Printable View
Actually...maybe not? http://kenpom.com/blog/index.php/web...poll_is_great/Quote:
Originally Posted by steven52682
The preseason poll is actually a better predictor of results than the in-season one. The problem is that the in-season one overreacts to bad losses and good wins.
UConn definitely is a top 25 team. Are they really a top 10 team? Hard to say.
Ken Pom's Game plan show's Duke winning every game. But at the bottom it show's predicted record of 29-2 (14-2 in conference). So which 2 conference games is he predicting Duke loses? Am I reading his page correctly?Quote:
Originally Posted by JasonEvans
I have sworn off reading the polls until after January 1st.
Right below the projected record it says:Quote:
Originally Posted by bluepenguin
Kind of like the probability of flipping a fair coin and getting heads is 50% for each individual flip, but the probability of getting heads 3 times in a row is less than 50% (12.5%, in theory).Quote:
Projected record is based on cumulative probabilities of winning each game and may not equal the individual game predictions.
It is probabilistic: Duke is favored in every game (win probability > 50%) but is almost certain to lose some games. For example, if Duke has a 90% chance of winning each of 30 games, the expected number of losses would be three.Quote:
Originally Posted by bluepenguin
sagegrouse
I don't think that Pomeroy's analysis really applies to UConn because it only concerns the results and movements of teams ranked #1. His post does raise an interesting question, though, in what the polls are supposed to represent. Pomeroy has this to say.Quote:
Originally Posted by loran16
Now, you’re saying, “Ken, the ratings aren’t designed to identify the best teams! No pollster is doing that!” My response to you would be “Stop yelling.” And then - why aren’t they doing that? I also ask you – what are they trying to do? You probably can’t answer the last question, and I couldn’t either so I contacted the AP to determine what instructions are given to voters. They did not respond.
I believe that Ken's analysis is correct except that he is missing two big words at the end, right now. The Polls have always been a snapshot of the current landscape and not a predictor of future performance. The only one that is commonly used as a predictor is the preseason poll, which is probably why it is better at predicting things. Will UConn be a top 10 team come Selection Sunday, probably not. I'm not sure they will even be a top 25 team as I personally don't believe that they can simply have Kemba Walker carry them with little help for an entire season. Unless someone else steps up big time, UConn will lose every game against a half decent team when Walker has a sub-par performance and that is not a formula for consistent success. Right now, though, any rankings based on performance only should have UConn in the top ten.
Unfortunately for those of us who love the dork polls, they are all still influenced by preseason ratings, so we can't know for sure where UConn stands base solely on performance.
So how far should Minn. fall? They beat UNC who was ranked number 8 but yeaaa and now they lost UVA.
great stat. It bodes well.Quote:
Originally Posted by gam7