.png)
Which Statistics are Critical to Success in Volleyball?
Prologue
One of most enduring interests I have in performance analysis is understanding which stats are critical to success in volleyball. If you know that then you can then focus on those skills (that the stats measures) in training and matches and this should in turn help you win more matches than you would have otherwise. This is really what performance analysis is all about.
This interest was actually the subject of my Masters dissertation in 2013, where I looked into the identification of valid performance indicators in non-elite volleyball. I revisited this topic again during the first lockdown in 2020. I had statted so many more matches since 2013 that I thought I could redo my Masters dissertation with different sets of data, which differentiated between gender and level.
There were some surprisng results so I wanted to compile them all together to see if there were any broader patterns that we could identify. I did just that.
Introduction
Just as a reminder, normalises data into percentages or averages so it creates a level playing field so we understand and compare the levels of performance. Performance indicators are when they are associated with a successful performance or outcome. This is important because it provides a deeper understanding to what is critical to success. in volleyball can be defined as final league points, winning or losing a set (set outcome) and winning or losing a close set (close set outcome, i.e. a set score with a 2 point difference, e.g. 25-23).
I have done five valid performance indicator studies, all using different datasets. You can read them on the links below. I would say it is worth a read because it gives a nice explanation of what I did before, the purpose of the new study and a detailed explanation of the results. I also included a literature review of valid performance indicator studies published in research articles as part of my masters dissertation.
5 | City University London + Opponents (BUCS South Eastern Division 1), 2017-20, Set outcome 4 | IBB Polonia Men + Opponents (Men’s Super League), 2018-20, Set outcome 3 | Women’s Super 8s + League, 2015-20, League correlation (2017-18) + Set outcome 2 | Women’s Super League, 2019-20, League correlation + set outcome 1 | Men’s Super 8s, 2012-13, League correlation + set outcome (Masters Dissertation) 0 | Academic Literature Review, elite men’s volleyball (2004 – 2012)
One thing is clear from all my studies in the identification of valid performance indicators in non-elite volleyball: it is the best and second best valid performance indicator is the points efficiency ratio [PER = Points Won/ Points Lost] and hitting pct. [H% = (Kills – Errors)/ Total Attack Attempts], respectively. This would make sense because to win a match, you need to win three sets and to win one set, you need to score 25 points before your opponents does. So the ability to concurrently score points and not lose points in errors should be critical to success. Attack would make sense too as most points won and lost are done via the attack.
However, there were some surprising results that came from the above studies. The main one was that Defence In-Play Pct. and the Block-Defence In-Play Pct. looked to be highly valid in some studies. The literature review into valid performance indicator studies that I wrote for my Masters dissertation showed mixed to weak support for defence as a valid performance indicator. My Masters dissertation results showed no significant results for defence as a valid performance indicator. Because of this, I wanted to compiled all the studies together into one table to see if these performance indicators had a global significance and to see if there were any other patterns.
Method
Table 1 shows what the sample consisted of in each study. I’ve also added another dataset of the Men’s Super 8s, 2013-14 season, from my work at the Super 8s Statistics. Due to the basic stats I collected, I could only look at the league correlation of the points efficiency ratio and hitting pct. only. All sets were video recorded. Performance data statted with the DataVolley program post-2015 and with StatEasy pre-2015.
Performance indicators being tested and its calculations can be seen in Table 2. There are major differences in the operational definitions of the stats taken prior to 2015 (when using StatEasy) and after 2015 (when using DataVolley). The big one is that how I rated the reception quality, and subsequently the serve, is much different. The zone for a perfect and positive pass was a lot bigger prior to 2015 than it is after 2015 (read more here on how I statted the reception quality changed over time). I also coded defence stats prior to 2015 but I have not done so after 2015. To calculate defence stats now, I used opposition attack zeros as a proxy measure for digs and opposition attack kills as a proxy measure for missed digs.
Pearson’s correlation was conducted to identify correlations between the performance indicators and final league points. This produces a r-score. Basically, the closer the r-score is to 1 then the closer to a perfect correlation it is. Wilcoxon signed ranks test was conducted to identify differences between winning and losing performances within a set. This could be differentiated with a specific point set difference. This produces a z-score. Basically, the bigger the z-score, the bigger the difference. Both r and z-scores will come with a p-value. We are looking for any p-value below 0.05 (typically written as p < 0.05). It basically means there’s a 95+% probability that there is an actual, or , correlation or difference. The smaller the p-value, the higher the probability that there is a correlation or difference.
