introduce a new training session. The intervention was substantial, affecting recruitment, staffing, resources, preparation, organisational arrangements and elements of the school's traditional approach to rugby.
Nor is the central concern that an ambitious program narrowly missed an ambitious target. The available results show no meaningful competitive improvement against the historical baseline, and several outcomes, including representative-player production and more recent team results, deteriorated. The results section examines those differences rather than asking readers to accept them on assertion.
It is the contrast between the intervention and its outcomes that makes the case noteworthy. A substantial investment followed by sustained improvement would raise one set of questions. A substantial investment followed by little or no improvement raises another, particularly once its financial and organisational costs are considered. The recorded deterioration in several outcomes adds to the puzzle.
The interest lies not merely in the size of the changes or the poor results, but in the distance between the changes undertaken and the improvement they were intended to produce.
This does not by itself establish that the intervention caused the deterioration. The distinction between what happened after a change and what happened because of it will remain important throughout the case study.
4.4. An existing system that had already demonstrated significant capabilities
Perhaps one of the most important features of the case is that Terrace did not begin its improvement effort with a system that had never demonstrated anything of value.
Before the intervention, the rugby program had a long record of producing elite players and was generally competitive around the middle of the GPS competition despite operating differently from many scholarship-recruiting rivals. The timing of the 2010 First XV makes it particularly revealing. It was the last team before the disastrous 2011 season that helped trigger the improvement program, not an exceptional cohort selected at random from the distant past. It entered as the preseason favourite for the wooden spoon, with no victories expected. Instead, it finished fourth and produced an exceptional number of representative players. That achievement occurred immediately before the decision to remake the program.
A single exceptional team does not prove that the system could have reproduced those outcomes routinely. Nor does it establish precisely which features of the program made them possible. But it does establish a question that should precede any major redesign: what was the existing system already capable of doing, and how were those capabilities produced?
Were its strengths recognised, understood and preserved? Did the intervention build on them, replace them or unintentionally make them harder to reproduce?
These are not questions about nostalgia for an earlier era. They are questions about whether an organisation can improve a system without first identifying what it should retain. The possibility that valuable capabilities were overlooked or displaced is one of the most consequential avenues for later analysis.
4.5. The apparent acceptance or normalisation of poor outcomes
Organisational performance does not depend only on the decisions of coaches or administrators. Schools have leadership groups, governing bodies, parents, players, old boys and other stakeholders, each with different interests and varying degrees of influence.
The persistence of Terrace's outcomes raises an additional question: why have disappointing results apparently failed to produce sufficient collective pressure for a fundamental reconsideration of the program?
That is not the same as claiming that everybody involved is satisfied. Individual players, parents, staff or administrators may be deeply dissatisfied, and private concerns are not always visible. Nor can we assume that there have been no internal reviews or attempts to change things. The observable issue is the apparent tolerance of a prolonged pattern of results and the limited evidence, at least in the public account, of a decisive departure from the broader approach.
This is especially interesting in a school environment, where the players bear many of the consequences but have little control over the system in which they participate. Each cohort has a limited number of seasons in which to benefit from the program.
The case therefore raises questions about organisational expectations, accountability and the point at which an undesirable outcome becomes accepted as normal.
4.6. A grassroots setting with lessons across sporting levels
Terrace competes in an exceptionally strong schoolboy rugby environment, but it remains a school rugby program. That makes the case different from the professional and international examples that dominate much of the discussion about sporting improvement.
Lessons drawn from successful professional teams often depend on conditions that other organisations cannot reproduce: full-time athletes, large budgets, extensive specialist support, recruitment markets and control over preparation. What works under those conditions may not translate readily to a school or community club.
The Terrace case begins at a level of sport familiar to a much larger range of organisations. Yet the underlying improvement problems are recognisable at every level: understanding existing capabilities, deciding where to invest, developing the players available, coordinating coaches and programs, and responding when results do not meet expectations.
The specific solutions are not universally transferable. A school program with annual turnover operates under different constraints from a professional club or national team. But the principles governing how organisations diagnose problems and attempt to improve can be examined across those environments.
There is also a useful irony in a grassroots organisation adopting features associated with professionalisation. It provides an opportunity to ask whether becoming more professional in appearance or structure necessarily means becoming more effective.
4.7. Longitudinal evidence rather than one exceptional result
Many familiar sporting case studies begin with a singular achievement. A team wins a World Cup or a championship, and the study then works backwards to explain why. Such success may reflect years of preparation, but the decisive outcome can ultimately turn on a handful of matches, unusual circumstances or single errors and moments that do not fully represent the organisation's underlying capability.
Terrace provides the opportunity to examine something different: the repeated operation of a program over many seasons. The principal post-intervention comparison extends across thirteen complete competitive seasons, with the treatment of disrupted years explained when the results are presented.
Across that period, successive cohorts arrived and departed, teams were assembled repeatedly, coaches and staff operated within the program, and the organisation had many opportunities to observe outcomes and respond. This is more informative about habitual performance than any single season, whether exceptional or disastrous.
The question becomes not simply whether Terrace could produce a good team once, but what level of performance and player development the program could repeatedly produce.
Thirteen seasons are not thirteen controlled experiments. Competitive conditions, opponents and student cohorts all changed. Nevertheless, the extended record provides a way to look for sustained patterns rather than mistaking a momentary result for a reliable organisational capability.
The difference between winning and losing a World Cup might come down to a single red card, a refereeing decision or an unfortunate moment of play. Thirteen seasons of underperformance against an established historical baseline demand a different kind of explanation.
4.8. The apparent ease of correcting many of the problems
An unsuccessful improvement program can leave an organisation with consequences that are difficult to reverse. Institutional knowledge may disappear, facilities and budgets may be tied to particular approaches, and an entire pathway may have to be rebuilt before a different strategy becomes possible.
One of my reasons for returning to Terrace is that many of the constraints and bottlenecks I have identified appear to be comparatively straightforward to address. Some concern how existing resources are used, how decisions are made, or how the program is organised. The argument is not necessarily that Terrace must dismantle what it has built; some elements of the intervention might work considerably better if incorporated into a different approach.
If that assessment is correct, the interesting issue is not merely that the original strategy may have been ineffective. It is that relatively accessible alternatives may have remained available while the outcomes continued to deteriorate.
This must be examined carefully. A decision can be easy to reverse on paper without its effects being easy to undo. Rebuilding capabilities, relationships or a development pathway can take time. And no later correction can return the lost opportunity to a student whose school rugby years have passed.
That distinction makes the case especially valuable: the cost of changing direction may be small relative to the cumulative cost of continuing along the same path.
4.9. A microcosm of problems found throughout rugby
I have also returned to Terrace because many of its apparent improvement problems resemble approaches seen elsewhere in rugby, including at higher levels of the game.
When results disappoint, organisations often respond by changing personnel, recruiting more talent, expanding specialist support, increasing training or adopting features associated with more successful programs. Some of those decisions can be entirely appropriate. The problem arises when they are treated as self-evident solutions without a sufficiently developed understanding of the system they are supposed to improve.
This is the broader concern with what might be called an anti-systemic approach: attempting to improve individual parts while paying too little attention to the relationships, constraints and existing capabilities that determine the performance of the whole.
Terrace is potentially a useful microcosm because its scale makes the relevant decisions and long-term outcomes more accessible than those of an entire professional or national rugby system. It offers a contained setting in which to examine problems that may be considerably larger and more complicated elsewhere.
The existence of similar practices in different organisations does not establish that they failed for the same reasons. But if recurring mechanisms can be identified and tested, the lessons from Terrace may become relevant to a much wider discussion about how Australian rugby, and rugby more generally, approaches improvement.
4.10. Multiple measures of effectiveness
The case is not limited to a single premiership table or a count of First XV victories. Terrace's history allows several different aspects of performance to be examined, including match results, winning percentages, ladder positions, representative player production and patterns of player development.
There are also measures that may help explain how the program operates. Player selection, positional movement, displacement and team continuity do not by themselves tell us whether a school has succeeded, but they can reveal aspects of the conditions under which its teams and players are expected to develop and perform.
This matters because different measures can point in different directions. A program might recruit stronger players without becoming better at developing the students already enrolled. It might improve participation while experiencing weaker First XV results. It might increase coaching resources without increasing its capacity to produce representative players.
Rather than allowing one statistic to define the entire program, Terrace offers the opportunity to ask whether apparent improvements in one area correspond to improvements elsewhere, and whether the outcomes most relevant to the intervention actually changed.
The existence of several indicators does not guarantee that they will all support the same conclusion. That is precisely why examining them separately is useful.
4.11. A relatively consistent competitive reference point
Longitudinal evidence is particularly valuable when there is some continuity in the environment against which performance is measured. Terrace has played in an established competition with recurring opponents, a broadly familiar season structure and a substantial historical record.
That provides an external point of reference. The question is not simply whether Terrace won more or fewer matches in absolute terms, but how it performed within a competition in which many of the same institutions appeared year after year. It creates opportunities to examine relative performance, recurring opponents and the strength of particular seasons.
This does not make comparisons effortless or perfectly controlled. The schools change too. Opponents recruit, lose players, improve their own programs and experience unusually strong or weak cohorts. Competitive standards may also shift over time. A nominally identical finishing position can therefore represent different levels of performance in different years.
Still, the repeated competition against familiar schools gives this case a more useful comparative foundation than a history assembled from disconnected competitions or constantly changing opponents.
It allows the analysis to move beyond the raw win-loss record and ask what Terrace's changing results mean in relation to the environment in which they occurred.
4.12. Different, potentially competing definitions of improvement
Finally, Terrace is an interesting case because the purpose of a school rugby program is not exhausted by the First XV's competitive results.
A school may reasonably value participation, student experience, academic achievement, player wellbeing, individual development, representative progression and competitive success. Those objectives may support one another, but they can also pull decisions in different directions. An approach that maximises one outcome may do little for another, or may even undermine it.
That complexity matters in a case where the school's traditional educational philosophy differed from the increasingly recruitment-driven culture of the competition. It matters even more when assessing a substantial intervention undertaken, as I understood it at the time, to improve First XV competitive results.
What, exactly, was the organisation trying to improve? What would have counted as success? Were the intended gains worth the costs and trade-offs? And how should benefits in one part of the program be weighed against disappointing results in another?
The purpose here is not to assume that winning should override every other educational consideration. It is to avoid allowing different definitions of improvement to substitute for one another after the fact. A program can succeed in some respects and fail in others. A serious assessment needs to say which is which.
Taken together, these characteristics explain why Terrace has remained such a useful reference point in my work. The case offers an unusually large intervention, an extended record of outcomes, existing capabilities worth understanding, and questions about prediction, persistence, accountability and the possibility of changing direction.
They do not provide twelve proofs of a single causal explanation. Nor do they require the reader to agree with every criticism of the program before finding the case valuable. Their importance is that they create a rich opportunity to ask what improvement actually requires, how we recognise it, and why serious efforts to achieve it can produce so little of what was intended.
5. Clarification One: What Does "Failure" Mean Here?
Before examining the intervention and its outcomes, two points need to be made clear. The first concerns the word failure.
When I describe the Terrace Rugby improvement program as a failure, I am not criticising the school for failing to win a GPS premiership. Nor am I arguing that Terrace should be judged by the same objectives, resources or recruitment practices as Nudgee College, Churchie or The Southport School.
Those are different questions.
Terrace has long placed great importance on academic achievement and a broad education. Its approach to sport has traditionally reflected the idea that participation is one part of school life, not its defining purpose. More recently, the college has described its sporting approach in terms of participation rather than performance.
