Case example: Gold medal match, world championships: Poland v Brazil
Let‘s first look at some figures from the final.
We could say that this was a match on an equal playing field but with a slight advantage for Brazil. But Poland won the final 3:1. This immediately indicates a weakness in the figures. The psychological component cannot be measured. The only explanation for Poland winning is that they played well when it mattered, but on average they were not as good as Brazil.
Let’s look more closely at the serve. Poland had an error rate of 16% and Brazil of 18%. I am sure that most people will think, “That’s relatively high. Almost one serve in five was a mistake.” Now let’s give this number a context. The average error rate of the team that won gold in the last six gold-medal matches (WL, WC, EC) is 20%. To take this further: there is a negative correlation between serve errors and the passing value of the opponent, and a positive correlation between serve errors and PS%. This means that the higher the serve error rate, the worse the passing of the opponent, and the higher the scoring probability of one’s own serve. This makes sense, as a higher error rate in the serve generally means a high serve force. Here are a couple of extremes:
World League 2014: USA beat Brazil in the final 3:1 with the USA Serving E%=19%, Passing: 2.54, PS%= 38% and BRA Serving E%= 10%, Passing= 2.26, PS%= 36%
World League 2013: Russia beat Brazil in the final 3:0 with RUS Serving E%=25%, Passing: 2.29, PS%= 44% and BRA Serving E%= 8%, Passing= 2.27, PS%= 32%
So one can say that Poland served well for a gold-medal match and found the right balance between serve force / flow of play and service risk. And Poland played well when it mattered.
In order to better understand the serve forces of Brazil and Poland, the data is visualized here.
It can be clearly seen that Poland consistently served between the players, whilst Brazil served at the players. This data visualization provides meaning and says more than simple figures ever can.