Wednesday, June 6, 2018

Thoughts on Evaluating Passing in Volleyball

Given that receiving serve is a team's first and best way to create a chance to win the rally, evaluation of passing should take into account how the pass tips the balance between the offense and the defense. I am not proposing evaluating passing in a novel way, as I think that the three-point scale is a fine tool (as is a four-point scale but more on that later). I do think that it is worthwhile to look at how the evaluations we make should be determined. Pass location, speed, and height affect the efficiency of the offense and, therefore, passes should be evaluated on these criteria. I think that it is important that the evaluator recognize two factors that change as pass location, speed, and height change, what the pass allows the setter to do and how much pressure the pass allows the offense to put on the defense.

The location of the pass is the most important factor that determines how effective the offense can be. While teams may differ in where they consider a "perfect" pass to be, these differences don't have a significant effect on how pass location determines offense efficiency. What does matter is how well middle attackers make themselves available as the setter moves away from "perfect". This is what expands the range of passes from which a setter can effectively set all three front row attackers. If middles cannot make themselves a viable option then the likelihood of scoring decreases significantly. This idea blurs the line between skills. Without changing passing at all, we can make our first ball offense better by making our middles better, which allows us to grade our passing differently. You can see this in the higher levels of the game, where a ball passed in zone three on the attack line can still be considered a three-option pass but that same pass at lower levels is, at best, a two-option pass. So there is a balance for us to find in improving our side out efficiency; we can spend time improving our passing but we can also spend time improving our middles to create the same effect. Regardless of how we choose to spend our training time, we should recognize that our evaluations of passing should take into account how effective our middles can be based on the pass location.

Speed and height of a pass have an effect on side out efficiency as well. Like "perfect" location, "perfect" speed and height are subjective. What matters is that we look at how those factors change offensive efficiency. The differences here are in levels of acceptable risk and their accompanying rewards. Passing faster/lower puts more pressure on the defense by giving them less time to gather information and respond to the offense. But unlike some changes in location, passing faster also increases the pressure on the offense. Faster passes mean less time for the offensive players to respond too, especially the setter. Passing faster generally means that location must be better but if location is better than the resulting pressure on the defense is higher. One little-recognized side effect of passing lower/faster is that the setter has to make body position changes to receive such passes, which can make a difference in set accuracy. Passing higher/slower may not increase pressure on the defense by itself, but doing so relieves pressure on the offense, which returns us to location being the determining factor for offensive efficiency. When evaluating passing, we should consider the height and speed of the pass to determine how much pressure the pass puts on the offense as well as on the defense.

Pass location and speed/height clearly have effects on what the setter is able to do with the pass. In my opinion, what the pass allows the setter to do and what the setter should do are two different things. There are arguments to be made for if/how each of those should effect the evaluation of the pass. If the pass allows the setter to set all three front row attackers, this is a situation that the defense has to deal with. That defense may not know what the offensive tendencies of their opponents are so they must prepare to defend against any of the three front row options, even though the setter has been trained to eliminate an option from consideration. So I prefer to evaluate passing in a more system-agnostic way, meaning that I prefer to consider what the setter can do more than what the setter should do in their offensive system. I choose to do this because I think that the defense usually knows less about what may happen than the offense does so the pressure on the defense remains the same even though the system limits the offensive options. (To be nerdy, this demonstrates that volleyball is a game of imperfect information with simultaneous moves.)

I think that the best way to describe the pressure that an offense can put on a defense is using a four-point evaluation scale. The only difference between a four-point and a three-point scale is the four points available for a "perfect" pass that the three-point scale does not offer. I see the benefit of scoring the "perfect" pass higher as a way to distinguish passes that put a maximum amount of pressure on the defense. In my opinion, there is a very large range of passes that can be considered to allow three front row options to be set but not all of those passes are created equal. A "four" pass allows the setter to be balanced, neutral, and likely jumping when setting. A "four" allows clear runways for all attackers since the setter is not moving into any attacker's path. These factors combine to create a maximum amount of pressure on the defense because the offense is best poised to run as it is designed. As the setter begins to move to receive the ball, at least one of these advantages begins to degrade. Even though the setter can still effectively set all three front row attackers, there are slight changes in the pressure placed on the defense as tempos, angles, and approaches change for the offense. The changes may be small but they are potentially important to each team's ability to win the rally so there is a case to be made for distinguishing between a "perfect" pass and all passes that allow effective setting of all front row attackers. I recognize that those effects may be small enough that they may not be important at all levels of play so I do not feel that evaluating on a four-point scale is necessary but it is certainly useful, especially for levels of play where small advantages can create important differences in outcomes.

There is more that we can do to get at how passers help their team's offense. Looking at passing averages are where we start but we can easily dig deeper to better understand what those averages might mean. As an example, look at three passers who all have passing averages of 2.0.
Passer 1 Passer 2 Passer 3
2 3 3
2 3 3
2 3 3
2 3 3
2 3 3
2 1 3
2 1 1
2 1 1
2 1 0
2 1 0
If we only look at passing averages, these three are the same but we have two other numbers that we can look at that can tell us more about the passers. First, let's consider In System Percentage (IS%), which tells us how frequently the passer puts the team in a position to set all three front row attackers. (If there isn't much of a difference between how likely you are to side out on a 2 versus a 3 pass at your level, then feel free to change how you define IS%.)  Passer 1's IS% is zero, Passer 2's is 50% and Passer 3's is 60%. A coach may value passers 2 or 3 more highly because being in system 50-60% of the time may give the team a better chance of siding out than a steady stream of two-option passes. The next number we can consider is Zero/Error Percentage (0% or E%), which tells us how frequently a passer is aced. While passers 1 and 2 are both at 0%, Passer 3 is at 20%. A coach may be more hesitant about that passer relative to the other two because 20% of the time, they immediately lose the rally for their team. This example isn't intended to make a clear case for any of the three passers, just to show how we can learn more about our passers just by looking at the same data in a slightly different way.

There are more advanced metrics that we can consider, like FBSO% and xSO% but these metrics are more about how effective the offense is rather than isolating the passer. While the pass definitely has a strong effect on how good the offense can be, this post is intended to be more about evaluating the pass itself. I'll describe these two metrics in another post.

n.b. I put quotation marks around "perfect" and "four" because I don't want to give the impression that there is a Platonic "perfect" or "four" pass. I could have done the same for "three" or "two". What makes a pass "perfect" is dependent on the level of play, the setter, and offensive philosophy. "Perfect" should be whatever you think it should be in your conditions. I suggest that you figure out what those terms mean based on the criteria I present above.

What I Did on My Summer (and Spring) "Vacation"

I noticed recently that I had not posted much in the past few months. There is a part of me that feels bad about that but I also realize tha...