One of the major functions of accounting is to report unbiased numbers. In this paper, we demonstrate how favoritism may bias results and also how greater transparency can mitigate this bias. We examine whether one subtle feature of transparency – the physical presence of a stakeholder during the evaluation process – mitigates favoritism biases in subjective evaluations. Using archival data from professional ski jumping, we find that, controlling for objective performance indicators, subjective evaluations suffer from favoritism. In particular, evaluators favor athletes of their own nationality and athletes that have a compatriot on the evaluation panel. We test our transparency hypothesis taking advantage of the situation where, during lockdowns, sports competitions took place without an audience on-site. We predict and provide evidence that the physical presence of an audience during the decision-making process is associated with lower levels of favoritism in subjective evaluations. As such, we contribute to the accounting literature by highlighting how transparency may decrease the likelihood that subjective evaluations are biased by favoritism.