The Landscape Across Rankings

Matchup Metrics Comments Off on The Landscape Across Rankings

A composite of all the experts!

Trust is not simply a matter of truthfulness, or even constancy. It is also a matter of amity and goodwill.

Gary Hamel

Coaches who vote in the AVCA Poll are the volleyball experts – No Doubt! There are 23 of them, representing all geographical regions (1,2,3 & 4), no more than 2 from any conference, all who take the time to punctually submit their ballots allowing fans the enjoyment to process their collective thoughts every Tuesday afternoon. This week, I am delighted to report, the sum from all 23 of them, each ranking 1 thru 20, finally added up 4,830 on the button. (23 x 210 per ballot = 4830). Their poll is an example of “Wisdom of the Expert Crowd.”

Massey, Inside Hitter, and T100 are based on models intended to use mathematics to do exactly the same as the coaches en masse. That is to reduce the characteristic noise about the signal that is strength of these volleyball teams, and let me tell you there is an awful lot of noise. Variability in play by the teams, variability in schedule strength, random factors related to officiating (We all know about that, just ask us!), and even luck can collectively produce the noise to which I refer. Every win and loss must be rated appropriately based on a multitude of factors because assessment by simply counting them up is terribly poor at indicating who is better than whom, in particular when two teams lack the opportunity to face off on the court. 

Just 3 weeks in, differences in the three math models to dampen their respective noise often creates more variability between them than to the coaches who observe their teams on the court. However, as the season goes on and the bins are filled with more data, these mathematical systems converge to far lesser differences from one another. I have no doubt the T100 will show little variability in its signal compared to the other two in April as the sample sizes approach thresholds providing heightened confidence, and by extension, more certainty for that which we claim. 

The human condition is subject to cognitive biases that logic and algorithms intended to process data aren’t. However, humans invented these, so their flaws are of a different kind. I likely have as much confidence in the T100 as those behind Inside Hitter have in theirs. I’m not sure I should have as much confidence as Massey, given I consider him one of 4 who is on the “Mount Rushmore” of sports statisticians, but it is hard to invest so much focused intent and not have faith, especially given the results. And by results, I do not mean forecasting winners and losers. A model is best measured when being able to demonstrate the 100 matches forecasted exactly at 80% in the end shows roughly 80 favorites won the contest, etc. The key isn’t to be correct in predicting the outcome as much as it is to be correct in the variability about it. Calibrating a model to demonstrate this is the mission, and the rest often takes care of itself.

I truly enjoy checking in on what others are cooking up by their latest ratings. The computer ratings tell a different story every week as to how wins and losses are regarded. And I always enjoy studying the points and rank order by the coaches, especially last year when every ballot was transparent, allowing to see the variability in the individual coaches for various teams and conferences, particularly their own. This year it isn’t available, and that’s not only okay, it is the way it should be. It isn’t as if these coaches are my representative in Congress, right? To whatever degree any doesn’t vote their conscience with anonymity, it is certainly bound to be less so than with having known in previous years anybody could lookup how they filled out their ballot.

So why the title “MORE AMITY?” The MORE stands for Median Order Rank of the Experts. In other words, look at the ranks of any team among the best 25 for every source and take note of each median. The 4 expert sources: AVCA, Massey, Inside Hitter, and T100, Yield this output. It is nothing more than a composite forming probably the best ranking of D3 Men’s Volleyball teams on the planet, if you believe in the Wisdom of the Crowd as much as I. That is the AMITY. It is also the good will toward all these experts which seems apparent having read the above, at least I hope so. So with no more explanation here it is:

MORE AMITY – Median Order Rank by Experts – AVCA/Massey/Inside Hitter/T100 – Yield
As of 1/30/24AVCAMASIHT100MEDIANMEANSTDEV
1Wentworth16211.52.52.4
2Stevens24423.03.01.2
3Southern Virginia35133.03.01.6
4Cal Lutheran41865.04.83.0
5Vassar63945.05.52.6
6St. John Fisher58376.05.82.2
7Loras82586.55.82.9
8Springfield77757.06.51.0
9Nazareth101061110.09.42.2
10Carthage111114911.011.32.1
11North Central [IL]912151011.011.52.6
12Marymount189101412.012.84.1
13Juniata1213161212.513.31.9
14NYU1314111313.012.81.3
15Dominican (IL)1415171515.015.31.3
16Nichols1917121616.516.02.9
17Misericordia1522131917.017.34.0
18Benedictine (IL)1716211717.017.82.2
19Messiah1619221818.518.82.5
20MIT2325202222.522.52.1
21Mount Union2620192623.022.83.8
22SUNY New Paltz2121262523.023.32.6
23Baldwin-Wallace2026262123.523.33.2
24Lasell2226242323.523.81.7
25Hobart2624182625.023.53.8
26UC Santa Cruz2623262425.024.81.5
27Rutgers-Newark2618262626.024.04.0
28Randolph Macon2626262026.024.53.0
29Augustana (IL)2626232626.025.31.5
30St. Joe’s LI2426262626.025.51.0
31Hunter2626252626.025.80.5
A 26 rank seen above is a place holder for any true rank in that system which is not among the top 25.

To see a January graph of the top 20 AVCA Polls showing team movement throughout the month, then you can simply click HERE.