That is a legitimate choice. A school is not obliged to prioritise winning rugby premierships, and there is no reason to assume that an educational institution must organise itself around sporting success.
But that distinction does not answer the question this case study is asking.
Terrace was a participation-oriented school before the improvement program, just as it describes itself as one today. Its educational priorities did not suddenly come into existence after the intervention.
The central comparison, therefore, is not Terrace against Nudgee, Churchie or any other school. Those opponents matter as competitive context, but they are not the benchmark against which the intervention is being assessed.
We are comparing Terrace after the improvement program with Terrace before the improvement program.
Before the changes, Terrace was generally able to compete around the middle of a strong GPS competition despite largely declining to recruit players in the way other schools did. It also had a substantial record of developing representative players. The intervention changed important aspects of how the program was resourced and operated. It is entirely reasonable to ask whether those changes produced better outcomes than the system they replaced.
There is also a responsibility to the players themselves. Students competing in this environment commit considerable time and effort to preparation, training and matches. That commitment deserves an effective program, whether or not winning a premiership is the school's highest priority.
The difference between regularly winning four matches in an eight-match season and regularly winning one is not trivial. Neither outcome necessarily brings a premiership. But they represent markedly different competitive experiences for the players undertaking the same essential commitment to the sport.
If players are committing comparable or greater time and effort, while the organisation is devoting substantially more resources to the program, the question of whether those investments are producing better outcomes becomes more pressing, not less.
This is not to say that winning is the only valuable outcome of school sport. Participation, enjoyment, retention, wellbeing and development all matter. Improvements in those areas should be recognised where they can be demonstrated. But invoking participation does not, by itself, explain a deterioration in First XV results relative to a period when participation was already central to the school's philosophy.
Participation is an objective, not an exemption from improvement.
The point is not that Terrace has failed because it is not the best rugby school in Queensland. The point is to ask whether the program became more effective than it was before the changes, particularly in the outcomes those changes were intended to improve.
6. Clarification Two: This Is Not a Response to 2026
The second clarification concerns timing.
Given Terrace's particularly disappointing 2026 season, it would be understandable for someone encountering this document now to assume that it is a reaction to those results. It is not.
As explained at the outset, I have used Terrace as a case study of coaching and organisational improvement for more than a decade. The underlying concerns, observations and questions long predate 2026. Some of the criticisms of the proposed direction were articulated before the intervention began.
The 2026 season did, however, influence when I chose to assemble and publish the historical account. Early in the pre-season, it was evident that Terrace was facing another difficult campaign. Rather than publish while that campaign was still unfolding, I waited for the final results so that the latest complete season could be incorporated into the analysis.
The 2026 First XV produced one of the worst competitive records in approximately a century of Terrace rugby. Only three teams had recorded worse results. Yet the 2026 team operated within a program benefiting from levels of investment, infrastructure, recruitment and professional support unavailable to many of its historical predecessors. When those advantages are considered alongside the results, there is a compelling argument that 2026 represents one of the most significant competitive underperformances in Terrace's rugby history.
Terrace won just one match that season. Beyond the historical win-loss record, 2026 also raises questions about selection, player displacement, team stability, the composition of the playing group and the competitive context in which the results were achieved.
It may even warrant a detailed comparison with 2011, the historically disastrous season in which Terrace failed to win a match. Such a comparison cannot rest on the number of wins alone. The two seasons differed in important ways, including the extraordinary disruption caused by the 2011 floods, the availability and quality of opposition, playing arrangements and the characteristics of the teams involved.
That comparison is worth exploring separately. So are the unusual displacement and instability measures associated with 2026. Neither belongs in detail in this introductory document.
More importantly, the question of whether the improvement program achieved its intended outcomes was already well established before 2026 began. Another year of evidence is useful, but the central argument does not depend on treating one bad season as uniquely decisive.
The 2026 season did not create the case for examining Terrace's improvement program. It added another year of evidence to a case that was already well established.
7. How Improvement Will Be Assessed
The basic test applied throughout this study is deliberately uncomplicated:
Did the improvement program produce better outcomes than the program it replaced?
That does not require Terrace to win a premiership, dominate the GPS competition or match the resources of other schools. It requires the intervention to deliver some meaningful improvement against Terrace's own historical performance.
The central purpose of the changes, as understood through my involvement with the organisation and extensive discussions at the time, was to improve the competitive results of the First XV following the disastrous 2011 season. It is therefore appropriate to place considerable weight on First XV outcomes when evaluating the intervention.
Those outcomes can be examined in several ways: matches won, winning percentage, finishing position, performance relative to opponents and changes over longer periods rather than individual seasons. Representative player production and developmental outcomes provide other important measures of what the program has achieved. Some measures, such as selection continuity and player displacement, may also help illuminate how the program has operated, even when they are not direct measures of success in themselves.
No single number needs to carry the entire argument. Nor should one exceptional season, whether successful or unsuccessful, determine the verdict on a long-running program.
There is a further consideration. Merely maintaining previous performance would not necessarily justify the intervention. The changes involved considerable commitments of resources, alterations to established practices and departures from aspects of the school's traditional approach. Those decisions came with financial, organisational and cultural costs, as well as opportunities that might have been pursued differently.
If an organisation invests substantially in order to improve, it is reasonable to expect a return in the outcomes the intervention was intended to affect. The return need not be spectacular. A sustained gain of even one additional win per eight-match season would represent meaningful competitive improvement. But a program that becomes more expensive and more disruptive without becoming more effective raises a different question from one that simply fails to achieve a premiership.
The resources needed to achieve exceptional performance should not be confused with those needed to achieve reasonable performance. A school might require substantial additional capability to dominate a strong competition. It does not follow that moving from routinely winning one match to winning four requires more recruits, facilities, staff or training time. Where those resources already exist, persistent underperformance may reflect the program's inability to convert existing capability into results. Improving the conditions in which a program operates is not necessarily the same as improving the performance of that program.
This does not mean that benefits elsewhere should be ignored. Better player wellbeing, lower injury rates, improved participation or a better student experience would be meaningful outcomes if demonstrated. Equally, improvement in one part of the program does not automatically establish that the First XV intervention achieved its principal purpose. Different outcomes should be considered on their own terms.
There is also a difference between observing improvement and proving what caused it. A stronger or weaker result after an intervention does not establish that the intervention alone produced the change. Changes in player cohorts, opponents and the broader competition must be considered. The results portion of this document will explain relevant comparisons and exceptional seasons where they arise, rather than introducing a separate technical methodology here.
Ultimately, the assessment has two connected elements: Did Terrace become more effective at producing the outcomes it sought, and were any gains sufficient to justify what was invested and changed?
8. How the Wider Case Study Will Unfold
This contextual report has three substantive parts. Part One establishes the setting, purpose and boundaries of the case. Part Two examines the changes surrounding the improvement program. Part Three presents the subsequent results and relevant historical comparisons.
Together, those parts will form a common contextual reference. They are not intended to be the last word on the Terrace experience.
The wider case study will develop through Rugbyology over time, in whatever form best suits the particular subject. A short LinkedIn post about an improvement principle might use one Terrace example. A longer essay could explore the consequences of a particular recruitment practice. A video or podcast might examine selection stability, player development, system design or how organisations respond when their assumptions are challenged.
Some of those pieces may form recognisable sequences, but there is no intention to force the material into a numbered series with a predetermined endpoint. Terrace will instead be a recurring thread running through a broader body of work about improvement.
I intend to make the foundational account available both as a readable publication on the Rugbyology website and as a downloadable, formatted PDF. A dedicated case-study page can also provide a permanent home for the work, linking the historical document to related essays, videos and shorter pieces as they appear.
This should work in both directions. Readers who encounter a Terrace example elsewhere can return to the historical context. Readers who arrive through the foundational document can explore the arguments, concepts and questions that grow from it.
The document itself will not be frozen. This is a first version, not an attempt to produce a final account that can never be amended. New information may clarify the original intervention. Additional measurements may become useful. Further competitive seasons will add to the record.
Those future results should be incorporated regardless of whether they reinforce or challenge earlier interpretations. A successful season deserves to be recorded just as much as a disappointing one. At the same time, one good or bad season should be considered in the context of the longer pattern rather than treated as a complete explanation by itself.
Over time, the related material may become substantial enough to justify a more comprehensive consolidated study. For now, the priority is simpler: establish the history, make it accessible and allow the analysis to develop without requiring the entire project to be finished before anything can be published.
9. The Central Questions
Two questions sit at the heart of this work.
The first is specific to Terrace: How can an organisation make such a substantial and sustained effort to improve its rugby program, yet fail to achieve meaningful competitive improvement, with several important outcomes actually deteriorating?
The second is larger: Do sporting organisations genuinely understand how improvement occurs, or do they too often make changes based on assumptions about what ought to make them better?
These questions are connected, but they are not interchangeable. The first asks us to examine a particular history, its decisions and its results. The second asks what that history can teach us about the way organisations diagnose problems, recognise their existing capabilities, choose interventions and respond to evidence.
This contextual report does not settle those questions. It establishes the background and outcomes needed to examine them carefully. Later Rugbyology work will take up individual mechanisms, explanations and lessons.
The hope is that the experience of one school rugby program can help people working in very different environments ask better questions about their own efforts to improve.
Terrace is the case study. Improvement is the subject.
Part Two: The Magnificent 7.5
What Terrace changed, and what changed for Terrace
One of the most important things about the Terrace rugby improvement program is the number of circumstances that changed in its favour.
Some of those changes were instigated by Terrace. In a seismic redefinition of its rugby program, the College began recruiting players, brought athletic development inside the school, invested in sporting facilities and moved towards a more supportive institutional attitude to rugby and professional sporting careers. Other changes were not Terrace's decisions at all. A GPS calendar reform removed an unusually restrictive preparation timetable, while changes to Queensland school-entry ages helped alter the composition of senior teams. A further change to recruitment protocols may also have affected the competition, although its practical significance is less certain.
I have grouped them as the Magnificent 7.5. The title is a little playful; the opportunities were not. This is a selection of the most consequential developments, not a claim that seven and a half independent reforms were implemented as part of a single plan. They occurred at different times and reinforced one another in different ways. Nor is this a complete account of all changes made by or to Terrace during this time period.
The purpose here is to establish what changed, what those changes made possible and why better outcomes were a reasonable expectation. The results come in Part Three. The relevant benchmark is not a premiership every year. It is whether Terrace became more successful than it had been under the more restrictive conditions that preceded these changes.
A. What Terrace changed
1. Recruitment: the reversal of a defining identity
Of all the changes, none was more seismic than recruitment.
For the entirety of the program's history before 2012, the College's refusal to recruit players or offer sporting scholarships was far more than an enrolment position. It was a defining part of the rugby program's identity. Terrace is a Catholic school established by the Christian Brothers in the Edmund Rice tradition. I used to joke that, in rugby conversations, Terrace identified more strongly with not offering sporting scholarships than it did with Catholicism.
It was a joke, but it captured something real. From roughly 1997 until 2012, the subject seemed to be everywhere. “We don't import” was a familiar expression of how Terrace saw itself. You heard the point from players, coaches and people around the program, often without anybody needing to explain it.
It performed two quite different functions. For players, not recruiting was a source of pride. They were competing with the boys who had come through the school, not a team assembled late in the piece to win a premiership. For coaches, in particular those who were responsible for unsuccessful seasons, it was also a ready-made explanation for why Terrace struggled against opponents who could bring established players into their senior teams. In my experience, the absence of scholarships was invoked so often that it had become a built-in explanation for disappointing results.
Those two attitudes were not contradictory. A policy can be both a principle worth defending and a limitation used to explain failure. That combination is part of what made the eventual reversal so extraordinary.
