Gators In KenPom & Net [Latest at #133]

It’s a good question. I’m not sure if “game” is the right word, but teams definitely value metrics more today. I guess it would be “gaming” the metrics if teams stopped putting in their subs after the game was out of hand in order to maintain a high margin of victory.

It should also be noted that KenPom’s metric is a relative one, where 0 is always the rating of the average team, so it’s impossible for all teams to raise their rating. Every time the Gators raise their KenPom rating, it necessarily comes at the expense of their opponent’s rating. So if these high ratings are coming from intentional coaching actions, it must be that the losing teams aren’t valuing the metrics to the same extent as the winning teams, otherwise their actions would just counteract those of the victors.

 

That leads me to believe they've tweaked what they use in the metrics.
Or, maybe the Powers-that-Be play with the formula to make sure that Duke remains at the top, for the TV ratings.

 

What strikes me is that all 4 of those teams—Auburn, Duke, Houston, and Florida—are really different and win in pretty different ways. Different tempos, different offensive and defensive philosophies, different competition levels (even contrasting our non conference with Auburn’s). But they are all easily top 4 in various configurations across all the advanced sites (KP, Torvik, Miyawaka, NET).

Part of me thinks we just know better how to get a statistical edge in games that actually contributes to wins (even if we all go about it differently), such as the four factors. I mean, if you rebound better, shoot better from 3 and the paint, get to the line more, and turn your opponents over more, you are going to be pretty good, haha. So is that what accounts for so easily beating most bad teams, so handily beating good ones, and only losing to really good ones? With the exception of Duke, it’s also interesting that those teams do not have a high volume of elite prep or NBA draft prospects on their rosters.

One other possibility is that portal/NIL has enabled programs to evaluate kids better on the basis of actual college production that will translate to the new system/conference and also to put together more experienced teams, which tend not to fluctuate from game to game alike younger kids do.

Will be fun to watch down the stretch.

 

If there's a statistical edge that can be exploited, CTG will know it.

 

The nugget in Eric’s story about memorizing license plates was revealing.

 

I still think one of the most impressive things I’ve heard to date was him tracking the plus/minus for his teammates in his head, while he was playing. That’s not a simple task to track live even if that’s your only job during the game. Tracking multiple teammates live, in your head, while playing, that’s next level stuff.

 

I'm SMH about how Duke has a relatively high SOSNetRating when they play in the ACC? I know Duke played Kentucky and Auburn OOC, but those teams in the ACC are trash...

 

Related: According to bracket matrix, the Gators are now projected as a #1 seed in all 87 brackets that they track. This puts their bracket average slightly higher than Alabama’s who is projected as a 2 seed in seven of the brackets.

The Bracket Project's Bracket Matrix - 2025

 

A big problem with NET ratings is that not all Quad 1 games are the same. Ditto for Quad 2 etc.

Duke has a very good Quad 1 victory when they beat Auburn on their own floor. That ranks just the same as Florida beating Auburn on Auburn's home floor. According to NET they count the same. LOL.

 

No. The impressive part is doing that and still be able to operate shoelaces, an automobile, a toothbrush, and a girl.

 

Yeah, there's an obvious flaw in this whole "Quad" game system. The break points are simply placed where they are for no good reason. A true road game against a #75 opponent is a Quad 1, whereas against a #76 team is a Quad 2. Jeebus. How much difference can there be between teams ranked 75th and 76th? Don't say "1", wise guys.
There's a much better math solution to this.

 

Well said, Planet. This is a problem with all systems applying categorical cutoffs to continuous data. E.g. the same problem exists with letter grades in school, where 81% and 89% are both considered Bs, and 79% and 90% are considered to be a C and an A, respectively.

One thing that I think the quad system gets right is recognizing the difficulty of winning on the road. I’m not sure that a home win over the #25 team is exactly equivalent to a road win over the #75 team, but it sure isn’t equivalent to a road win over the #25 team. The best of our imperfect options should take into account both your reasoned criticism of arbitrary cutoffs and the very real phenomenon of home court advantage. Somehow.

 

We laid an egg and scrambled it last night but we're still #4 in Kenpom, Torvik, and Evanmiya this morning. That's some good news, at least.