As I have already mentioned in this post, this post and in many others, I am interested in testing the assumptions with which we carry out our daily work. One of the assumptions that we work on, is that timeouts are useful. To test this assumption, I participated (with reader Ben Raymond, who did the actual statistics parts, i.e. the ‘work’) in a study of timeouts in the Polish and Italian men’s leagues from the last season. For the 2015/2016 Italian League, we were able to obtain data all matches except one (ie 131/132). Our results are effectively the same as the results for the Polish Plus Liga, with one tiny difference. But I don’t want to get ahead of myself.
The basic observations we noted for the Polish League are replicated in Italy. Taken together this suggests that coaches at least implicitly follow a similar strategy in their usage of timeouts.
- The number of timeouts taken by each team reasonably correlates to the set win / loss ratio of the teams.
- The frequency of timeouts taken steadily increases over the course of a set until the opponent has a score of 21. In Italy, however the technical timeout is taken at 12.
- Looking at score differential, at nearly all periods of the set, teams are more likely to call a timeout while trailing than while leading. The favourite score to take a timeout is 24-22.
- Coaches most often take timeouts with a point differential of -2, -3 and -4.
One assumption, or conventional wisdom, that I have written about is that there are more service errors after timeouts. We did not find that to be true. As in the case with the Polish League the error rate after tactical and technical timeouts was the same or less than in general play. This could be due to coach’s instructions, or player experience, or the server’s concentration being broken.
|Serve category||Serves||Errors||Error Rate|
|First of set||506||51||0.101|
|Followed tactical timeout||1478||253||0.162|
|Followed technical timeout||471||66||0.136|
To test whether in fact players did approach serving differently at different match junctures, we looked also at ace percentage and perfect pass percentage. The overall results in these categories differed slightly from the Polish League service errors in that aces and perfect reception rates were essentially identical in general play and after tactical timeouts.
In the Italian League it seems that serving is not affected in any way by a tactical timeout.
As in Polish League, roughly two thirds of all serve attempts in the Italian League are won by the receiving team. This figure is similar in general play, on the first ball of the set, after a tactical timeout and after a technical timeout. The difference are not statistically significant.
|Serve category||Opp Serves||Sideouts||Sideout Rate|
|First of set||506||322||0.636|
|Followed tactical timeout||1478||1008||0.682|
|Followed technical timeout||471||294||0.624|
As with the Polish League there is no significant effect of a timeout taken at any point in the set. The same goes for any score differential.
Service Series / Runs
The final area that we considered in the project was how these parameters changed during a service series. In contrast to the Polish League, we found that the likelihood of service errors DOES significantly change during a series. The service error rate is significantly less on the first serve of a series than on subsequent serves. As with the Polish League, sideout percentage does change, for the worse. After the third serve in a series, sideout percentage decreases but not decrease is not significant. In this part of the analysis, we did find one specific situation in which taking a timeout significantly improved the sideout percentage of the following rally. That situation is in when a timeout is taken after zero break points have been won. That is, BEFORE a run begins. My best guess is that all of those timeouts are taken at the end of a set when the score is equal.
The findings of the study are taking a tactical timeouts is effectively the same as not taking a timeout. That is tactical timeouts have a negligable impact on the game. The single exception is the timeout taken before a service series has begun. Most likely such timeouts would be taken at the end of a set. Given that other timeouts are ineffective, the findings of this study suggest that keeping timeouts in reserve until the end of the set would be a valuable tactic.
Further posts in the series can be found here.
The full data from the Italian League can be found here.
The commentator just mentioned the posts regarding timeouts 🙂 Positively commenting on you challenging conventional wisdom.
United States v Italy – Group 1: 2016 FIVB Volleyball World League
There is hope 😉
Thank you very much for collecting, evaluating and interpreting these wonderful data, Mark & Ben.
I take you to conclusively show that taking timeouts per se doesn’t increase the chances of winning the next point – save the mentioned exceptions before a service series and some situations which are so rare that you can’t expect significant results.
However, I’m sympathetic to the idea of not giving up conventional wisdom to easily neither 😛 Here is a hunch: There are good timeout teams or coaches and there are worse. So my question would be, whether there appear some particular teams in your data, that are significantly more successful after timeouts – be it the timeouts taken by themselves or by the opponents – than the average team in the respective league.
If that would be the case, it would be interesting to see, how many timeouts they have taken and at what point of the match, in the second step. But maybe the number of timeouts taken by a specific team in one season are to small anyway to come up with something conclusive.
I think that Berti might be on to something and would be interested in the results of his query as well. Based on the info above you might be prone to saving timeouts for later in the set which for myself would be true. I always try to have at least one timeout remaining after 20. I think the other question I have that ties in with Berti would be how many coaches call timeouts to be perceived as doing something? In other words it is better to be perceived as doing something, because appearing to do nothing is worse. I think you have written about that as well in a previous post. I also wonder if timeouts make more or less difference depending on the level and quality of coaching. Nonetheless, I love the information you have provided.