Optimal Choices – Serve


Something I find fascinating is the idea of making the right choice. As coaches, we often have conversations with players about their decision making process – whether it’s asking why she set the quick in that situation, why she dove into the angle on her block move, why she served the libero, why she tried to tip that ball, etc. Sometimes this leads to a constructive chat about their thought processes and sometimes this leads to a coach giving the player their own answer to the situation. In either case, the ultimate goal is to assess the choices available and chose the right one. But what is the right one?

How do we make the right choice? I’ll ask you to suspend your preconceived notions about volleyball for a moment and let’s approach this with the naiveté of a beginner – in the truest sense of the word.

If we look at this question in the context of serving, you’ll see my solution above. We’ve established from earlier posts that looking at your opponent’s FBSO eff may be the best way to evaluate serving. With this in mind, let’s look at what we can control when it comes to serving. I would argue that there are 3 things: where you serve from, where you serve to, and the type of serve you hit. If we factor in all 3 of these, our sample sizes shrink a little too much, so we’ll reduce this to: where you start and where the ball ends up.

In the viz above, I’ve eliminated combinations of start & end zones with fewer than 10 attempts per rotation. I also thought about including the serving team and their rotations into this, but the more I thought about it, the less important I felt it was overall.

And while you could drill down deeper and add more context, this viz acts as a cheat sheet for serving at the listed opponents. The efficiencies you see listed are the serving team’s maximum Point Scoring Efficiencies (basically in inverse of FBSO eff, so higher is better for the server here) in each of your opponent’s rotations. Start zones go down the y-axis, end zones across the x-axis. Keep in mind that startzone 9 is between 1 and 6, startzone 7 is between 6 and 5 along the endline.

Go ahead and find some of the darkest chunks in the viz, like serving against Illinois’ Ro2 from 9 to 7 – a PS eff of 0.600 on first balls. Or from 9 to 1 against OSU’s Ro2 (0.364). There are certainly start/end zone combinations with much worse results for the serving team – and so it becomes odd that a team would choose NOT to serve in this fashion.

This is of course a preliminary look at how this information could be used – and we could certainly break it down farther by looking at set distribution after you serve into Illinois’ Ro2 from 9 to 7. When you attack that pass-hitter, where do they go? On a perfect pass vs. a medium pass? Can we design our blocking scheme around this information. Based on the data, can we predict with a certain level of confidence what the opponent is going to do before they do it? This is the crux of scouting isn’t it – knowing your opponent as well as they know themselves?

The nice thing about serving is that it’s closed loop. You can tell a player, hey, Illinois is in Ro2, let’s attack the pass-hitter from zone 9 and serve it down the line. It’s not a skill that requires reading. You’re not looking at the approach, the shoulder – you’re not watching the wrist or trying to outsmart the attacker. You’re in control of your actions and you’re playing the numbers. Like a pitcher knowing where a batter is weak. Let’s attack the angles of the opponent serve receive where we see a chink in the armor. Maybe its 5 to 5 or 1 to 1. Maybe it’s straight down the sideline. Who knows…

The data knows. Put your gut feelings away for a minute and see if the numbers support what you’re spouting. Pump the brakes on alternative facts and look at what’s going to keep your server on the line longer.