The basic skills of Dodgeball - Throwing and Dodging
I previously showed how throwing relates to hit percentage. In this post, I will show how hit and defense percentages relate to point scoring ability, and I will explain how one can estimate the technical skill of players using an index based on these two values.
Take Home Messages
- Throwing skill is twice as important as dodging ability
- Throwing and dodging skill together determine around 75% of an individual’s ability to score points
The plot below shows the hit percentages and points scored per minute for each player in the Swedish Mixed Team, averaged from all games during the 2018 Euros. Unsurprisingly, there is a positive slope and statistically significant correlation.
Correlation between hit percentage and points scored per minute for the Swedish Mixed Team, averaged across all games from EuroDodge 2018.
As I mentioned in the previous post, if one is above the line, then the score is higher than expected given their hit percentage, and if one is below, then one has scored less than expected. For example, Female2 has an unexpectedly high scoring rate for her low hit percentage. Theoretically, this could be because she is a bad thrower and throws 10 shots per minute but rarely gets eliminated, but in this case, it is because she had a surreal amount of catches which are worth 2 points (opponent gets eliminated and the team gets one player more on court). Similarly, Male3 underperformed given his hit percentage, which is likely due to the fact that he had very low hit and defence percentages and usually got eliminated in the first 30 seconds of each set, leaving no time to score positive points.
Similarly, defence percentage also shows a positive slope, but slightly less so than for hit percentage. The outcome when a ball is thrown is more dependent on the skill of the thrower than that of the defender. This gives a slightly lower correlation, with greater spread of values above and below the line and lower slope than for hit percentage. This is because there is lower variability between players, i.e. they are more similar, in terms of defensive ability.
The plot above shows both of these values, and how they relate to rate of point scoring (size of markers). I’ve also included lines of equivalence, which indicate how one can move across the plot without affecting point scoring ability; players along the same line are equally good in terms of total technical skill. It shows that throwing is relatively more important, since movement along the x-axis leads to crossing of more lines than the same amount of movement along the y-axis.
Using something called multiple regression I found that hit and defence percentages account for about ¾ of the difference in points scored per minute (this leaves ¼ to be explained by other variables, which I will outline in a later post). It also showed that the relative importance of these skills differs. As illustrated in the plots above, and corroborated by the multiple regression analysis, the hit percentage has roughly a two times bigger effect on point scoring than the defence percentage. In other words, one will have a greater return of investment in terms of success on the court if one spends time practicing their throw rather than their defensive ability, and it is more likely that the team with better throwers win over a team with equally good defensive players.
Given that these two values are easily observable, I calculated their weighted average to get what I refer to as technical skill, which is a percentage indicating an individual’s overall success rate when attacking and defending.
Since throwing is twice as important as defending for the rate of scoring, I weighted their percentages accordingly (sum of 2*hit% and defence% divided by 3) to get a weighted technical skill score which has a higher correlation with point scoring ability and probability of winning. The plot below shows how this weighted technical skill relates to rate of point scoring; there is a high and statistically significant correlation, and an improvement in technical ability of only 1% is expected to result in hitting one more opponent, or one less elimination for the player, each 6.67 games.
Since the players are not exactly on the line, there are additional factors that affect point scoring which are not included in the technical skill index. Since throwing and dodging are the two most clearly observable and identifiable technical skills, I’ve referred to everything else as Tactical Skill. Things such as positioning on court, deciding whether or not to take a certain shot, taking a risky shot and then being exposed without a defending ball, speed on the court etc. all determine how well a player utilises its technical skill in order to score points. As previously, the players that are above the line have scored more points than expected from their level of technical skill, and vice versa for those below the line. The distance from the line is determined by a player’s tactical skill and sums up the last quarter of what determines point scoring ability. More on that in the post about tactical skill.