This was a defended position, not an accident
The College's stance was sustained across successive leadership periods. Its own publications had acknowledged the competitive difficulties posed by recruiting rivals while defending an educational commitment to develop the students already enrolled. Relevant arguments appear in the 1998, 2000 and 2006 yearbooks. In 2010, the issue became part of a wider public debate about schoolboy rugby, scholarships, fairness, the displacement of existing students and physical mismatches.
These were genuine educational questions. A school is more than its First XV, and the arguments against recruiting accomplished players into a student cohort cannot simply be dismissed as indifference to sport. Terrace had made a deliberate choice about the kind of school sporting program it wanted to operate.
Yet the competitive consequence was obvious. A team might lack depth or established quality in a few important positions, but Terrace could not address those gaps simply by bringing in players from elsewhere. It had to develop students already at the school, change the way it played or accept the limitation.
I remember a coach telling me after a difficult season that his side was three players away from being “world beaters”: a prop, a powerful back-rower and a centre. To my mind, that was not a credible diagnosis of the team's difficulties. There was already enough playing capability to achieve considerably better results. The supposed absence of three players offered a much simpler explanation than examining how the available players were coached, organised and used. Different coaches used different examples, but the central refrain was remarkably consistent: Terrace was struggling because it could not recruit the players its rivals could.
Recruitment had become Terrace rugby's ultimate “if only”.
If only we could recruit a prop. If only we could find a dominant ball-carrier. If only we could bring in the players that other schools could.
Therein lay the beauty of the excuse. It could be used to explain almost any disappointing season, while preserving the belief that success was only a few recruited players away. And while Terrace refused to recruit, it could never be tested. The missing players could always be imagined as the difference.
In 2012, the “if only” became testable
After the exceptionally difficult 2011 season, Terrace began recruiting players for rugby. Four or five recruited players arrived in 2012. That was not a marginal adjustment to an existing practice. It reversed something the College had defended, and the rugby community had discussed almost constantly, for years.
Subsequent First XV teams also frequently included players who had entered the College after the normal secondary-school intake. My review suggests four or five such players in many teams, with any recruitment at the normal entry point potentially additional to that figure. Later enrolment does not, by itself, establish a sporting scholarship, a fee concession or even a recruitment approach, and the precise arrangements should not be inferred from an enrolment date. The change in Terrace's ability to seek established talent, however, was unmistakable.
I have sometimes compared the old position to an amateur side competing in a professional competition. That is an analogy, not a claim about the formal status of any GPS school. The point is the asymmetry. Other teams had a mechanism for addressing weaknesses in their playing roster that Terrace, by choice, did not use. In 2012 Terrace decided it could use that mechanism too.
Consider what that should have meant. Recruitment did not need to transform every player or guarantee a premiership. Even an additional win or two a season would represent a substantial improvement across an eight-match GPS season. Given how frequently coaches had previously identified the inability to recruit as a decisive competitive handicap, expecting a meaningful improvement once that handicap was removed was hardly an extravagant proposition. One or two wins is an illustration of the scale of expectation, not an estimate of recruitment's measured effect.
There is a second reason the expectation was so strong. The existing Terrace program was not incapable of developing players. In 2010, before recruitment was introduced, Terrace produced an unusually strong group of representative players. It produced more Australian Schoolboys than seven of its eight GPS opponents.
In other words, Terrace appeared to have something worth preserving: an ability to develop players from within its own cohorts. Recruitment offered the possibility of adding accomplished players to that capability. It should not have been necessary to choose between developing existing students and addressing particular positional shortages through recruitment.
The explanation became especially testable when coaches who had worked in the non-recruiting era, including the coach responsible for that three-player assessment, later had the opportunity to work with recruited players. The anticipated transformation did not follow. That subsequent experience does not prove that recruitment was incapable of helping, but it seriously challenges the earlier claim that a few missing players were the decisive problem. The results in Part Three allow the broader explanation to be examined against the record.
Recruitment therefore deserves more weight than any other item in this section. The College did not merely gain another resource. It abandoned a defining position, removed a limitation that had long dominated explanations of poor performance and acquired the capacity to build teams in ways previously unavailable to it.
2. Bringing athletic development inside the College
The second change was the creation of an internal Athlete Development Program in 2012. It was a sensible decision in principle, and one I supported. But to understand its significance, it is important to understand what Terrace was replacing.
From rudimentary preparation to a specialist service
In the late 1990s, Terrace had a very basic gym and its strength and conditioning program was overseen by the team physiotherapist. It was an early period in the development of physical preparation in school rugby. Physiotherapy expertise did not necessarily translate into specialist strength and conditioning knowledge, and the gap was sometimes striking. There were instances of people responsible for physical preparation borrowing training resources from players to learn the fundamentals themselves.
After a particularly poor period in the early 2000s, Terrace engaged an external company to provide athletic preparation. For approximately eleven years, the company delivered a strong, well-structured program. Its coaches were qualified, had specialist knowledge and experience, and had skin in the game. It was a substantial advance on the arrangements of the late 1990s.
That history matters because 2012 did not mark the arrival of professional athletic preparation at Terrace. It marked a decision to bring an existing specialist function inside the school.
There was also a familiar pattern in the explanation for disappointing rugby results. Coaches had long pointed to recruitment as a constraint. In the years leading up to 2012, the external nature of the athletic development service also became a target of criticism. The external provider unfairly bore the brunt of blame for disappointing rugby results, even though those results did not, by themselves, demonstrate shortcomings in the athletic preparation service. This was a recurring pattern involving external providers within Terrace rugby.
An external service is not necessarily an inferior service. Its quality depends on the people, expertise, organisation and decisions involved, not on whose payroll the coaches occupy.
Why bringing it in-house made sense
Nevertheless, I regarded the decision to create an internal development pathway as the right one in principle. Terrace had the opportunity to design a physical development curriculum across the entire schooling journey, from students' early years through to senior competition. Rather than organising preparation around successive First XV seasons, the College could build physical capabilities progressively over many years.
There were other potential advantages. Its own staff could coordinate more closely with rugby coaches, other sports and the school calendar. They could connect physical preparation to the specific demands of training and competition, and maintain continuity as cohorts moved through the school and across sporting seasons.
The College would also have more direct responsibility for staffing, standards, planning and delivery. That created opportunities for improvement, but it also meant that poor appointments or decisions could have greater consequences for the program.
Combined with recruitment and improved facilities, this created a considerable opportunity. Terrace could not only access a broader pool of players, but also take greater control over their long-term preparation and that of the students already enrolled.
The distinction between the decision and its execution
The important qualification is that internalising a service does not, by itself, improve the service. Terrace was not replacing an absence of specialist provision. It was replacing a capable external provider. The opportunity was to retain or exceed that level of expertise and gain the advantages of integration and continuity.
I believed the organisational decision was sound, but I had serious reservations about elements of its implementation. Those concerns should not be confused with opposition to creating an internal program. They are concerns about how the opportunity was used.
The consequences were not always hypothetical. Following the change, training decisions on some occasions adversely affected players' availability for matches or left groups poorly prepared to compete. Problems of training load and readiness could also distort evaluations of players being considered for the First XV. These episodes illustrate a wider danger: adding specialist provision can introduce new problems if its work is not competently managed and integrated with the competitive program.
This distinction becomes important when we later examine outcomes and the actual operation of the program. For now, the relevant point is that Terrace had made another deliberate change that should have improved the conditions for developing and preparing rugby players. Whether greater organisational control produced greater athletic capability is a separate question.
3. Facilities: from the bottom of the competition to a claimed world-class standard
The third major change was the scale of the College's investment in sporting facilities. Once again, the contrast with the conditions under which earlier Terrace teams had operated was substantial.
Before the upgrades, Terrace’s gym and athletic development facilities were basic and among the worst in the GPS competition. By approximately 2013, that situation had been transformed. Terrace had developed a first-rate strength and conditioning facility that would have placed among the best in the competition. This was already a significant improvement in what the College could provide for its athletes.
But Terrace did not stop there. Having moved from one of the least impressive gyms to one of the strongest, it went on to replace that arrangement with something more ambitious. The College acquired an additional property across from the main school and developed a new Performance Centre. A standalone two-storey gym was opened in 2022. The earlier upgrade had not merely brought Terrace up to an acceptable standard; the College was prepared to supersede an already first-rate facility with another major investment.
The College has been explicit about the standard it believes it now provides. In its published account of the Performance Centre, Terrace says that students have access to “world class strength and conditioning equipment”. It presents that equipment as part of an Athlete Development Program intended to teach young athletes to move efficiently and safely. “World class” is the College's own description, but the significance for this case is unmistakable: Terrace was not presenting its physical preparation resources as merely adequate. It was promoting them as exceptional.
The investment was not confined to gym buildings and equipment. There were extensive improvements to playing and training fields, new grandstands, lighting, changing facilities and supporting technology. Some of the Tennyson work followed the 2011 floods and necessarily involved restoring damaged facilities; other works went beyond restoration. Both contributed to a substantially different sporting environment, although their direct effects on rugby preparation varied. A grandstand, for example, is not the same performance resource as a gym or training field.
There is no audited aggregate figure in this study for the facilities and related program expenditure. But the successive gym developments, the acquisition of additional property, field works and associated infrastructure alone represent investment running into millions of dollars. The ongoing cost of specialist athletic-development and coaching staff over the subsequent years adds another substantial commitment. Not every dollar was spent exclusively on rugby, but these were resources and capabilities from which the rugby program could benefit.
This mattered in combination with the internal Athlete Development Program. Terrace now had both greater control over how young athletes were prepared and much stronger physical resources with which to prepare them. Appropriate equipment, space and training environments could support progressive physical development, a broader range of activities and more effective preparation over successive school years.
None of this means that better buildings guarantee better rugby players. A first-rate facility still has to be used well. But that is precisely why the investment belongs in this account: Terrace had removed another substantial limitation and had given itself the physical conditions in which better preparation should have been possible. Whether that opportunity translated into improved outcomes is a question for the results and analysis that follow.
4–5. A different attitude towards rugby and sporting careers
The fourth and fifth changes are best understood together. They concern the relationship between a school and its rugby program: what it believes sport is for, how much competitive success matters and how much support it is prepared to give those pursuing it. They were separate developments on different timelines, but they describe a broader shift in Terrace's institutional priorities.
For much of the period before the improvement program, the College's senior leadership was strongly resistant to rugby's competitive ambitions. This was more than a reluctance to recruit players. At times, rugby's prominence within the school itself seemed unwelcome, to the point where it was reasonable to wonder whether the leadership would have preferred a much smaller role for the sport, or none at all. That is a description of the institutional environment, not a claim to know anyone's private intentions.
In 2015, leadership changed. The incoming Principal had an extensive rugby background as a player and coach and subsequently served on the Queensland Rugby Union board. Whatever one thinks of the priorities of either administration, the contrast in their enthusiasm for rugby was considerable. A program that had often seemed to be pushing against its institutional environment was now operating within one much more sympathetic to the sport and its ambitions.
Sport as preparation for a career
An important change in thinking had already emerged after the disastrous 2011 season, before the leadership transition. I remember the Director of Sport telling me that Terrace needed to recognise professional sport as a legitimate career, and therefore something the College should prepare students for and support.
That statement signalled a different conception of the school's responsibility. Supporting a talented rugby player was no longer necessarily a concession to the desire to win schoolboy matches. It could also be understood as helping a student pursue a serious future, just as the College supported academic, artistic and other vocational ambitions.
That change in philosophy helped make a more ambitious sporting program institutionally intelligible: specialist preparation, pathways, coaching and investment could be justified as educational opportunities as well as competitive advantages.
