A coach asked in an online forum:

Does anyone have a more precise serve receive formula they prefer over the 3-2-1-0 rating to rate a serve receiver? For example, do you ever take into consideration serve difficulty, position on the court, server’s position at contact, or type of serve?

The 3-2-1-0 system for rating passes has been around a long time. It was originally formulated by the late Jim Coleman. Some people use a 0-4 system, with a 1 for an overpass and the others bumped up. If you use DataVolley and/or work outside the US you are probably familiar with the ++/+/!/- system. I talk about these here. As outlined in this post, however, there are issues with the averages/percentages we get using these systems.

So let me take on this question from two directions.

Getting more precise with the basic analysis

A major shortcoming of the 3-2-1-0 system, or any of the variations, is that it doesn’t necessarily connect us well to what we’re after. Ultimately, we seek higher passing accuracy to facilitate a more effective attack. We want to pass as well as we can to maximize our chance to score. But what really is the difference between a 3-pass and a 2-pass, or a 2-pass and a 1-pass in that regard?

If you give one pass a 2 rating and another a 1 rating, the suggestion is the first pass is twice as good as the second. But are you twice as likely to score off a 2-pass as a 1-pass? You’re odds are certainly better, but exactly 2x better? Probably not. Depending on your team, the real odds could be higher or lower.

Likewise, a 3-pass probably doesn’t mean 50% better odds of scoring than a 2-pass.

For this reason, along with others from the post I linked above, using an average based on one of these pass rating systems should be considered a rough guide. You’ll want to look at something like Expected Sideout %, or Expected First Ball Sideout %. Here’s an article that goes into this in more detail.

Passer and server location, serve type, etc.

Once you have a methodology for actually rating passes you can start to categorize them by things like where the server is, location of the passer, type of serve, etc. This article provides an indication of how far you can go down that path with something like DataVolley.

You can absolutely use this information, and teams do in their scouting. I mentioned two examples of this with regard to serve targeting in this post. We can also use it in a self-scout fashion to identify areas of strength and weakness in our own team’s performance.

Here’s the big caveat, however.

You probably need at least 30 observations to make a reasonable analysis. And you want unbiased observations (e.g. not just against a single team or type of team). In other words, they should come from a representative cross-section of matches. The more you drill down your analysis, the fewer observations there will be in any given match. That means having to bring in more matches.

Just as you don’t want to depend on a hitter’s efficiency from only 10 swings being representative of their ability, you don’t want to do that for a passer either.

Ideal world

In an ideal world we would rate individual passers and teams overall for comparative purposes by controlling for an array of factors. Having enough data to do so, however, may put us outside the scope of time we have in which to operate. As a result, we often have to work with more broad analysis than we’d optimally like.

6 Steps to Better Practices - Free Guide

Join my mailing list today and get this free guide to making your practices the best, along with loads more coaching tips and information.