Serve-Receive Evaluation…5-4-1-0 isn’t quite correct either….

So, I could write a bunch (have written a bunch) on serve-receive. A Proper Evaluation of Serve-Receive and Proper Serve-Receive (The Director’s Cut), for instance…

For those of you under 30, here’s the TL;DR:

3-2-1-0 is not a good system for evaluating passes.

For everyone else:

The 3->0 system treats the differences between a 3 and 2, a 2 and 1, and a 1 and 0 all equally. The thing is, they ARE NOT equal differences. You improve your offense 9,940% or so when you go from 0 to 1 and you improve 400% when you go from 1 to 2 and only 30% or so going from 2 to 3. Thus, the passing scale needs to be unbalanced to reflect our chances of scoring based properly on our hitting in those situations. Make sense? That’s how I got to 5-4-1-0…though that rounded things off to numbers easy to keep track of.

Reading up on baseball metrics, it dawned on me that my 5-4-1-0 fix isn’t the answer either! There are two big problems with it.

First, it takes into account the chance of a kill–but it doesn’t take into account things such as pushing the opponent out of system and forcing errors. A good pass leads to a good swing which can keep the opponent from effectively counterattacking. Oops.

So if we take that into account*, our percentages go to something close to 65% – 45% – 12% – 1%. We can round those to 7-5-1-0. But it still isn’t correct–because we are only measuring the points generated by balls kept in play where there is a chance to direct the ball over the net. The system does not take into account the negative value of getting aced at all.

*Please remember, I don’t have computer stats or massive samples to hack at this with–the stats/math are being done by hand and Excel and the stats are a combination of old Big 10 stats along with juco numbers and a season of *trusted* HS stats. I don’t think this changes the principle involved, however.

So–while we have a 65% positive outcome on a 3-option pass (.65 points generated, rounded to 7, then multiplied by 10 so we have a straightforward number)–the 0-quality pass currently has no negative value…it simply lacks positive value, but we KNOW getting aced means an absolute loss of a point. (-1.00 point lost, rounded to -1, multiplied by 10 to equal -10).

This means instead of 3-2-1-0 or 5-4-1-0, to accurately evaluate a pass’ quality and its effects on our offense, we should change the scale to: 7-5-1-(-10).

So old system passers: A: 2-2-2-2-2-2 2.00 4-4-4-4-4-4 4.00 B: 3-3-3-3-0-0 2.00 5-5-5-5-0-0 3.33

Now instead we get: A: 5-5-5-5-5-5 5.00 B: 7-7-7-7-(-10)-(-10) 0.47

This makes the critical nature of avoiding getting aced even more obvious. So let’s look at a three juco libero as a freshman and sophomore. Did the passer get better?

#1: Freshman year: 2.60 on a 3 pt scale Sophomore yr: 2.58 –clearly not better…basically the same With the 7-5-1-(-10): Freshman: 5.53 Sophomore: 5.73

#2: Freshman year: 2.01 on a 3pt scale Sophomore Yr: 2.04 With the 7-5-1-(-10) Freshman: 3.81 Sophomore: 4.22

#3: Freshman year: 1.99 on a 3pt scale Sophomore Yr: 2.03 With the 7-5-1-(-10) Freshman: 3.71 Sophomore: 3.58

Number-crunching for several years of data at the juco level and 1 1/3 of high school, the new system clearly shows differences between athletes’ passing values. I’m convinced this is accurate and a better way of evaluating serve-receive passing. [1], [2], and [3]

[1] There IS a catch–ultimately, each team–at each distinct level–will have its own percentages for winning points with 3-option, 2-option, and 1-option passes. Me using 7-5-1 (6.5 and 4.5 to be technically correct) is based on info from NJCAA volleyball primarily…men’s volleyball is absolutely going to have different values, same with HS, etc, etc.

[2] While more accurate, I don’t think this is a tool that can be used in-game without computer assistance. To be useful in game, stats HAVE to be simple and easy to digest. 7-5-1-(-10) is NOT intuitive in that regard. Thus, it is better for teaching the importance of just keeping a ball alive or for coach evaluation after practices/matches. It’s useful there.

[3] The easiest way to do this is going to be to chart passes by setter options–the ol’ 3-2-1-0. That’s the easiest way for people to do stats (also, it’s the way computer apps insist you do it…). Afterwards, plug those 3/2/1/0 totals into a spreadsheet and let the spreadsheet do the math heavy lifting.