Reply to thread

Message: [QUOTE="jncuse, post: 4889114, member: 1969"] I'm not going to claim his preseason rankings are flawless or even particularly great- in this current environment its hella difficult to rank 362 teams. You need to use some algorithm [U][I]But their impact does 100% go away by season end, even if there is some weight now. [/I][/U] Syracuse preseason ranking right now has minimal impact on where it currently is. Its not "important" to us right now. (It is important to some teams - See Wyoming example below) [B]Let's make this 100% clear though since there is some confusion. The preseason predictions have ZERO weight on the final KP rankings. KP did not claim otherwise in the interview. [/B] Its not a self-reinforcing system. Just as KenPom said there is preseason weight at the 5 and even 10 game mark (he is an open book), he will 100% tell you there is none by season end. [B] Your misinterpretation that he is saying that in this interview is leading you to make an incorrect downstream conclusion. [/B] KP could actually run his rankings right now with no preseason weight at all. He chooses not to (See my Wyoming below) [B]Here is an article from 2012 where KenPom addresses why he still uses pre-season weights around this time of the year. He is actually running rankings without pre-season weight at this point as there is now enough connectivity, [/B]but feels that teams that are playing well above his pre-season ranking or well below it, will tend to have some partial regression (or partial improvement depending on what direction the variance is). It's a long read, but he supports his position with data to keep some weights this late into the year. Seems like around December 24, 2011 he had the weights at about 20%(predicted) and 80% (Year to Date). Whether he should still do that on December 20th, 2023 might be more debatable where pre-season predictions are even harder and more in-accurate now then before. I would drop them as soon as possible. But I assume he still runs some data to support it. But rest assured those weights do go away before the year is over. [URL unfurl="true"]https://kenpom.com/blog/preseason-ratings-why-weight/[/URL] I will use Wyoming from that article. Wyoming was the most wild example from that year. Wyoming 2012 Prediction - 273 Wyoming KP rating at Dec 24, 2011 (with no preseason weights) - #43 Wyoming[U][I] published[/I][/U] ranking at Dec 24, 2011 (with preseason weight) - #88 Wyoming Final KP ranking - #98 He makes it very clear that the FINAL ranking has no weights in it. Regarding Syracuse since our current ranking (#91) is so close to our pre-season ranking of #105, the use of pre-season weights (say it still had 20% value) will have very little impact on our ranking. It will only move teams 10-20 spots if they are way, way above or way, way below their starting point. [/QUOTE]

Verification: What is a Syracuse fan's favorite color?