The unofficial online home for fans of the University of Cincinnati Bearcats. Part of Collegiate Sports Nation (CSN).

I know this will sound odd to you but Kenpom isn't a tried and true religion. What eyes see still matters.

(Yesterday 11:35 PM)rath v2.0 Wrote:  I know this will sound odd to you but Kenpom isn't a tried and true religion. What eyes see still matters.

I know it might sound odd to you but I watch a crap ton of basketball.

I've probably watched more Houston than any team outside of uc. My eyes are why I am sold on them. They have so many guys. They do everything pretty well. They finally have a lottery pick type dude.. that's an elite team.

(Yesterday 08:46 PM)Bear Catlett Wrote:  Normally I would laugh at O$U losing again to fall to 11-11... if only they didn't beat our a$$ by 28 earlier in the season.

Oh, I can still laugh. But I'm also tired of recent basketball losses to OSU, dating back to the home and home losses we saw under Cronin.

The reason Kenpom and other efficiency metrics are useful is that over time, team records tend to converge to their efficiency. Kenpom is more useful than W/L records by far for that matter. The idea being that games are affected by variables out of a team's control like road games (worth ~1.0-2.5 ppg in a given year) and luck (the way refs call the game, the way shots fall in an arena, etc). So if a team plays 5 games that are all within 1 possession, that team could be 0-5 in those games, 2-3, 3-2, or 5-0 and all be similar in caliber, given that a team might get a 2 or 3 point advantage due to the factors mentioned.

Obviously, over time those game-to-game variances converge and you get a better picture of a team and their record. But in a 5 game stretch, the massive variances in college basketball could render some randomness to how the records play out.