Neither supportive leadership nor the acceptance of professional sport as a career automatically wins rugby matches. Their importance lies in what they permit an organisation to attempt. Senior decision-makers influence budgets, staff appointments, enrolment policies, facilities, the legitimacy of competitive ambition and the freedom of coaches to pursue it. Their outlook can either enable or constrain almost everything happening at program level.
This kind of organisational change is harder to put a number on than the arrival of four recruits or a new gym. But the same GPS competition offers two striking reminders that the institutional environment can matter enormously.
Nudgee College: a rapid reversal
In 2004, Nudgee College finished last among the schools competing for the GPS First XV premiership, recording zero premiership points. This was an extraordinary result for a school with such a formidable rugby history. The following year it recovered to equal fourth, and in 2006 it was equal second. Across 2008 to 2026, Nudgee won or shared eleven GPS First XV premierships.
That collapse coincided with a deliberate reduction in the institutional emphasis placed on rugby, including a retreat from recruitment and some of the measures previously taken to support success. The direction provoked strong resistance within the school community, as has been publicly documented, and was quickly reversed.
Nudgee's subsequent recovery does not mean that leadership and recruitment policy alone explain every result. But the sequence demonstrates why the institutional priorities governing a rugby program deserve attention alongside what happens on the field.
Brisbane Grammar: from a record defeat to a premiership, and back again
Brisbane Grammar offers a different but equally dramatic sequence. On 13 June 2009, its First XV lost 103–0 to Brisbane State High. Three seasons later, Grammar won the 2012 GPS rugby premiership.
That turnaround coincided with a more ambitious institutional approach to rugby. That direction was not sustained. Grammar's subsequent results were markedly different: in the eleven seasons from 2013–2019 and 2022–2025, its First XV won just ten of eighty-eight premiership matches, including four seasons without a win or draw.
Again, the competitive sequence is not in itself proof that a leadership decision caused the rise or fall. Playing cohorts, coaching, recruitment, opponents and other factors may also have contributed. Yet these are not trivial fluctuations. They are changes of a magnitude that ought to make us interested in the organisation around the team, not only the players and coaches on the field.
What the comparison means for Terrace
These examples are not offered as a formula in which a pro-rugby Principal produces wins and an anti-rugby Principal produces losses. In fact, they make the question more interesting. A school's leadership and philosophy can alter the conditions for success without determining whether its program uses those conditions well.
For Terrace, the central point is that the institutional environment changed considerably. By the middle of the improvement period, the College was more accepting of high-level sporting ambition and had leadership far more sympathetic to rugby. That should have made it easier to pursue, resource and sustain improvement than under the previous arrangements.
The contrast between Terrace's two leadership eras becomes more revealing when set beside the results in Part Three. Here, the change itself deserves recognition as a potentially significant advantage, even though its effects cannot be isolated as neatly as those of recruitment, facilities or preparation time.
B. What changed for Terrace
The next developments were not simply decisions made by the College. They were changes to the wider conditions in which Terrace had to compete. They matter precisely because a change available to every school does not necessarily benefit every school equally. A common rule can remove a disadvantage that one participant previously felt more acutely than others.
6. A legitimate preseason: a competition change that particularly benefited Terrace
The sixth change was not made by Terrace at all. It came from a reorganisation of the GPS sporting calendar, and I would argue that no school stood to benefit more from it than Terrace.
Under the old calendar, GPS rugby and football shared an extended winter season, with fixtures spread across Terms 2 and 3. For a school like Terrace, that arrangement made it remarkably difficult to prepare a First XV properly. Many of the players likely to make the side were committed to other sports in Term 1, particularly rowing and cricket. They were not simply sitting around waiting to begin rugby training. They were participating in the wider sporting life of the College, as they had been encouraged to do.
In practice, Terrace's rugby preseason often began in earnest only around Easter. There was a familiar moment around Easter when the First XV squad could finally come together and begin serious contact work. With premiership fixtures approaching, that could leave approximately four weeks of meaningful preparation. Four weeks to build combinations, introduce and rehearse structures, prepare for contact and turn a collection of schoolboys into a competitive senior team was an extremely narrow window.
Why the disadvantage was not shared equally
Every school operated under the same GPS calendar, but they did not all have the same ability to work around it.
Terrace was a day school and, until 2012, did not recruit rugby players. Its First XV candidates were predominantly students moving through the College's ordinary sporting program, many of them active in several sports. Other schools had boarding populations, established rugby recruitment arrangements or a greater capacity to assemble groups of rugby-focused players outside the short formal preseason. They could begin preparing players and combinations considerably earlier in the year.
There was another difference that the official school calendar did not capture. Schools recruiting established representative players were also more likely to have groups of students involved in state academies, national talent squads and other external development pathways. Those players might receive additional coaching and physical preparation throughout the year, independently of their school's formal rugby season. Terrace had far fewer players in those pathways, while its longstanding culture also discouraged playing club rugby during the school year. For many Terrace students, the short school preseason was consequently the beginning of their concentrated rugby preparation, not merely the final stage of work already underway elsewhere.
That distinction matters. The rules governing competition dates were common to everyone, but the practical preparation opportunities were not. Some rivals had ways of accumulating substantially more rugby preparation before the season. Terrace, by the nature of its program and student population, was more constrained by the timetable.
The decisive calendar reorganisation was already in place in 2012, when GPS rugby moved into a consolidated Term 3 competition. Terrace's First XV held a preseason training camp beginning on 10 April that year, while Round 1 did not take place until 21 July. Those dates show how different the preparation opportunity had become from the previous arrangement, under which meaningful First XV preparation had often been compressed into roughly four weeks.
By separating rugby from the earlier winter timetable, the reform created what Terrace had so often lacked: a legitimate, sustained period of preparation before the rugby premiership. Players could finish their Term 1 commitments, join the rugby squad and still have substantially more time to develop contact readiness, team combinations, tactical organisation and match preparation. Trial matches could become part of a progressive build-up rather than something squeezed into the final weeks.
The scale of the difference is easy to illustrate. In 2026, Terrace played a trial match approximately eight weeks before Round 1. Under the old arrangement, the first trial was generally no more than four weeks before the premiership began, and eight weeks out was before the period when Terrace would usually have begun serious First XV preparation. In the new arrangement, the team was already sufficiently organised to play a match. That is a fundamentally different preparation opportunity.
An unusually favourable equalisation
The calendar reform benefited all GPS rugby programs, but equal access to more time did not mean an equal increase in opportunity. A school already preparing its rugby players for much of the year gained additional official preseason time. Terrace gained something closer to the sustained preparation period it had historically been missing altogether.
This is why I regard the calendar change as an unusually significant equalising development for Terrace, and arguably one from which it benefited more than any other school. It removed a structural disadvantage without requiring Terrace to develop an entirely new capability or make an equivalent investment of its own.
It also arrived alongside the decision to recruit and to bring athletic development inside the College. A longer preseason meant more time to integrate incoming players, coordinate physical preparation with rugby training and prepare the First XV as a functioning team. These changes were potentially complementary, not merely additive.
Whether Terrace made effective use of the extra time is a different question. The calendar did not make its coaches better or ensure that their preparation was well designed. But it provided an opportunity that the earlier Terrace teams, operating within a much shorter practical preparation window, simply did not possess.
7. Age parity: the extra year Terrace seldom used
The seventh change requires an explanation of how the competition worked, because the disadvantage was not simply that Terrace sometimes had younger players. Other schools had ways of making systematic use of an additional year of rugby eligibility that Terrace rarely used.
Under the GPS Open rules, a player could compete in the First XV during the calendar year in which he turned eighteen, but not the year in which he turned nineteen. Most Queensland students traditionally finished Year 12 during the year in which they turned seventeen. That left an important distinction between the age at which students ordinarily graduated and the oldest age at which they were eligible to compete.
Terrace was generally reluctant to have students repeat a year of schooling. Consequently, the few First XV players in their under-eighteen age-grade year (turning eighteen that calendar year) tended to be students who had started school later than their peers or had repeated a year much earlier in their education. An eighteenth-year player was the exception, not the standard.
Two ways of gaining an extra year
Recruitment gave other schools much greater flexibility to use that remaining year of eligibility. In GPS rugby, there were two particularly important arrangements.
The first was placing an incoming player in a lower year level. A player who would ordinarily have entered Year 11 could instead be enrolled in Year 10. That created the possibility of three seasons of senior rugby rather than two, with the oldest coming when he was physically more mature and potentially much more valuable to the First XV. A school could explain the decision as allowing the student additional time to settle academically and socially. That might be a legitimate educational consideration; it did not remove the sporting benefit.
The second was extending Years 11 and 12 over three years. Schooling flexibility existed for students whose circumstances justified an altered pathway, such as serious disruption to their education or exceptional commitments outside the classroom. In my recollection, rugby programs could also use that kind of arrangement when recruiting talented players. A student could enter Year 11 at the expected age but spread the senior academic program over three years rather than two.
That arrangement offered two distinct advantages. It could keep a player in the school and eligible for a third senior rugby season. It could also spread the academic demands of senior schooling across a longer period, potentially creating more time for training, recovery and preparation. It was not difficult to present this positively to parents: the student would have more time to manage both education and sport.
I have used the word loophole for these practices, but it matters what that means. An arrangement can be permitted by educational and competition rules and still create a significant competitive asymmetry. The examples were established practices within GPS rugby; they are not an allegation that a particular school broke eligibility rules. The precise historical conditions governing entry-year placement and extended senior study are separate from the competitive asymmetry described here.
The difference on the field
The result was that some rival First XV teams could field a substantial number of players in the oldest eligible age group. Seeing roughly half a squad in that category was not remarkable. At Terrace, it was unusual to have even one such player in a season.
Historical Terrace figures put the difference in perspective. Before the improvement program, players turning eighteen during the season accounted for approximately 0–6% of First XV player-match appearances. The corresponding figure was 50% in 2026. These are figures from my analysis of Terrace teams, not an independently measured comparison of every GPS opponent; the final publication should specify the underlying seasons and whether appearances count starts or all players who took the field.
Those numbers matter because of what one year can mean during adolescence. A player who is physically outmatched at seventeen may be considerably heavier, stronger, faster or more resilient a year later. The gains are not uniform, and age does not replace ability, but in a collision sport they can transform a player's capacity to compete.
At Terrace some of the relatively rare older First XV candidates had not been consistently outstanding schoolboy players on their way through. They had played in the D, E or F grades and only came into First XV contention after another year of physical development. Other players progressed from lower grades at their schools to serious representative contention following an additional year of growth, maturation and schooling. In some cases, the progression ultimately led to professional careers.
How Terrace gained age parity
Two developments changed Terrace's access to older players, and they were not the same kind of change.
Recruitment, beginning in 2012, was a decision Terrace made. It allowed the College to enrol some players who could compete in their final eligible age-grade year, narrowing an advantage rivals had historically used more extensively.
Queensland's school-entry reform was a change made outside Terrace. The introduction of Prep in 2007 was accompanied by a shift in school-entry ages. The first, smaller Prep cohort completed Year 12 in 2019 and was, on average, around six months older than the preceding graduating cohorts. The following full cohorts continued under the newer age arrangements. It was the change in age at school entry, rather than merely adding a preparatory year, that mattered to the eventual age of senior students.
The second development was available to all Queensland schools, but its consequences were not necessarily equal. A school that had already been using recruitment and flexible progression to field older players had less of a historical gap to close. Terrace, which had previously depended overwhelmingly on ordinarily progressing younger students, had much more ground to make up.
