When good luck turns against you: Evidence on performance responses to noisy feedback

Abstract

Providing and receiving feedback is a common practice. Yet, feedback has been shown to both enhance and, at times, impair subsequent performance (Kluger & DeNisi, 1996). A commonly overlooked aspect is that performance depends not only on ability but also on uncontrollable factors, commonly referred to as luck. When evaluations fail to disentangle the two, feedback can become over- or under-rewarding. This also affects result interpretation, since individuals typically credit themselves for success but blame bad luck for failure—a phenomenon known as the self-attribution bias (Miller & Ross, 1975). Despite its relevance, empirical evidence on how reactions to luck-distorted feedback affect subsequent performance remains scarce, largely due to difficulties in measuring true performance and noisy feedback in real-incentive settings.

We address this challenge using 8,234 team-match observations from elite European football as a natural laboratory. Football is a low-scoring sport characterized by substantial randomness, meaning that results often fail to reflect actual on-field performance. To separate skill from luck, we use Expected Goals (xG), which estimate the scoring probability of each shot based on detailed characteristics. Because shots occur more frequently than goals, xG provides a less noisy and more accurate proxy for underlying performance (Brechot & Flepp, 2020). We exploit the halftime scoreline as a salient and easily interpretable feedback signal and define the Feedback–Performance Discrepancy (FPD) as the difference between the halftime scoreline and the xG score. After controlling for team ability and contextual factors, FPD captures the exogenous, luck-driven component of feedback. To test behavioral reactions, we estimate OLS regressions with match-clustered standard errors and conduct a rich set of robustness checks, including multiple fixed-effects specifications.

Our findings consistently show that teams benefiting from good luck at halftime perform significantly worse in the second half, while unlucky teams improve. These patterns are evident in second-half xG scores (illustrated in Figure 1), pass accuracy, and long-ball accuracy. We interpret these results as evidence that teams benefiting from good fortune underestimate the role of luck—echoing the self-attribution bias—and consequently perform worse, consistent with previous evidence that overconfidence impairs decision-making. Additional moderation analyses support this interpretation. Conversely, we interpret the improvement of unlucky teams as effort mobilization in response to under-rewarding feedback.

The results have implications beyond sports, suggesting that an effective feedback strategy is to adopt a rigorous and constructively critical approach in intermediate evaluations—avoiding excessive praise and overconfidence—while ensuring that final evaluations fairly reward performance. This combination fosters motivation and preserves fairness in recognition.

References

Brechot, M., & Flepp, R. (2020). Dealing with randomness in match outcomes: How to rethink performance evaluation in European club football using expected goals. Journal of Sports Economics, 21(4), 335–362. https://doi.org/10.1177/1527002519897962

Kluger, A. N., & DeNisi, A. (1996). The effects of feedback interventions on performance: A historical review, a meta-analysis, and a preliminary feedback intervention theory. Psychological Bulletin, 119(2), 254–284. https://doi.org/10.1037/0033-2909.119.2.254

Miller, D. T., & Ross, M. (1975). Self-serving biases in the attribution of causality: Fact or fiction? Psychological Bulletin, 82(2), 213–225. https://doi.org/10.1037/h0076486