Results
Table 3 shows the results of the five studies I conducted. I excluded all results showing set outcome for all sets since all performance indicators were shown to be highly significant (p < 0.00001). It basically shows winning teams perform significantly better than losing teams in all skills measured. I limited it to sets that were close, i.e. with a 2 or 3 point set difference.
The point efficiency ratio still looks like the most valid performance indicator out there. The only condition where there is no significant results is when there is a 3 or 2 point set difference in the IBB Polonia London matches in both the 2018-19 and 2019-20 seasons. However, these are based on 19 and 11 sets, respectively, which is a really small sample. Hitting Pct. still looks to be the second best valid performance indicator, as there’s significant results across the board. There’s mixed results on whether the Hitting Pct. is done after the pass or in transition. I think it’s more to do with who can attack best in all situations.
There is support for the Block-Defence Pct. and Defence Pct. but it looks like only for the Women’s Super League and not for the Men’s Super League or BUCS level. There’s good support as a correlation to league points (p < 0.01). There’s a highly significant result for both Block Defence In-Play Pct. (p < 0.00001) and Defence In-Play Pct. (P < 0.001) in close set matches (2 point set difference). This means there’s a 99.999% and 99.9% chance, respectively, that the result is accurate!
Service rating looks to be only correlated to league points in the Women’s Super League only. I am surprised that there’s no significant results for reception rating, outside of my original Masters dissertation.
Block Efficiency Pct. has no significant results across the board. Again, this makes sense because not many points are won and lost via the block. This would mean a better performance for the block is needed. Perhaps a Block Rating is needed, e.g. 0 points for a block error, 1 point for a block out, 2 points for a block touch, 3 points for a rebounded block and 4 points for a block kill. However, it should be noted that block touches (blocker touches the attack but play continues on blocker’s side) and block zeros (block attempt but did not touch the attack) are not included in these calculations.
Discussion
I expect the points efficiency ratio and hitting pct. to be highly valid across any performance indicator study based on this simple premise: to win a league or tournament, you need to win matches, to win a match you need to win three sets and to win one set, you need to score 25 points before the opponent does. So the ability to concurrently score points and not lose points in errors should be critical to success. This is what the points efficiency ratio measures. And since most points are won and lost via the attack, the hitting pct. should be critical to success too. I would expect that the points efficiency pct. would be an even better performance indicator because it takes into account total point scoring attempts.
The Block-Defence Pct. and Defence Pct. looks to be valid but only in the Women’s Super League. This makes sense because if hitting pct. is critical to success in volleyball, then it must hold true that the opposition’s hitting pct. will play a factor too. The way the two performance indicators are calculated is simply the the inverse of the opposition’s kill percentage. I am planning on statting defence stats in the future to improve my statting skills, so that’ll be good to get some defence stats for individual players (the current calculation can only be done for the team).
I am really surprise how reception rating yielded no significant results outside of my masters dissertation. After the points efficiency ratio and hitting pct., the reception rating is the next performance indicator I tend to look at. I think the likely reason is how I rated the pass in 2012-13 is vastly different to how I rate the pass from 2015 to 2019 (see below). The perfect and positive passing zones were much bigger back then than it is back in 2019. Also, there is a very steep gradation between a positive pass and a negative pass in 2015-19. A pass that lands just outside the positive zone is rated the same as a pass that lands beyond the baseline. Clearly both are not the same in terms of attacking options.
I did address this last season (2019-20) when I added a neutral pass in my statting (see below). It provides a nice bridge between a positive and negative pass. I arbitrarily gave the neutral pass a weight of 1.5 because I wanted to keep the passing rating to a 3 point pass. However, upon some research, the weights I use for the 2020/21 season is a better reflection of its importance. It would be interesting to see if these weights will change the results so it becomes a valid performance indicator.
Conclusion
The results indicates that focus should still be in point scoring and attack performance indicators. More focus should be given to the block and defence stats and less focus on the reception.
Comments