The first coach to really make an effort to explain the mechanics of coaching to me was Jim Stone back when he was at Ohio State–the only coach to lead OSU to a Final Four or Big Ten title. Jim was great–he didn’t spout dogma. He gave me resources and let me think–taking advantage of the fact my background wasn’t from playing the game. That led me to statistics and reading stuff by Jim Coleman one of the early stat/analysis gurus of volleyball (and really, any sport).
**SHORT PLUG–>If you like reading this, consider going to Amazon and purchasing Like Heck She Isn’t a Volleyball Player. It’s inexpensive and supports my efforts at going beyond the box. Go ahead…I’ll wait.**
One of the things I loved then–and still do–about Coleman’s approach was that statistics have to be useful, either to guide future practices or to help with the match while it’s ongoing. Otherwise, they are basically fluff (like the individual “Points” stat used now for volleyball). He came up with a lot of useful, brilliant stuff, including the system most people use for rating serve-receive ability.
Coleman basically rated passes on a scale of 0-3, ‘0’ representing getting aced or the ball being overpassed while the other numbers represent the number of choices available for a setter. This meant the question became what constituted a good average. For me, I always figured it was between 2.1 and 2.3…that was before I coached Emily Orrick, the best juco libero ever–she put up a 2.61 and a 2.65 in her two years here. Okay–that’s irrelevant. Anyways….
We had a tournament about that point where we had two passers total passing numbers that looked like this (I’m making the numbers up to show my argument, by the way):
- 2, 2, 2, 2, 2, 1, 1, 2, 2, 2, 2, 2, 2, 1, 3 15/28, 1.86
- 0, 0, 3, 3, 1, 3, 3, 3, 3, 0, 1, 2, 3, 3, 0 15/28, 1.86
Those two passers are statistically the same. Both averaged 1.86 options/pass. But are they the same passer? I don’t think so.
So I started grinding numbers and realized that there’s a problem with the math. When we use the 1-2-3 system, it is set up so that a ‘2’ is worth double a ‘1’-value pass, and that a ‘3’ is worth 50% more than a ‘2’ and 300% more than a ‘1’. That seemed off. My gut told me that a perfect pass should be much more valuable than a pass where the setter had to forearm it or a non-setter played the ball (my gut’s only partially right as you’ll see).
With information provided by multiple coaches, most memorably Todd Dagenais at Central Florida, Penn State assistants (from both men’s/women’s teams), and Pete Hanson from the Ohio State men’s team, along with stats from the NJCAA-level and a couple HS programs, I was able to put together some serve-receive statistics. What I found was that the 1-2-3 sequence wasn’t valuing things properly.
The chances of scoring based on the pass quality (in percent). Remember, you can score on a ‘0’ since that includes overpasses which provide a chance for an opponent to make a mistake. :
- “0”: 2
- “1”: 12 An increase of 600% over a ‘0’ (instead of infinite…)
- “2”: 44 An increase of 367% over a ‘1’ (instead of 100%)
- “3”: 53 An increase of 21% over a ‘2’ and 442% over a ‘1’ (instead of 50/300)
Now–part of Coleman’s philosophy is that we MUST keep things simple. Tallying things as 2-12-44-53 isn’t easy during a match, but…rounding numbers is! But let’s look at this a different way.
On a ‘1’, I have a 10% chance of a kill, thus with normal pass-rating, a ‘2’ should create a 20% chance of a sideout (because it’s value is double) and a 30% chance on a ‘3’ (where the value is treble). Instead we get 10-40-50.
So why not change your pass rating system to 0-1-4-5 instead and get the true value of a pass as the rating? Well, first we need to know if it can differentiate better than the old system, so let’s go back to those two passers earlier…
NORMAL RATING SYSTEM
- 2, 2, 2, 2, 2, 1, 1, 2, 2, 2, 2, 2, 2, 1, 3 15/28, 1.86, 1 perfect pass
- 0, 0, 3, 3, 1, 3, 3, 3, 3, 0, 1, 2, 3, 3, 0 15/28, 1.86, 8 perfect passes
MODIFIED FOR S/O SUCCESS
- 4, 4, 4, 4, 4, 1, 1, 4, 4, 4, 4, 4, 4, 1, 5 15/52, 3.47
- 0, 0, 5, 5, 1, 5, 5, 5, 5, 0, 1, 4, 5, 5, 0 15/46, 3.07
Uh, oh–now we’ve got a big difference. The normal system gives us an idea of the average number of options a setter has from a pass, but the second is more important–it provides an idea of our expected sideout percentage from the passes, and while it sounds simplistic, the reality is whoever scores the most points wins a volleyball set.
What we now see is those two passers are not equal. While the second passer is perfect more often, her problems with the other serves drops her value significantly. Heck–if we replace those ‘0’ with ‘1’, the modified value will STILL be lower than the first player’s, even though the 0-3 system numbers are now superior.
Ahh, but there’s other important stuff here, not just a better way of comparing players to know who is performing better. The progression of the value is no longer linear. There’s a huge jump in value from a ‘1’ to a ‘2’ and a much smaller uptick from ‘2’ to ‘3’. This means you get more bang for your buck improving poor passes than working on making good passes perfect.
There’s a sub-lesson there, too. Youth coaches, by reflex it seems, pull some players out of passing duties because they aren’t good–but if this valuation holds for defense as well as serve-receive, then shouldn’t we work on passing with middles/others who get pulled from the back row? Even if they are only playing defense while they serve, turning them into mediocre passers can have huge benefits (along with things like self-confidence). Aiming for those perfect passes regularly, when the improvement over being consistently good is not significant, that’s time that could be spent improving those athletes’ other skills as well–I suspect that the value of improving other skills from poor to average is more valuable across the board than good to great (below the highest of levels).
Consider what this means for how you run a practice? Are you maximizing your chances for improvement, making it easier to win? Are you preparing your athletes fully for the next level of play?
**If you liked this and have some money burning a hole in your pocket, consider sparing some change and donating to the Dietz Foundation, my educational charity (501(c)3, designed to advocate for education through non-traditional means. The website’s up and the first creative project (a boardgame) is getting ready for production already. Your help can provide scholarships for kids looking to go to college and become tomorrow’s educators!