The Case for Doing Less: When More Data Stops Improving Decisions
The default instinct in elite sport is acquisitive. A new sensor, a new platform, a new model, each promising a sharper edge. The logic feels unimpeachable: more data means more insight, and more insight means better decisions. This article questions the middle step. The evidence increasingly suggests that beyond a certain point, additional data stops improving decisions and starts degrading them, by overwhelming the human capacity to interpret it, by diluting attention across metrics that do not change anything, and by substituting collection for judgement. The case for doing less is not a case for ignorance. It is a case for matching data volume to decision capacity.
The constraint is interpretation, not collection
The most consistent finding across the applied monitoring literature is that the binding constraint in elite sport is not how much data a club collects, but whether anyone can convert it into a decision. This is not a marginal observation, it recurs across wearables, GPS, athlete-management systems and injury monitoring.
The pattern is straightforward. Monitoring programs succeed where a specific person’s role is to translate data into coaching, medical and load decisions, and they fail where the technology is treated as self-explanatory. The most accurate, most expensive sensor in the building produces nothing of value if its output lands on a dashboard nobody has the time or remit to read. Adding a second sensor to that situation does not help, but it adds a second unread stream to the first.
This reframes the acquisition question entirely. The relevant variable is not “what can we measure?” but “what can we interpret and act on?” A club with one well-used data stream and a dedicated analyst is in a stronger position than a club with ten streams and no one to read them. The second club has more data and worse decisions.
Why more inputs can mean worse models
The intuition that more data improves prediction also breaks down on closer inspection, and the injury-prediction literature is the clearest example. As established in the systematic review by Bullock and colleagues, the published injury-prediction literature is dominated by models at high or unclear risk of bias, none of which had been externally validated at the time of review (Bullock et al., 2022). Adding more predictors to a model that is already poorly validated does not fix it, it frequently makes it more prone to overfitting, where the model learns the noise in its development data rather than a generalizable signal.
The same caution applies to the individual metrics feeding these systems. The acute:chronic workload ratio, long a centrepiece of load monitoring, was shown by Impellizzeri and colleagues to suffer from mathematical coupling and statistical artefacts that can generate spurious associations, with no evidence supporting its use for injury-reduction recommendations (Impellizzeri et al., 2020).
A department that adds this metric to its dashboard has not added information; it may have added noise dressed as information, which is worse, because it invites action.
The general principle is that data quality and decision relevance matter more than data quantity. A smaller set of well-validated, well-understood metrics that each connect to a specific decision will outperform a sprawling dashboard of weakly-validated numbers that connect to nothing.
The quality framework the field already has
This is not a fringe position. The sports-technology quality framework developed by Robertson and colleagues argues explicitly for a shift away from acquisition and toward evidence, validation and context (Robertson et al., 2023). The framework’s logic is that a technology earns its place not by existing or by being novel, but by demonstrating that it produces decision-relevant, valid information in the context it will actually be used. A tool that cannot meet that bar is not a neutral addition – it is a draw on attention, budget and trust.
Applied honestly, this framework is subtractive as often as it is additive. It implies that a serious department should be willing to remove monitoring streams that do not change decisions, not only to add ones that do. The discipline of removal is rarer in elite sport than the appetite for acquisition, but the framework points clearly toward it.
The monitoring decay cycle
There is a recognizable life cycle to over-acquisition, and naming it helps a department avoid it. A new technology is bought with enthusiasm. It is implemented intensively. For a period, staff engage with its outputs. Then, as the novelty fades and the interpretive burden accumulates, engagement declines. Within roughly a year to eighteen months, the system is still collecting data but no longer informing decisions, and it has quietly become a sunk cost that occupies budget and dashboard space without earning either.
This cycle is expensive in a way that does not appear on a balance sheet. The direct cost is the licence fee. The indirect cost is the attention diverted to standing up and maintaining a system that decays into disuse, attention that could have gone to deepening the use of data the club already held. The most valuable thing a performance department owns is not its data, it is the interpretive bandwidth of its staff, and over-acquisition spends that bandwidth on collection rather than decision.
The case for the second practitioner over the second platform
The practical conclusion that follows is uncomfortable for a market built on selling hardware and software: in many departments, the highest-return investment is not another data stream but another person who can interpret the data already being collected.
This is a recurring theme across the monitoring evidence. The clubs that extract genuine value from their systems are, with notable consistency, the ones where a specific individual owns the translation of data into decisions. That individual is the bottleneck and the multiplier simultaneously. Without them, additional data has nowhere to go. With them, existing data is more fully exploited. For a department weighing its next investment, the choice is rarely “this tool or nothing.” It is “this tool, or the next-best use of the same money” — and the next-best use is frequently the staff capacity to interpret what is already on hand.
The position a department can hold
Doing less is not doing nothing, and it is certainly not a rejection of data. It is a discipline of proportion. The position the evidence supports is this. Match data volume to interpretive capacity, and treat staff bandwidth, not storage or sensor count, as the binding constraint. Prefer a small set of validated, decision-relevant metrics to a large set of weakly-validated ones, because the latter invite action on noise. Be as willing to remove a monitoring stream that changes no decisions as to add one that does. Run every prospective acquisition through a quality framework that asks not “is this new?” but “does this produce valid information that changes a decision in our context?” (Robertson et al., 2023). And when choosing between another platform and another practitioner who can interpret what you already collect, recognize that the evidence repeatedly favors the practitioner.
The acquisitive instinct feels like rigor. Often it is the opposite: a way of appearing to address a problem — under-informed decisions — by buying more of the thing that was never the constraint. The harder and more valuable discipline is to ask what the data you already hold could tell you if someone had the time to read it.
How this series is made, and how to read it: this is editorial analysis, not a practitioner’s memoir and not a systematic review. PERFORM’s pieces are researched and drafted with the assistance of AI tools, then reviewed, edited and fact-checked by our editorial team against primary sources, peer-reviewed literature, clearly labelled preprints, industry reports, league and company announcements, and practitioners’ own published work. Where the evidence is strong we say so; where it is limited we treat it as limited; where a claim comes from a vendor or corporate announcement we treat it as a hypothesis, not proof. The views here are our editorial position, drawn from the published record rather than first-hand experience inside an elite performance department. Where practitioners are named or quoted, those words are their own. Where we couldn’t verify a claim, we left it out. And where you have the hands-on experience we’re writing about, we’d rather hear from you than pretend to it.
References
Bullock, G. S., Mylott, J., Hughes, T., Nicholson, K. F., Riley, R. D., & Collins, G. S. (2022). Just how confident can we be in predicting sports injuries? A systematic review of the methodological conduct and performance of existing musculoskeletal injury prediction models in sport. Sports Medicine, 52(10), 2469–2482. https://doi.org/10.1007/s40279-022-01698-9
Impellizzeri, F. M., Tenan, M. S., Kempton, T., Novak, A., & Coutts, A. J. (2020). Acute:chronic workload ratio: Conceptual issues and fundamental pitfalls. International Journal of Sports Physiology and Performance, 15(6), 907–913. https://doi.org/10.1123/ijspp.2019-0864
Robertson, S., Zendler, J., De Mey, K., Haycraft, J., Ash, G. I., Brockett, C., Seshadri, D., Woods, C., Kober, L., Aughey, R., & Rogowski, J. (2023). Development of a sports technology quality framework. Journal of Sports Sciences, 41(22), 1983–1993. https://doi.org/10.1080/02640414.2024.2308435