Figure 6

Results of the two distinct null models. A Comparison of the different randomizations: in the first model (career reshuffle), player i has the same tourneys, but the temporal sequence is different (same shapes in a different order), while in the second model (tourney reshuffle) some competitions are different from the original sequence of tournaments of i (some shapes have changed). The asterisk marks the unchanged time of the first win (while the shapes, i.e., the tourneys, are different). B Boxplots of the random distribution of the centrality of the first ten tournaments after players’ careers have been reshuffled. C Boxplots of the random distribution of the centrality of the first victory for each group of players after their careers have been reshuffled. D Boxplots the random distribution of the centrality of the first ten tournaments, having randomized all the tournaments among the players. E Boxplots of the random distribution of the centrality of the first win for each group of players, having randomized all tournaments among the players. In both models, different from the data, the effect of attendance is indistinguishable from the first win