There are a couple of extra graphics presented after the updated “MORE 25,” that readers may enjoy seeing. As for the MORE 25, there remain 4 Gold, 7 Silver, and 8 Bronze tier teams in the latest accounting. The AVCA Coaches Poll, out just an hour ago, shows 459 of the 460 total votes by its 23 members were distributed across just 22 of the 124 programs from the landscape. This means just 1 coach from the 23 voters placed a single team, one time, that was not from among these 22 teams, and it would have been in the 20th and final slot of his/her ballot. I often wonder exactly what drives such uncanny and remarkable agreement regarding “independent” opinions for the strongest 20 teams in the country? It makes perfectly good sense to see this level of agreement for the best 16 to 18 teams, but when it comes to choosing the last 3 teams on a ballot, for example, this coalition almost unanimously agree they should come from just 5 remaining candidates? If I took the last 10 teams from the MORE 25+ below not among the top 17: MIT, Dominican, Misericordia, Stevenson, Benedictine, Messiah, Nazareth, Augustana, Lasell, and Aurora, and then randomly put their school’s names on a piece of paper and gave it to any 23 coaches across the landscape, asking each to circle the best three from these 10, I would bet a Benjamin that more than 5 of these teams would be circled at least once! Otherwise, that would constitute unanimity for the opinion half of these 10 all being better than the other half. Exactly as this latest poll implies happened. Even computer models don’t conform to that degree of certainty after 80% of a season is completed? I might lose that Benjamin, but I’d be significantly better than 50-50 not to, I think. Speaking of 50-50, it occurs to me that it only requires 23 people to be in a room to have a 50-50 chance for any two having the same birthday (month & day). For the heck of it, how about the 23 voters put their birthday Month and Day on the next AVCA ballot to see if any two of them do have the same birthday? That might be a fun experiment! The chances any two of them do will be equivalent to a coin flip. I could double down with that wager just to make it sporting. LOL
Earlier this season I described how computer models would converge to more common ground as the season played out. At the time I was graphically comparing them to each other, which required 3 different scatter-plots, but since have produced the AZR (Average Z Score Rating) as a reasonable benchmark for the signal of team strength. So now the relationship can be seen by a single plot with its data points color coded by the model from which they come. More than three-quarters of the way through the regular season, you can see how the data from each form their own model and how all 3 of them converge to a reasonable wisdom of the crowd signal, too.
The matrix seen below is a win matrix of all D3 Men’s Volleyball Teams wins through last Thursday, 3/14. The 124 teams are placed in the order of their T100 rank from top to bottom and from left to right. Every black cell represents at least one win where the team on the left margin defeated the team on the top (or bottom) margin. Some of the cells have 2’s inside them if defeating an opponent twice, for example. Those that are white are simply zero, meaning they haven’t played.
The reason it was shrunk to the size you see is that 124 x 124 = 15,376 cells on this matrix. You can see what amounts to a density of more than 90% of these black cells above the main diagonal, which means the team presently rated stronger on the latest T100 won a match played over these last 10 weeks, not that the forecast in real time was accurate in predicting winners better than 90%. (Real-time forecasts this year are closer to 85%, as is expected given the nature of D3 Men’s volleyball and the scheduling practices for its teams.) The black cells to the lower left of the diagonal are what some might call “upsets.” They were unexpected outcomes at the time which still remain so as a consequence of any team’s overall performance throughout the year. They represent the “signature wins” of teams rated weaker and what will be considered “bad losses” for teams to end up rated stronger at the season’s end when the NCAA committee gets together to discuss worthy resumes for Pool B & C selection.
What I thought would be a fun look at some of these surprising outcomes is to create what I have called Transitive Trivia Tales – A game of four degrees of separation using the transitive property in mathematics. I have color coded them and enhanced the trek on the matrix below in what a math guy might refer to as iteration. The length of any “UP ARROW” is the magnitude of a present difference in T100 ranks of the opponents which describes the degree of the somewhat surprising outcome. The longest up arrows in each tale are: ORANGE- Buff State over Santa Cruz; RED- Arcadia over Cal-Lu; GREEN- Baruch over NYU. Is it really that surprising that 2 of these 3 are excellent West Coast teams coming East to play multiple teams on consecutive days?
NOTE: For anyone wondering about the blue outlined squares along the diagonal, these are the conferences when I sort the order of teams first by conference and then by T100 rank. When the regular season is almost finished, the visual of this matrix as seen through a conference lens will be available in “Conference Corner – Cuatro” the end of next week.