In 2026, half of Terrace's First XV playing appearances came from players in that oldest eligible year, an exceptionally high proportion against the school's historical record. Before the improvement program, the corresponding share was generally only a few percent. The 2026 figure may prove to be a record high when the full historical series is checked, but its scale is already clear. This does not establish that an older side should win every match. It does establish that another of the long-standing constraints under which Terrace had competed was substantially reduced.
Given the physical differences between seventeen- and eighteen-year-old rugby players, this alone provided a meaningful reason to expect better competitive performance. Combined with recruitment, improved athletic preparation, facilities and a longer preseason, it represented a particularly consequential change in the conditions available to Terrace rugby.
7.5. Recruitment governance: the uncertain equaliser
The last change is the half in the Magnificent 7.5, not because its potential significance was small, but because it is difficult to know how consistently the new expectations were applied.
Before the improvement program, recruitment across GPS rugby operated with very little effective oversight. The absence of a reliable, common restraint meant that some programs could go to extraordinary lengths to strengthen their First XV.
How far the old practices could go
One incident associated with the 2010 controversy illustrates the environment. One GPS school recruited a player who had already completed secondary school in 2009. He had captained his previous school and played representative rugby. He was brought back into school for another year, with rugby the obvious competitive attraction. This was not an ordinary case of identifying a promising Year 9 or Year 10 student and offering him an educational opportunity. It involved an accomplished school-leaver returning to the school system in circumstances that also made him available for senior rugby.
The example matters because it shows the extent of the advantage that could be obtained through recruitment arrangements, rather than because it establishes that a particular eligibility rule was broken. It also gives context to the highly public debate about schoolboy rugby recruitment, educational purpose and player safety around that time.
An attempt to put boundaries around recruitment
As the controversy developed, there were efforts to establish protocols among headmasters about recruitment and eligibility. These included expectations that a recruited student would have to attend his new school for a period before becoming eligible to represent it. The exact arrangements and how strictly they operated varied or were disputed; this was not a uniformly policed recruitment regime.
A useful piece of contemporary evidence is a 2022 letter from the Associated Independent Colleges (AIC) to the GPS. It describes recruitment protocols established following discussions in 2013, but argues that they had subsequently been ignored. That is revealing in both directions: there were attempts to govern the practice, yet the existence of an agreement did not necessarily mean that schools complied with it or that anybody consistently enforced it.
Nor were restrictions automatically favourable to Terrace. One of Terrace's early recruits was ruled ineligible, so the new arrangements could create an immediate disadvantage for a school only recently entering the recruiting environment. But across a competition in which other schools had long used recruitment aggressively, common limits could also reduce advantages those schools had previously enjoyed.
That is why I count this as a possible improvement in the conditions Terrace competed under, rather than a Terrace improvement initiative. The College did not need to recruit more players or spend more money to benefit from opponents having fewer ways to recruit theirs. The potential effect was another narrowing of the gap between the non-recruiting Terrace of the past and the stronger, better-resourced teams it had been required to face.
I cannot quantify how much the protocols changed the behaviour of rival schools. That uncertainty is precisely why the item earns only half a place. But it would be misleading to leave out an attempt to regulate one of the most important historical competitive advantages in the GPS, particularly when assessing how much Terrace's circumstances had changed.
The scale of the opportunity
Consider the contrast between Terrace rugby before 2012 and the conditions that subsequently became available.
A program that had defined itself partly through not recruiting could now seek established players to fill gaps. Those players could enter a more coordinated athletic development structure with better equipment and facilities. The First XV had a much longer period to prepare for competition. It increasingly drew on older players, and it operated within an institution more willing to recognise and support serious sporting ambitions. Recruitment protocols may also have narrowed some of the advantages previously enjoyed by rivals, though the extent of their practical enforcement remains uncertain.
Importantly, these changes were not seven unrelated additions. They could reinforce one another. More talented recruits were potentially more valuable with time to integrate; longer preparation was potentially more valuable with stronger physical-development systems; those systems were potentially more effective with better facilities and institutional support.
Nor should the changes all be credited to Terrace's decisions. Some required the College to act. Others simply improved the conditions in which it was operating. For an evaluation of outcomes, however, both are relevant. The historical program had competed under restrictions that the later program no longer faced to anything like the same extent.
The most striking example remains recruitment. For years, the absence of scholarships and player recruitment had been an identity, a source of pride and an almost universal explanation for why results were difficult to achieve. Then the College changed course. If that explanation had been substantially correct, removing the constraint should have mattered. And recruitment was only the first of several major changes.
It is worth considering these changes individually, not only as a collective package. Several were substantial enough to support a reasonable expectation of improvement on their own. Removing the recruitment disadvantage, gaining access to a genuine preseason, or fielding a more mature team could each plausibly have been worth an additional victory over the course of a season. That is not a statistical forecast. It is a measure of how modest the improvement threshold was compared with the scale of the opportunities.
The benefits cannot simply be added together, and the changes took effect at different times. But as they accumulated, the case for expecting some improvement became stronger. Terrace did not need to win a premiership to demonstrate progress. It needed to become better than the Terrace teams that had operated without those advantages.
None of this guarantees victory, proves that the changes were implemented well or assumes that the earlier program's strengths were preserved. Those are different questions. Nor is the flood-disrupted 2011 season an appropriate ordinary baseline for judging long-term improvement.
Removing disadvantages is not the same as creating advantages
Not all these changes necessarily gave Terrace an advantage over its competitors. Many primarily removed disadvantages under which it had previously operated. Facilities and athletic development, in particular, offered an opportunity to catch up with schools that had already invested heavily in those areas, with the potential to move further ahead if the new resources were used effectively.
Nor was Terrace operating in a static competition. Other GPS schools were also recruiting, investing and developing their programs. Since match results measure relative performance, it is possible for a school to improve aspects of its own rugby capability without increasing its winning percentage if its competitors improve as much or more. The historical record here cannot isolate every change in relative underlying capability.
But that does not make it unreasonable to expect better competitive results. The benchmark is not whether Terrace became the most successful or best-resourced school in the competition. It is whether Terrace became more competitively successful than it had been before the intervention, particularly after so many of its previous disadvantages had been removed.
In my assessment, an appropriately designed and integrated program, with coaching personnel and practices capable of exploiting the new opportunities, could reasonably have achieved around one additional win per eight-match season relative to Terrace's historical performance. That would be a meaningful and recurring gain without requiring premiership domination. Indeed, recruitment alone seemed capable of making that scale of difference if the new players were used effectively. This is an assessment of reasonable potential, not a formal target retrospectively attributed to the College or a claim that any one intervention guaranteed a result.
The distinction matters because, as Part Three shows, the results did not come close to that modest illustration. The changes were significant enough that improvements of this order might reasonably have been expected from some of them individually, yet their combined introduction produced no sustained competitive gain.
The essential distinction is between improving individual components and improving the performance of the system in which they operate. Terrace had materially changed what was possible. Whether its coaching and organisational arrangements could turn those opportunities into better results is a separate question.
Part Three turns to the competitive results and representative-player development.
Part Three: Results
More than a decade has passed since Terrace embarked on its unprecedented improvement program. That provides a significant period over which to examine its outcomes and assess whether the changes delivered the improvements they were intended to achieve.
This section examines the program's performance against historical benchmarks, comparable schools and broader measures of success. The immediate objective is to establish what happened, rather than why it happened. Questions of causation, organisational decisions and the underlying assumptions of the improvement strategy will be examined separately.
Headline Outcomes
Two measures provide a useful starting point for evaluating the program: on-field performance and the production of representative players.
While neither captures the full complexity of a school rugby program, they represent distinct and measurable outcomes. One reflects the competitive performance of teams, while the other provides an indication of the program's ability to produce players capable of progressing to the highest levels.
Considered together, they provide a broader picture of the program's performance than wins and losses alone.
On-Field Results
The most obvious place to begin assessing the improvement program is with the results themselves. After all, improving the school's competitive performance was a central objective of the changes.
By almost any measure, the results have been disappointing.
Across the 13 full seasons since the improvement program began, Terrace has recorded nine losing seasons, winning fewer than half its matches. Only four seasons avoided that outcome, and two of those finished with exactly a 50% winning percentage.
But perhaps the most revealing comparison is with the 2010 First XV.
That team entered the season as the preseason favourite for the wooden spoon. It was widely expected to finish last without winning a match. Instead, it finished fourth.
In the 13 full seasons since the improvement program began, Terrace has surpassed the results of that 2010 side just once.
After more than a decade of deliberate improvement, increased resources and significant changes to how the rugby program operates, only one season has bettered the performance of a team that was not expected to win a match.
And 2010 was hardly some extraordinary golden era of team success. It immediately preceded the worst First XV season in Terrace's recorded history to that point.
There is an even more revealing test. Remove the undefeated 2004 premiership season from the pre-improvement record entirely, but retain the post-improvement program's exceptional 2016 season. This is an intentionally generous comparison for the new program: the old program loses its strongest season, while the new program keeps its own.
Even under those conditions, the post-improvement program averaged just 0.06 additional win-equivalents per season over the adjusted historical baseline. Project that difference across its 13 completed seasons and the advantage is less than one win-equivalent in total: approximately 0.79. In a program transformed on so many fronts, even the equivalent of one extra victory in a decade would have been a modest gain. The most favourable adjusted comparison does not reach that scale.
Consider that finding against the scale of the changes described in Part Two: recruitment, substantial financial investment, specialist programs, facilities, and better conditions for preparation. Even after removing the previous program's undefeated premiership from the comparison, the observed advantage across thirteen seasons amounts to less than one additional win-equivalent over the adjusted historical rate.
This is an intentionally asymmetric benchmark, not proof of what any particular investment caused. The detailed calculation, alongside the more balanced test that removes the strongest season from both periods, appears later in Part Three.
The broader finding is the absence of meaningful competitive improvement. Across the complete post-intervention dataset, Terrace's winning percentage was lower than before the changes, and in most seasons it won fewer matches than it lost. The exact size of the difference varies with the assessment window and treatment of exceptional seasons, but a substantial improvement does not emerge.
Match results are only one way to evaluate a rugby program. The production of representative players offers a second measure, and the picture is arguably even more concerning.
Player Development
Perhaps the most surprising finding is not the decline in Terrace's on-field results, but what happened to its production of representative players.
Player development is notoriously difficult to measure. We cannot easily quantify how much a school improves an individual player. But we can measure how many players emerge from its program to represent Australia.
And historically, Terrace was exceptionally successful at producing them. Even during periods of poor team performance, the school continued to produce players who progressed to represent Australia. Representative production was not dependent on winning rugby matches.
This makes the subsequent decline particularly interesting.
The improvement program introduced substantial additional resources, specialist athletic development and the recruitment of talented players from further along the development pathway. Even if these changes failed to improve the First XV's results, we might reasonably have expected them to improve the school's production of representative players.
Instead, the opposite occurred.
Across comparable nine-year selection windows, Terrace's annual production of Australian Schoolboys representatives declined by 45.45% after the improvement program began. The scale and direction of that change are especially striking because Terrace was now actively recruiting talented players, including students who had already been identified in representative development pathways. The underlying counts were 11 selections in 2003–2011 and six in 2012–2020. These are representative-selection windows, not the season exclusions used for First XV match results.
The finding becomes more pronounced over a longer assessment window. Because representative pathways changed, this extension incorporates Australian Under-18 selections as well as Australian Schoolboys selections. On an annual basis, the decline reaches 53.25% when the analysis is extended to 2026. The calculation compares 11 selections over the nine pre-improvement years (2003–2011) with eight over fourteen assessed post-improvement years (2012–2019 and 2021–2026, excluding 2020). This representative-selection window differs from the 13-season on-field dataset and should not be conflated with it.
