Getting More Out of Volleyball Stats - Derived Stats and Thinking Out of the Box

I like stats. I really like stats. Good stats give you fairly objective information about how your team is doing, how your opponents are doing, and inform your coaching decisions both in game and in practice. The problem is getting good stats.

While coaching high school volleyball, I was required to provide the number of kills, blocks, aces and assists for my guys after the match. At a minimum that requires one manager, assistant, or player to tally those stats. Unfortunately these stats by themselves don't really tell us anything meaningful. To get the most utility out of these stats, I had a second person tackle hitting stats and tally attempts (0), errors (E), blocked attempts (B) and kills (1). This at least could give me hitting efficiency, (K – E – B)/(K + E + B + 1), or the handy In Play Efficiency, 1 – E/(K + E + B + 1). I could have gotten by without the blocked attempts, but it is useful to differentiate between a loss of point because of hitting into the net or out of bounds, and a loss of point because of hitting into the block. By using a 4 color pen and a different color of ink every set, I could see team and individual hitting efficiency by set. That would at least give me some idea of a player starting to struggle that might need encouragement or a chance to rally himself. I could look take a look at the assistant's or manager's clipboard and see that one of my players had been blocked on the last 3 attempts, or hit the ball out the last 4. I could also see that hot player with a bunch of kills in a row. It's all there in front of me, outside of possible bias (see: Confirmation Bias). It's something, and I was grateful for the gain in information, but I will always want more information. The consistent challenge is having the resources to get more information, either in cost of hardware and software to gather the stats, or the personnel (numbers and/or ability) to put in the effort to collect the data.

I spend a fair amount of time trying to come up with new ways to gather more information out of the same base resources, and I have come up with an idea that could yield some good additional information out of the same hitting stats. Instead of just putting a 1 for a kill, now put a 1 for a kill or a T (for Tool) for a kill off the block. E is still the same for an error, and B for a blocked attempt. 0 is still an attempt that was dug by the opposing team, but now there is also a C (for Cover) for an attempt that was covered by your team. A final new one would be a D for one touched by the block and then dug. I suppose 0 and D could be swapped if that makes more sense to the stat taker. The 1 and the T are fairly similar in appearance and what it would take to write, and both represent a point. E and B are similar enough and represent a point for the opponent. 0, C, and D are similar enough and represent an attempt without a point one way or the other. For the stat taker this doesn't add that much more to the job, but it could give some good information for the coaches.

For one, this could give the percentage of attempts that the opponent's block touches, (T + B + C + D)/total attempts. It could also give the percentage of times a ball that comes back to your side of the net is covered, C/(B + C). It could also give the percentage of times you score when the ball is touched by the block, T/(T + B + C + D). Perhaps a blocked hitting efficiency, (T – B)/(T + B + C + D). This data could give some good information to the coaching staff, I think particularly a cover percentage stat.