In the post Scoring Serving and Passing Effectiveness I outlined the commonly used 0-3 serve reception rating system. In the post Considerations in serve reception ratings I shared some alternative views from another blogger more based on FIVB-style ratings. Here I want to extend the discussion based on something another blogger – Jim Dietz – wrote on the subject.
Jim too begins with the 0-3 system as the starting point. What he then does, though, is to think in different terms with respect to averages.
In standard use you would average out a player’s passes based on their various ratings to get final value somewhere between 0 and 3. The problem that system has is that it can provide a really misleading picture of a player’s performance. For example, a player who passes ten 2-passes will have the same average (2.0) as a player who passes five 1s and five 3s. But are those equal averages really indicative of equal quality. Jim makes the reasonable point that they are not.
In place of the 0-3 system, Jim offers a different solution. He suggests one based on the probability of getting a kill from a given quality of pass. His suggested figures come from some data he received from multiple sources. I’ll leave you to examine them for yourself.
I will offer up two considerations to think about in all this, though.
First, is first-ball sideout percentage (FBSO) really the best basis for establishing the reception rating? Would we be better off using sideout percentage? It’s easy, perhaps, to immediately say let’s use FBSO as it most closely links to the quality of the serve reception. The reason I suggest maybe we don’t use it relates to my second consideration.
Should we have something other than a 0 for an overpass? A zero in this alternate way of thinking means we expect to lose the rally 100% of the time. This is certainly true of aces. Not so much with overpasses, especially at lower levels. At the same time, though, an overpass cannot produce an FBSO from an attacked ball (I’m not counting overpass kills here). That means scaling it’s value in terms of FBSO won’t work.
Something to consider.
Regardless of which way you go, though, you have to relate it to your level. If you don’t have the statistics available, figure out the FBSO/SO rates for your level of play. Then apply them to scaling the pass values as Jim describes.
6 Steps to Better Practices - Free Guide
Subscribe to my weekly newsletter today and get this free guide to making your practices the best, along with loads more coaching tips and information.
4 Responses
One thing–just for clarification–it’s important (to me) that stats be quick/easy to use. That’s why I keep ‘0’ for an overpass. The percentage of success is minimal, so easier to round it there than adjust the scale to something like 5-20-80-100. Yuck.
I agree regarding sideout% vs. FBSO–but since I was working with data from multiple teams, I used what they provided–won’t ever look free data in the mouth!!
Yeah, definitely take what data you can get and say “Thanks!”.
To your point about simplicity, though, the ultimate would be to end up with a 0-1, or 0 to 100% scale based on the expected outcome. So rather than saying we passed a 2.1 or 8.3 or 10.2 or whatever, we say we passed a 64% expected FBSO/SO. Not only would that link passing to outcomes but it would also make things more comparable from team to team.
In your example of 10x 2 pass v. 5x 3 and 5x 1… Using the probability/expected method, in my league, receiving all 2 passes gives an expected SO% of 65.8%. Mixing 3s and 1s gives an expected SO% of 63.3%.
And the probability of winning a sideout after and overpass is roughly 25%.
Thanks for sharing those figures. That 2.5% difference probably doesn’t seem like a lot to some folks, but it’s conceptually equivalent to a 25 point difference in your hitting efficiency. That’s a meaningful change in the context of a match, and especially a season.