There is also a separate national comparison based specifically on Australian Schoolboys production. In the pre-improvement period Terrace ranked 10th among Australian schools for representatives produced. In the corresponding post-improvement period it ranked 17th. These are period-specific production rankings, not its cumulative all-time standing, and they should not be confused with the extended Schoolboys/Under-18 count.
How does that compare with what might reasonably have been expected?
Simply applying Terrace's pre-improvement rate of Australian Schoolboys production over a 13-season span would yield approximately 16 representatives, without assuming any improvement at all. That is a historical-rate projection, not a claim that individual players were destined for selection; the formal percentage comparisons above use their own stated selection windows.
And Terrace's earlier success came before the scholarships, recruitment strategies and specialist development resources introduced as part of the improvement program.
This decline is not restricted to school-age representation. Terrace is historically the second most successful producer of Australian Under-20 representatives, yet its production at that level has also fallen by 37.5% since the changes began.
None of this, by itself, establishes why representative production declined. But it establishes another important outcome against which the improvement program must be assessed.
Terrace did not merely fail to improve its production of representative players. It produced substantially fewer of them, despite investing in the very resources that might reasonably have been expected to increase their number.
And that raises a much more interesting question about what the program was actually improving.
Data Selection and Treatment
Comparing the performance of a rugby program across different periods presents several methodological challenges. Not all seasons were played under comparable conditions, and changes to competitions, representative pathways and external circumstances can complicate direct comparisons.
To account for these differences, some seasons have been excluded from particular analyses, while others have been retained despite unusual circumstances. These decisions are intended to ensure that comparisons are as consistent and meaningful as possible, without selectively removing results simply because they are favourable or unfavourable to the program.
The following sections explain the treatment of specific seasons, the rationale behind those decisions and how they affect the interpretation of the results presented in this study.
Handling COVID-19
Perhaps the most obvious challenge in comparing results across the assessment period is the disruption caused by the COVID-19 pandemic.
Both the 2020 and 2021 seasons were significantly affected, although in different ways.
In 2020, the competition experienced substantial interruptions, with matches cancelled and an irregular playing schedule. The disruption was sufficiently significant that no premiership was awarded.
The 2021 season presents a different problem. Although all eight scheduled matches were played, the first three were classified as trial matches and did not count towards the premiership. Terrace lost all three, leaving just five matches that contributed to the official competition results.
These two seasons were structurally non-comparable with the other years examined in this study. The competition conditions, scheduling and treatment of results differed fundamentally from those of a conventional GPS season. Consequently, including them would introduce inconsistencies into comparisons intended to assess performance across broadly equivalent competitive conditions.
Rather than attempt to reconstruct or adjust the results to account for these differences, both seasons have been excluded from the comparative analysis of on-field performance.
This approach is applied consistently, irrespective of whether the excluded results were favourable or unfavourable to Terrace. The objective is not to improve the appearance of either assessment period, but to preserve the integrity of comparisons between seasons conducted under equivalent competitive structures.
Handling 2011
The 2011 season presents another significant challenge to the comparability of historical results.
That year, Terrace recorded the worst First XV season in its history to that point. While several factors contributed to this outcome, the extraordinary circumstances created by the Brisbane floods placed the school at a substantial competitive disadvantage.
The floods effectively deprived Terrace of its home-ground advantage for the entire season. Every match was played away from home, while every training session had to be conducted at an alternative, unfamiliar venue.
These circumstances were not shared by any other school in the competition. Terrace was therefore operating under conditions fundamentally different from those experienced by its opponents and from those of any other season in the assessment period.
This distinction is important. The poor results of 2011 cannot necessarily be attributed entirely to the floods, and other aspects of the program's performance remain open to scrutiny. Nevertheless, even without those additional factors, Terrace faced significant disadvantages arising from circumstances beyond its control.
The issue, therefore, is not simply that 2011 was an exceptionally poor season. The competition proceeded largely as normal, but Terrace experienced a uniquely severe structural disadvantage that no other school faced. This makes its results fundamentally different from those of other seasons in the assessment period.
For this reason, 2011 has been excluded from the primary comparative datasets for on-field performance.
The Australian Schoolboys comparison, however, retains 2011. Representative selection is an individual outcome, not a direct function of a team's win-loss record, and the floods alone do not justify removing an entire year of schoolboy selection data. Terrace produced no Australian Schoolboys representatives in 2011. Retaining that year therefore produces the more conservative estimate of the later decline: 45.45% over the two nine-year periods, rather than approximately 51.52% if the zero-selection year were removed from the pre-improvement period. The shorter-window representative outlier test is explained separately below.
To maintain transparency in the on-field analysis, the study also identifies where including the 2011 match results materially changes the findings. This allows readers to assess the effect of the exclusion rather than relying solely on the adjusted figures.
Periods of Analysis
Having established which seasons are excluded from the on-field performance analysis, we can divide the remaining data into four distinct periods for comparison.
The first, Post 1 (2012–2019), covers the initial eight seasons following the introduction of the improvement program in 2012, ending immediately before the disruptions caused by COVID-19.
The comparison period is Pre (2003–2010), the eight seasons immediately preceding the disrupted 2011 season. Matching the length of Pre and Post 1 allows a direct comparison before and after the intervention.
The third period, Post 2 (2022–2026), covers the five completed seasons following the COVID-19 disruptions. This provides a more recent assessment of the program's performance and allows us to examine whether outcomes have changed over time.
Finally, Post All (2012–2026) combines both post-intervention periods into a single dataset spanning 13 completed seasons. The disrupted 2020 and 2021 seasons remain excluded.
| Period | Years Included | Seasons |
|---|---|---|
| Pre | 2003–2010 | 8 |
| Post 1 | 2012–2019 | 8 |
| Post 2 | 2022–2026 | 5 |
| Post All | 2012–2019 and 2022–2026 | 13 |
Together, these four periods allow us to compare the program's performance before and after the intervention, distinguish between its earlier and more recent outcomes, and assess its overall performance across the full post-intervention period.
The representative-selection analyses may use different historical windows where necessary to accommodate changes in national selection pathways. Those comparisons are identified separately rather than assumed to follow the on-field performance periods exactly.
On-Field Performance Analysis
With those comparison periods established, we can examine competitive performance in greater detail.
This analysis considers winning percentage, competition ranking, and mean and median finishing position. These measures draw on the same underlying match results and should not be mistaken for independent statistical confirmations. They nevertheless describe different dimensions of the record: winning percentage captures match outcomes, ranking and finishing position show Terrace's standing relative to its competitors, and the mean and median help distinguish average performance from the influence of exceptional seasons.
Of particular interest is not simply whether performance improved following the intervention, but how it changed over time. By distinguishing between the initial eight-year period and the subsequent five seasons, we can assess both the overall outcomes of the improvement program and the trajectory of its performance as the intervention matured.
The central comparison is between Terrace before and after the intervention. The changes in resources, recruitment, staffing and institutional commitment created an expectation of meaningful improvement, not necessarily premiership success, but performance better than the school had previously achieved.
Winning Percentage
The most straightforward measure of competitive performance is winning percentage, which provides a consistent basis for comparing results across the four assessment periods.
During the pre-improvement period (2003–2010), Terrace recorded a winning percentage of 42.9%. This declined to 37.5% during Post 1 (2012–2019), representing a relative decrease of 12.6%.
Rather than recovering as the improvement program matured, performance deteriorated further. During Post 2 (2022–2026), the winning percentage fell to just 32.5%, a 24.2% decline relative to the pre-improvement period.
Across the entire 13-season post-improvement period, Terrace recorded 37 victories from 104 matches, an overall winning percentage of approximately 35.6%. The pre-improvement rate of 42.9% was equivalent to roughly 3.43 wins per eight-match season; the post-improvement rate was approximately 2.85. In ordinary sporting terms, Terrace averaged about 0.58 fewer win-equivalents per season, not more. The difference is 7.3 percentage points, or a relative decline of approximately 17.0%. This describes the historical results, not proof that the intervention caused them.
| Period | Winning Percentage | Relative Change from Pre |
|---|---|---|
| Pre | 42.9% | Baseline |
| Post 1 | 37.5% | −12.6% |
| Post 2 | 32.5% | −24.2% |
| Post All | ≈35.6% | ≈−17.0% |
| Pre (including 2011) | 38.2% | Alternative baseline |
Importantly, this finding is not dependent on the exclusion of the exceptionally poor 2011 season. Including 2011 in the pre-improvement dataset reduces the historical winning percentage to 38.2%, which remains higher than the overall post-improvement result.
Two findings are particularly significant. First, the improvement program has produced a lower winning percentage than the preceding program, even when the latter's worst season is included. Second, performance has deteriorated further over time, with the most recent five-season period recording substantially worse results than the initial eight seasons following the intervention.
The decline is therefore not simply an unfavourable comparison between the periods before and after the intervention. It also reflects a worsening competitive record as the improvement program progressed.
Winning Percentage Relative to the Competition
While winning percentage provides a useful measure of performance over time, ranking those percentages against the other schools in the competition provides additional context.
During the pre-improvement period, Terrace ranked fifth in the competition for overall winning percentage. This fell to sixth during Post 1, although that position somewhat disguises how close the school came to finishing considerably lower.
In Post 1, Terrace was separated from seventh and eighth place by just 1.5 percentage points. A single additional loss would have been sufficient to push the school down to eighth.
The results for Post 2 reveal a much more substantial deterioration.
Between 2022 and 2026, Terrace ranked eighth in the competition for winning percentage. More significantly, the school finished 14 percentage points behind seventh place. This was not a marginal difference in ranking, but a substantial gap between Terrace and the next highest-performing school.
The picture across the complete post-improvement period is similarly concerning. Terrace ranked eighth overall, with a winning percentage seven percentage points below seventh place.
These comparisons reveal more than a simple decline in winning percentage. Terrace has also deteriorated relative to its competitors, moving from fifth in the pre-improvement period to eighth across the complete post-improvement period.
Perhaps most importantly, the gap between Terrace and the schools immediately above it has become substantial. What was a relatively competitive sixth-place position during Post 1 has become a distant eighth place during Post 2.
The decline in competitive performance is therefore evident not only in Terrace's own historical results, but also in its position relative to the other schools against which it competes.
Mean Finishing Position
Another measure of competitive performance is mean finishing position, calculated by averaging each school's final ladder position across the relevant assessment period.
With nine schools competing in the GPS competition, finishing positions range from first to ninth, with a lower mean indicating stronger performance.
During the pre-improvement period, Terrace recorded a mean finishing position of 5.125, ranking fifth among the nine schools.
This deteriorated during Post 1, with the mean finishing position falling to 6.0. More significantly, this placed Terrace eighth in the competition when schools were ranked by their mean finishing positions over the same period.
The decline continued during Post 2, with Terrace recording a mean finishing position of 6.2, again ranking eighth among the nine schools.
Across the complete 13-season post-improvement period, Terrace's mean finishing position was approximately 6.08, also ranking eighth overall.
The results reveal a consistent deterioration in competitive performance. Before the intervention, Terrace's average finishing position placed it fifth among its competitors. Following the intervention, the school fell to eighth, a position it has maintained across both post-improvement periods.
Although the difference in mean finishing position between Post 1 and Post 2 is relatively small, the broader comparison is significant. Terrace has moved from the middle of the competition to near the bottom when measured by average ladder position.
This reinforces the findings from the winning percentage analysis. The deterioration in results is evident not only in the proportion of matches won, but also in Terrace's ability to finish ahead of its competitors over the course of a season.
Median Finishing Position
While mean finishing position provides an overall measure of competitive performance, the median offers a complementary perspective. By identifying the middle finishing position across the assessment period, it is less influenced by unusually successful or unsuccessful individual seasons.
