Does anyone know what the “Luck Rating” metric is on Kenpom? How is it measured?
Luck – A measure of the deviation between a team’s actual winning percentage and what one would expect from its game-by-game efficiencies. It’s a Dean Oliver invention. Essentially, a team involved in a lot of close games should not win (or lose) all of them. Those that do will be viewed as lucky (or unlucky).
what a stupid metric, you can’t measure something like that. what if it’s not luck and it’s actually a team being clutch down the stretch?!
So, Michigan won over us last year was luck because I think so.
It cracks me up that he would try to pass of a “luck rating” to make it sound like his formula is accurate and any deviation would be explained by luck instead of things like, coaching decisions, players being better prepared from one game to another, match ups (team A matches up well with B which matches up well with C which matches up well with A), refs “letting players play” vs “blowing the whistle at every contact,” player suspensions, injuries, some player progressing through the season and others regressing and an infinite of other things that are impossible to detail in any computer rating. It would be more genuine to call it an “Error Rating” but what it really is, is a “Hindsight Rating” that you can apply to past performances and is impossible to apply to future performances. Predicting winners and losers can be done with a very high mathematical efficiency. Trying to predict point spreads with an efficiency higher than a basic calculation and a roll of the dice to add or subtract points is a fools errand.
I think you’re hung up on the word “luck”
More likely this metric measures teams that win games that stats alone wouldn’t get, including but not limited to the things you mentioned.
All it means is that in games that metrics show it could go either way this team is either winning a higher or lower percentage than the formula shows, so you can add in that to try to overcome the 1’s and 0’s of efficiency and power rankings.
I have less of an issue with the “Luck Rating” metric than I do with the “Strength of Schedule” Metric. It seems like the SOS Metric seems to really value brand and conference over actual strength. I’ve watched too much basketball to believe that the current 11-15 should be ranked ahead of the Coogs strictly by the numbers. I just don’t see it.