Ok, and I don’t entirely disagree, but what adverse impact is this having on the outcome of the season, or the game more generally?
1 Like
I think like other preseason and early polls, it shapes a lot of the discussion around who is good and that has a downscale effect on what is already a subjective process of determining who gets into the tourney. It’s not the end of the world but i think it would make sense to wait to release rankings when there is enough real time data.
Personally, I don’t think UVA’s rise toward the top of the rankings early last season helped much with seeding in the NCAA tournament.
4 Likes
The thing is, there is no right or existentially just way to do it. Preseason bias is going to be present no matter what, with or without computer rankings. In theory, by the end of the season, when everybody has played or is connected to everybody else, a computer ranking is arguably the best way to account for all the combinations of who beat who and by how much. Of course, then we’ll be debating how to weight the early season/non-con vs in conference road games, etc., but hey that’s why there’s a tournament at the end.
2 Likes
I mean there may not be a right way but there is a better one. Preseason bias does exist but giving it a seemingly objective metric that turns out not to be increasingly irrelevant is not the way to go.
But the preseason bias is decreasingly relevant as the season goes on, is it not? So to the extent kenpom, etc have an impact on tournament selection, the impact of pre season bias should largely have faded from the models and the rankings should just reflect actual results, for the most part.
2 Likes
@Marktykobe reading all this talk of Kenpom
9 Likes
Sure but it still plays a role. Feels like an easer fix just to remove what is just irrelevant metric that serves to support preseason bias than to hope it goes away.
I’m guessing we’d all agree that all 363 D1 teams are not equally good even when everyone is 0-0 at the start of the season. I think most everyone was more excited about beating Florida than thumping NC A&T. So we all rely on some prior knowledge in addition to what has happened in the current season to form judgments about a team’s strength, and we all pay attention to various teams’ relative strengths as part of understanding our own teams. The best way to do that is certainly up for discussion, but I don’t see a way of getting out of paying attention to some form of prior information about a team.
3 Likes
Agree, plus the NET is the only official NCAA ranking (technically), and they wait to drop that at least until we have some data.
Just looked it up – last year was Dec. 5, which means post-Feast Week and “challenge” weeks. At least a chunk of the noncon.
Plus (and trigger warning for all you stats haters who don’t like numbers b/c your 5th grade math teacher yelled at you) – the issue with any small sample size you grab to try to represent a larger set is twofold: (1) normal variation, and (2) “something changed.”
For me, this is basically the Vegas tourney question – did we look super good b/c of normal variation or because something changed (injuries, other teams caught up, etc.) ?
If you all want, I’d be willing to explain why I think box and whisker plots are the best way to view a team’s adjusted offensive and defensive efficiencies.
5 Likes
Only if you form it into a question for the J-Willy show. Basically, write it all out, and at the end, say “Thoughts?” 
2 Likes
We’ve got him — it’s Leon Bond. And he’s going to be a very good tweener, just like the guy referenced above.
Difference is, one of them wanted to be here. One of them didn’t. Love that kid who played at Benedictine and Trinity, but once a Dukie, always a Dukie, and you’re dead to me.
1 Like
I think this is slightly misleading. I’ve always gotten the sense that a big reason why he isn’t here is because Tony wasn’t that high on him.
1 Like
Same sense. Pretty sure he would have been a Hoo if CTB had pushed. Point is still valid, we need a glass cleaning PF…pretty sure the that’s what the staff hoped for Minor…
2 Likes
We’ll never know because he never tried to commit. He was waiting on UNC and Duke offers. Once the Duke one came, we stopped recruiting him.
2 Likes
Problem is there is never sufficient data in college basketball. The season isn’t long enough for the number of teams. Thus they try to compensate with historical program ratings, similar player projections, etc. tortvik has talked about it before but if you remove the weighting the algorithm becomes way less accurate at predicting things like conference champions and the tournament. There is no ‘accurate’ way to do rankings of a 30 game season with 360+ teams competing. Too many teams don’t even play teams from the same conferences let alone have common opponents
I’m pretty sure once we got Hauser and McKoy that’s when we backed off.
Maybe I’m wrong though. Not that it matters anyway. Wanted him to lose at Duke. And I’ll want him to lose when he comes to JPJ.
2 Likes
Henry I think would’ve come here if the positional timing had been different. We were desperate for players after the 2019 post-natty exodus and we went out and got Justin McKoy and Sam Hauser, both 4’s, and all of a sudden 2020 Henry Coleman didn’t have an obvious scholarship slot or maybe minutes.
3 Likes