During the pre-improvement period, Terrace recorded a median finishing position of fifth, which also ranked fifth among the nine competing schools.
In Post 1, the median deteriorated to seventh, placing Terrace eighth in the competition when ranked against the median finishing positions of the other schools.
During Post 2, Terrace recorded a slightly better median finishing position of sixth, ranking equal sixth among its competitors. However, this apparent improvement in relative ranking requires some context.
With only five seasons in Post 2, the smaller sample makes the median particularly sensitive to individual season results. Furthermore, three schools recorded identical median finishing positions of sixth, effectively occupying positions six through eight in the comparative rankings. Only one school recorded a worse median than Terrace.
Across the complete 13-season post-improvement period, Terrace recorded a median finishing position of seventh, ranking eighth among the nine schools.
This provides another perspective on the same underlying match record. Terrace's typical finishing position deteriorated from fifth before the intervention to seventh across the complete post-improvement period, while its comparative ranking fell from fifth to eighth.
Although Post 2 recorded a slightly better median than Post 1, its mean finishing position was worse. Given the smaller sample and the clustering of schools with identical medians, little significance should be attached to this difference.
The more substantial finding is the consistency of the long-term deterioration. Both mean and median finishing positions indicate that Terrace has moved from the middle of the competition to near the bottom since the introduction of the improvement program.
The Grammar Problem
Any assessment of Terrace's post-improvement performance must consider the extraordinary circumstances surrounding Brisbane Grammar School.
In 2009, Grammar experienced the worst First XV season in its history. Like Terrace would do following its disastrous 2011 season, Grammar responded by embarking on a significant improvement program.
The turnaround was remarkable. Just three years later, in 2012, Grammar won the GPS premiership.
However, what happened next makes Grammar particularly relevant to this analysis.
Following its premiership success, Grammar experienced an extraordinary and sustained decline in competitive performance.
Across the subsequent 11 seasons, 2013–2019 and 2022–2025, Grammar recorded just 10 wins. Four of those seasons ended without a single competition point, with the school failing to register either a win or a draw.
To put that into perspective, Terrace has recorded just one winless and drawless season in its entire history: the disastrous 2011 season that helped precipitate its own improvement program.
Grammar recorded four such seasons in just 11 years.
This creates an important problem when interpreting Terrace's comparative performance.
Throughout the post-improvement analysis, Terrace repeatedly ranks eighth among the nine GPS schools. However, the school occupying ninth place is not simply another struggling rugby program. It is a school that experienced one of the most prolonged periods of competitive failure in its history.
The distinction is significant.
Finishing eighth in a nine-team competition is one thing. Finishing eighth when the school immediately below you has won just 10 matches in 11 seasons, including four entirely winless and drawless seasons, is something quite different.
This does not invalidate the rankings. However, it does change how they should be understood.
Grammar's difficulties also affected the composition of Terrace's own winning record. Of the 37 victories Terrace recorded over the thirteen assessed post-improvement seasons, eight came against Brisbane Grammar, representing 21.6% of all its wins. Grammar's weakness is therefore relevant not only to Terrace's eighth-place relative standing, but also to interpreting the winning percentage on which much of the analysis rests. That does not tell us which results would have changed against a stronger Grammar side; it does show that more than one-fifth of Terrace's victories came against a particularly unsuccessful opponent.
Terrace's position near the bottom of the competition is concerning in its own right. The extraordinary weakness of the school immediately below it makes that position considerably more revealing.
Who Is Terrace Actually Beating?
Winning percentage tells us how often a team wins, but it tells us very little about the quality of the opposition it defeats. To understand Terrace's competitive performance more fully, we need to examine not simply how many matches it has won, but who those victories have come against.
Across the 13 post-improvement seasons, Terrace recorded 37 victories. The average winning percentage of the opponents it defeated was just 28.4%.
However, this measure contains a small mathematical distortion. Every team Terrace defeats necessarily has a loss against Terrace included in its season record, reducing its overall winning percentage.
To account for this, an Adjusted Strength of Victory measure was calculated by removing each defeated opponent's match against Terrace and recalculating its winning percentage against its remaining seven opponents. Draws were counted as half a win.
For example, a team that finished with six wins and two losses, including a defeat to Terrace, would ordinarily have a winning percentage of 75%. Removing the Terrace result produces an adjusted winning percentage of 85.7%, reflecting its performance against the rest of the competition.
Using this adjustment, the average strength of Terrace's defeated opponents increases from 28.4% to 32.4%.
Even after accounting for this distortion, the results suggest that Terrace's victories have been disproportionately concentrated against weaker opposition.
The Strength of Defeated Opposition
The distribution of Terrace's victories makes this particularly apparent.
Of the 37 post-improvement victories, six came against teams that failed to win a single match throughout their season. That represents 16.2% of all Terrace victories.
Extending this analysis, 13 victories (35.1%) came against teams that won no more than one match during their season, while 20 victories (54.1%) came against teams that won no more than two.
Most significantly, 28 of Terrace's 37 victories, or 75.7%, came against opponents that finished their seasons with a winning percentage of 37.5% or lower.
In other words, approximately three-quarters of Terrace's victories came against teams that won no more than three of their eight competition matches.
At the other end of the distribution, Terrace recorded just nine victories against teams that finished with a winning percentage of 50% or greater. Only five of those victories came against teams that finished with five or more wins.
Perhaps the most striking finding concerns the competition's stronger teams. Across the post-improvement period, Terrace played 34 matches against opponents that finished their seasons with six or seven victories. Terrace lost all 34.
This reveals a significant limitation in the program's competitive performance. Terrace has not simply struggled to win consistently. Its victories have been overwhelmingly concentrated against the weaker teams in the competition, while it has been unable to defeat opponents producing consistently strong results.
Victories by Opponent
Examining Terrace's victories by individual school provides further context.
| Opponent | Victories | Share of Total |
|---|---|---|
| BGS | 8 | 21.6% |
| IGS | 7 | 18.9% |
| BSHS | 6 | 16.2% |
| ACGS | 5 | 13.5% |
| BBC | 4 | 10.8% |
| TGS | 4 | 10.8% |
| TSS | 2 | 5.4% |
| NC | 1 | 2.7% |
| Total | 37 | 100% |
Brisbane Grammar accounts for the largest share of Terrace's victories, with eight wins representing 21.6% of the total. As discussed above, that concentration is particularly relevant given Grammar's extraordinary competitive decline.
Together, Brisbane Grammar, Ipswich Grammar and Brisbane State High account for 21 of Terrace's 37 victories, or 56.8% of all wins.
Although the identity of an opponent does not necessarily establish its strength in any particular season, the distribution demonstrates that Terrace's victories have been far from evenly spread across the competition.
Comparing Post 1 and Post 2
The distinction between Post 1 and Post 2 becomes particularly revealing when the strength of defeated opposition is considered.
As previously established, Terrace's winning percentage declined from 37.5% in Post 1 to 32.5% in Post 2. However, the deterioration extends beyond the frequency of victories.
The opponents Terrace managed to defeat also became weaker.
| Measure | Post 1 | Post 2 |
|---|---|---|
| Seasons | 8 | 5 |
| Victories | 24 | 13 |
| Winning percentage | 37.5% | 32.5% |
| Average defeated-opponent win % | 30.7% | 24.0% |
| Adjusted strength of victory | 35.1% | 27.5% |
| Wins against teams with winning records | 4 | 1 |
| Wins against winless teams | 2 | 4 |
During Post 1, the average winning percentage of defeated opponents was 30.7%, increasing to 35.1% under the adjusted measure.
In Post 2, those figures fell to just 24.0% and 27.5%, respectively.
The decline in adjusted strength of victory is particularly relevant because it is not merely a reflection of Terrace's own results. By excluding the head-to-head match, the measure examines how successful Terrace's defeated opponents were against the rest of the competition.
The composition of those victories is equally revealing.
Across the eight seasons of Post 1, Terrace recorded four victories against teams with winning records and two against teams that failed to win a match.
During the five seasons of Post 2, Terrace managed just one victory against a team with a winning record, while four of its 13 victories came against teams that finished their seasons winless.
Consequently, victories against winless teams accounted for 30.8% of all Terrace wins in Post 2, compared with just 8.3% in Post 1.
This distinction reveals a deterioration that winning percentage alone cannot adequately capture.
Not only has Terrace become less successful at winning matches, but the teams it has managed to defeat have, on average, become less competitive themselves.
The program's declining winning percentage therefore understates an additional dimension of its competitive regression. Fewer victories are being achieved, and an increasing proportion of those victories are coming against the competition's weakest opponents.
Taken together, these findings reinforce the broader pattern emerging from the analysis. Terrace's performance has deteriorated since the introduction of the improvement program, and the more recent period provides little evidence of a competitive recovery. Instead, both the frequency and the quality of its victories have declined.
Who Is Terrace Losing To?
Examining the strength of defeated opponents provides one perspective on a program's competitive performance. However, the nature of its defeats can be equally revealing.
In Terrace's case, the post-improvement period has been characterised by two particularly concerning patterns: an inability to defeat stronger opponents and a series of losses against teams whose broader results suggest Terrace should have been competitive, if not favoured to win.
Losing to Stronger Opposition
As established in the previous section, Terrace has struggled to defeat the competition's strongest teams. That record has become particularly pronounced in recent seasons.
Since 2017, Terrace has lost 31 of its 32 matches against teams that finished their seasons with more than four victories, a defeat rate of 96.9%.
In other words, across this period, Terrace has managed just one victory against opponents that finished with winning records.
While losses against stronger opponents are not inherently unexpected, the near-total absence of victories across such a large number of matches suggests that Terrace has struggled to compete successfully against the upper half of the competition.
Unexpected Losses
Perhaps more revealing is what has happened against weaker opposition.
On four occasions during the post-improvement period, Terrace has lost to a team whose only victory of the entire season came against Terrace. In each case, the opponent finished last in the competition.
Prior to the intervention, such a result had occurred just once.
These defeats are particularly significant because they represent more than simply losing to a weaker opponent. In each instance, Terrace was the only team that its opponent managed to defeat throughout the entire season.
Although unexpected results are an unavoidable feature of competitive sport, their recurrence provides another measure through which the program's performance can be assessed.
Losses Despite Superior Results Against Common Opponents
Another way to examine unexpected defeats is to compare how Terrace and its opponents performed against the same seven other schools.
This approach removes the head-to-head result and considers only their performances against common opposition. It therefore provides a measure of relative competitive performance independent of the match being assessed.
For example, if Terrace defeated five of its seven common opponents while another school defeated only two, Terrace would hold a three-win advantage. A subsequent loss to that school would represent a defeat despite substantially superior results against the rest of the competition.
The comparison between the pre- and post-improvement periods reveals several differences.
| Measure | Pre | Post All |
|---|---|---|
| Losses despite outperforming opponents against common opposition | 7 | 10 |
| Losses despite an advantage of 2+ wins | 2 | 5 |
| Losses despite an advantage of 3+ wins | 0 | 2 |
| Largest common-opponent win advantage in a defeat | 2 | 6 |
During the pre-improvement period, Terrace recorded seven losses against teams it had outperformed against common opponents. Only two of these occurred when Terrace held an advantage of at least two victories.
Importantly, Terrace never lost a match when it had outperformed its opponent by three or more wins against their seven common opponents.
Since the intervention, that has happened twice.
The most extreme example is particularly striking. During the pre-improvement period, Terrace's largest common-opponent advantage in a defeat was two wins. In the post-improvement period, Terrace has lost a match despite recording six more victories than its opponent against their common opposition.
That represents a substantial disparity in demonstrated performance across the remainder of the competition.
