It has finally happened for the first time – All 5 models are in complete agreement for which teams are among the Top 10. Not that they have consensus for the exact placement of any one of them, however. Southern Virginia remains the landscape’s only unbeaten, and with three #1 ranks from the 5 models to Stevens’ two, continues to secure its place at the top of the MORE. That 3-2 margin mirrors the AVCA as 14 of its 23 first place votes on ballots went to the Knights, also. It has been a long time since the AVCA point leader in any week has secured less than two-thirds of the first places from the coaches! I guess that can happen when not only two of the three undefeated squads take their first loss, but the Knights also show a chink in the armor that to this point hadn’t been seen in the last 3 months, either.
It seems apparent the committee may have their hands full with Pool B’s third selection this year. Santa Cruz just overtook Nichols for the first time in the MORE all season. They now sit at #11 to the Bison’s #12. One has to wonder, “If the Banana Slugs can get by Cal-Lu later this week, would it be enough to wash the bad taste of that Buff State loss out of the mouth of the selection committee?” There is little doubt Santa Cruz has played the more difficult schedule even with the SOS metric exaggerating them both just a smidge. (The difference is inflated by roughly .01 as per yesterday’s post – Exactly .012 for those who want the exact number.) But their record suffered for it accordingly. I can’t help wondering if Springfield’s latest “burps” against New Paltz and Eastern Nazarene may end up creating a ripple effect to hurting Nichols’ chances, they having lost a couple close ones to them?
Is a selection supposed to be who is believed to have earned it over the course of the whole season, or who is believed to presently be the better team right now going into the NCAA Tournament? It is almost ironic a discussion is focused on the #11 & #12 teams fighting for a bid when we already know for a fact at least one of the best ten below is guaranteed to not be there, maybe even a few more if Juniata avenges their Southern Va. defeat, if Lasell can win the rubber match we expect them to have against Wentworth in 11 days, or should any other team win the UVC besides Vassar or Fisher? Imagine if Misericordia catches Stevens off guard to win the MAC at the same time Lasell is up on Wentworth and Juniata ahead of Southern Virginia late in the afternoon of April 13th? If it goes down like that then one of the best 3 teams in the land would not be Nationals bound. No worries though, because that is only a 1 in 200 chance to happen. Although, I remember a time about 15 years ago I was managing a large investment fund with 6 Vegas Favorite Sweet 16’s, and watched all 6 lose, and that was also a 1 in 200 event, too.
Here is the MORE:
I never heard back from the KPI creator. Sure wanted to know how his SOS metric consolidates wins and losses. It, like the NCAA’s, also is a biased statistic for pretty much all the same reasons. There is little doubt for it being a derivative of the whole “Opponent WIN% & Opponents of Opponent WIN% combinations” beloved by the NCAA’s RPI for many years. I suspect this to be the culprit when looking above to see Carthage, Augustana, and Dominican being 4, 6, & 8 KPI ranked spots below the median value. In fact, the table below shows the variance of each model’s rank of these 25 MORE teams relative to the medians used to rank them above. Notice how much higher the KPI is than the others.
| AVCA | MAS | IH | T100 | KPI |
| 4.76 | 4.44 | 7.96 | 3.88 | 9.80 |
Those who have been following all year might be wondering, “Why bring in the KPI if knowing its SOS is a biased statistic?” First, I didn’t know its SOS was biased until recently, but I did know the KPI seemed to treat the Midwest a little too harsh in its rating. Even though it created a little more noise overall, I was pleased that it made the MORE signal a little better by being a part of the process. However, in order to not let the KPI exert too much leverage on the system, I started to use medians rather than means in the AZR. Average is the common name for mean, but some refer to all measures of central tendency as averages even if it is technically not legit. To avoid any confusion with the “A” meaning average, I will be changing the name AZR to CZR starting next week. i.e. Central Z-Score Rating because I am using medians instead of means – I like the idea of calling it “Caesar.” LOL (As far as I’m concerned, it rules all metrics out there as a highly effective wisdom of the crowd composite of strength of D3 Volleyball teams.)
Oh, and one more thing about bringing in a new component of a team which creates a little more noise, but at the same time makes the signal it sends a little better. Have you ever seen a coach sub in an OH that is a little more wild and likely to error, but brings a passing component sure to improve the team’s overall chances nonetheless? That is kind of what the KPI is to the MORE.