Score contribution per author:
α: calibrated so average coauthorship-adjusted count equals average raw count
Under optimal tournament design, equilibrium effort is invariant to the shape of the mean-zero additive stochastic component, often referred to as a “shock” or “noise”. We report data from laboratory experiments providing the first test of this prediction. Consistent with theory, we find that average effort does not significantly differ between a negatively skewed and uniform shock distribution. In addition, we test a second theoretical prediction that, in winner tournaments, when the shock distribution is asymmetric as in our design, one should exert minimum effort whenever one's competitors are exerting above equilibrium effort. With a symmetric shock distribution as in our design, efforts should generally remain substantial, even when one's competitors are exerting effort above equilibrium value. Our data reveal that subjects actively engage in the tournament even when faced with aggressive competitors under both shock distributions.