The pre- and post-improvement periods contain different numbers of seasons, so the absolute frequency of these defeats should not be compared without accounting for the different sample sizes. Nevertheless, the emergence of losses involving substantially larger performance advantages is notable.
The distinction is not simply that Terrace has experienced more unexpected defeats in absolute terms, but that some of its post-improvement defeats have occurred despite far greater evidence of competitive superiority against common opposition.
A Broader Pattern of Competitive Performance
Considered alongside the earlier analysis of Terrace's victories, these results reveal two different dimensions of competitive difficulty.
Against stronger opponents, Terrace has struggled to achieve victories, particularly in the more recent seasons.
Against weaker opponents, Terrace has recorded several unusually poor defeats, including four losses to teams that failed to beat anyone else.
Neither finding, considered independently, provides a complete account of the program's performance. Strong teams occasionally lose to weak opponents, and struggling teams can experience prolonged periods without defeating successful opposition.
However, the combination is more revealing.
Terrace has struggled to defeat teams performing well across the competition, while also suffering defeats against opponents whose results were substantially worse than its own.
This reinforces the broader findings of the results analysis. The concern is not merely that Terrace has won fewer matches since the introduction of the improvement program. The distribution of its victories and defeats also raises questions about the program's ability to produce consistently competitive performances against opponents of differing strengths.
The Most Recent Four Seasons
Perhaps the most concerning feature of the results is that the deterioration has continued into the most recent seasons.
Between 2023 and 2026, Terrace recorded just 10 victories from 32 matches, representing a winning percentage of 31.25%. This is lower than the 32.5% recorded across Post 2 and substantially below the pre-improvement benchmark of 42.9%.
However, as the earlier analysis of defeated opponents demonstrated, winning percentage alone does not tell the complete story.
Of those 10 victories, four came against teams that failed to win a single match throughout their respective seasons.
In other words, 40% of Terrace's victories over the past four seasons have come against teams that were unable to defeat anyone else in the competition.
This leaves just six victories across four seasons against opponents that managed to win at least one match themselves.
These figures are particularly concerning when considered alongside the broader trajectory of the improvement program. The initial eight post-intervention seasons produced worse results than the preceding program. The subsequent five-season period deteriorated further, and the most recent four seasons have produced an even lower winning percentage, with an unusually large proportion of victories coming against winless opposition.
Although these overlapping assessment periods do not establish that performance has deteriorated in every successive season, they provide little evidence of a sustained competitive recovery.
More than a decade after the improvement program began, the most recent results remain among its least successful, both in the frequency of victories and the strength of the opposition being defeated.
Accounting for Outlier Seasons
Before drawing conclusions from the results, it is worth considering the influence of two exceptional seasons that may distort comparisons between the pre- and post-improvement periods.
In 2004, Terrace won an undefeated GPS premiership. Twelve years later, in 2016, the school recorded six victories and finished second in the competition.
Both seasons stand substantially above the surrounding results in their respective assessment periods. Consequently, it is reasonable to question whether these exceptional performances disproportionately influence the overall findings.
Neither exceptional season should be allowed to determine the assessment of a program spanning many years. We can test the sensitivity of the results in two ways: first by making an intentionally favourable adjustment for the new program, then by applying the same exclusion to both periods.
A Deliberately Favourable Comparison
One useful test is to make an intentionally asymmetric adjustment. What would the results look like if we removed the most successful season of the old program, 2004, but retained the strongest season of the new program, 2016?
The undefeated premiership disappears from the baseline, while the 2016 second-place finish remains in the later record. This is not the preferred historical comparison and does not replace the full dataset. It asks how far the evidence must be tilted in the new program's favour before the result becomes even marginally positive.
| Measure | Pre, excluding 2004 | Post All, retaining 2016 |
|---|---|---|
| Seasons | 7 | 13 |
| Winning percentage | 34.82% | 35.58% |
| Win-equivalents per season | ≈2.79 | ≈2.85 |
| Post-improvement advantage per season | ≈0.06 | |
| Equivalent advantage over 13 seasons | ≈0.79 win-equivalents |
Draws count as half a win. The adjusted pre-improvement period contains approximately 19.5 win-equivalents over seven seasons, compared with 37 wins over thirteen seasons after the intervention. The resulting annual-rate difference is approximately 0.0604 win-equivalents, or approximately 0.79 projected across thirteen seasons. Displayed percentages and averages are rounded.
The 0.79 figure is a comparison of season-average rates projected over thirteen seasons, not a literal count of additional matches won or an estimate of the causal effect of recruitment or spending. But it illustrates the central problem: even a comparison deliberately weighted in favour of the post-improvement program yields the equivalent of less than one extra win across the entire assessment period.
A fairer outlier sensitivity test applies the same treatment to both periods, removing 2004 from the pre-improvement record and 2016 from the post-improvement record.
On-Field Performance Without the Outliers
Removing the exceptional seasons reduces the winning percentages of both periods, but the underlying pattern remains unchanged.
With 2004 excluded, Terrace's pre-improvement winning percentage falls to 34.82%, ranking sixth among the nine GPS schools.
Removing 2016 from the complete post-improvement period produces a winning percentage of 32.29%, ranking eighth.
This represents a relative decline of 7.3% in winning percentage, even after the two exceptional seasons have been removed.
The same pattern is evident in the measures of finishing position.
| Measure | Pre (excluding 2004) | Post All (excluding 2016) |
|---|---|---|
| Winning percentage | 34.82% | 32.29% |
| Winning percentage rank | 6th | 8th |
| Mean finishing position | 5.7 | 6.4 |
| Mean finishing position rank | 6th | 8th |
| Median finishing position | 5th | 7th |
| Median finishing position rank | 5th | 8th |
Across all three measures, Terrace's post-improvement performance remains inferior to its pre-improvement performance.
Importantly, the comparative rankings also remain worse. Even without its undefeated premiership season, the earlier program ranks ahead of the post-improvement program in winning percentage, mean finishing position and median finishing position.
The results therefore cannot be explained simply by the influence of Terrace's exceptional 2004 premiership.
The Impact on Representative Production
The exceptional season for team results is not necessarily the exceptional season for representative production. 2004 was Terrace's standout pre-improvement year on the field, with an undefeated premiership. For Australian Schoolboys production, however, the relevant pre-improvement outlier was 2010. That First XV began as the preseason favourite for the wooden spoon, with no victories expected, but ultimately finished fourth and produced three Australian Schoolboys representatives. The contrasting results make clear why competitive performance and representative production need to be examined separately.
The original nine-year Australian Schoolboys comparison covers 2003–2011 and 2012–2020. Terrace produced 11 representatives in the earlier period and six in the later one, a 45.45% decline in annual production. The exceptional representative-producing cohorts were 2010 and 2016, with three selections in each year. Removing those two cohorts leaves eight representatives in eight pre-improvement years and three representatives in eight post-improvement years.
| Australian Schoolboys measure | Pre-improvement | Post-improvement |
|---|---|---|
| Original representatives (nine years) | 11 | 6 |
| Exceptional cohort excluded | 2010 | 2016 |
| Representatives removed | 3 | 3 |
| Remaining representatives | 8 | 3 |
| Remaining seasons | 8 | 8 |
| Representatives per season | 1.000 | 0.375 |
On that like-for-like, cohort-specific basis, average annual Australian Schoolboys production was 62.5% lower in the post-improvement period. This is a sensitivity test, not a replacement for the primary 45.45% comparison or a reason to discount either exceptional cohort's legitimate achievement.
Australian Under-20 production remains relevant to the broader development record, as discussed earlier. It is not included in this outlier adjustment: selection can occur after a player leaves school, and the intervening development pathways, timing and eligibility introduce complications that are not resolved by removing the same two school cohorts.
Interpreting the Outliers
Removing exceptional seasons does not necessarily produce a more accurate history. The undefeated 2004 premiership, the second-place finish in 2016 and the representative cohorts of 2010 and 2016 are all legitimate achievements. They belong in the record.
The sensitivity tests answer different questions. For on-field results, removing the strongest competitive season from each period leaves the post-improvement program with a lower winning percentage and worse comparative finishing measures. For Australian Schoolboys production, removing the exceptional representative cohort from each period leaves a 62.5% decline in annual production. Neither test should be confused with the full historical comparison, and neither identifies the cause of the outcomes.
What remains is the central finding: the substantial changes surrounding Terrace's improvement program did not translate into meaningful improvement in its observed competitive results or its rate of Australian Schoolboys production.
Conclusion: The Difference Between Change and Improvement
The central finding of this case study is not that Terrace failed to become the strongest rugby school in Queensland. It is that a program deliberately redesigned to improve competitive performance, and afforded substantially greater opportunities to do so, did not deliver meaningful improvement against its own historical results. In the primary comparison, the winning percentage actually declined.
The changes were real. Recruitment became possible where it had previously been rejected. Athletic development, facilities and institutional support were expanded. A longer preseason and a more favourable age profile reduced disadvantages that had once been central to the explanation for Terrace's difficulties. Not all of these developments were the College's doing, and none guaranteed success. But together they created a compelling reason to expect measurable improvement against Terrace's own historical record. The ambition did not have to be spectacular: even one additional win each season would have been meaningful.
That expectation was not met. Across the comparable seasons, Terrace's winning percentage declined from 42.9% before the intervention to approximately 35.6% afterwards, a difference of about 0.58 fewer win-equivalents per eight-match season. Its relative position in the competition deteriorated, and representative-player production declined sharply against the relevant historical benchmarks. The more recent results provide little evidence of sustained recovery. The absence of meaningful improvement persists when 2011 is restored to the earlier on-field baseline and when exceptional cohorts or seasons are excluded separately for the outcomes being examined.
Even the comparison most favourable to the intervention produces almost nothing. Remove Terrace's undefeated 2004 premiership from the earlier baseline but retain its successful 2016 season in the later record. The post-improvement advantage is still just 0.06 win-equivalents per season: less than one projected additional win-equivalent over thirteen seasons. Even one extra win in a decade would represent a more visible return than this deliberately favourable comparison produces. This is not the primary comparison. It is a deliberately generous test, and a revealing one given the millions of dollars invested in facilities and the substantial ongoing commitment to staffing, preparation and recruitment.
Those findings establish an outcome. They do not, by themselves, establish its cause. They cannot tell us which interventions were ineffective, which may have been implemented poorly, which earlier capabilities were lost, or what would have happened had Terrace chosen a different path. Nor do they rule out gains in participation, student experience or other educational outcomes that have not been measured here. Those possibilities deserve investigation rather than dismissal.
But neither should they become substitutes for assessing the stated competitive ambition. A program can become better resourced, more elaborate and more professional in appearance without becoming more effective at what it was established to do. Change is not evidence of improvement. Investment is not evidence of improvement. Even the sincerity and scale of the effort are not evidence of improvement. Outcomes have to be examined against the intended purpose.
That brings us to the more demanding question raised at the beginning of this study. What happens when an organisation has made a reasonable decision under uncertainty, but the results increasingly challenge the assumptions behind that decision? Were the disappointing outcomes examined as evidence that the original theory of improvement might have been wrong? Were its underlying assumptions reconsidered, or were changes made largely within the same approach? The historical record assembled here makes those questions worth asking. It does not yet answer them.
The next stage of the Rugbyology case study is not to catalogue mistakes or appoint culprits. It is to examine the reasoning behind the intervention, the mechanisms through which it was expected to work, the capabilities of the previous program and the evidence that might have justified changing direction.
An organisation doesn't have to learn from its own mistakes for others to learn from them.
Terrace is the case study. Improvement is the subject.
No spam, no sharing to third party.
Member discussion