PennStateHoops.com Discussion Forum

Kenpom's new "similarity scores" feature is a great time waster


#1

You can read a little bit about it here ( The kenpom.com blog ), but basically any score over 900 means it’s a pretty good match. Start going through the Penn State roster to see some fun comparisons. ;D

kenpom.com subscription.


#2

As someone who beat kenpom by 21 games(14 straight in late season as I got better at it) last year(out of 100-150 games) in the eyeballs vs. geeks contest, I’d like to comment a bit on his work. Someone once asked here if his work could be used to wager on games, and it was one reason, maybe the best reason on why I played that contest(the second being showing up Lar. I think I have them in the right order ;)). On that, it’s inconclusive, or at least I didn’t do the work necessary to know. For example, I might take a game vs. kenpom, where he had a team winning 67-66. Bodog might have it the other way, the favorite a 2 point favorite, and my pick would win by 1 point. So, was kenpom wrong? Not really, as if you bet his team, taking the 2 points, you would have won. Unfortunately, I did not keep those stats. But, my sense was that if he was at odds vs. the bookmakers, you should take the bookmakers choice, as they seemed to cover on most of the closer contests. Did kenpom make the odds closer than they should have been? If people used his work to wager, maybe. I don’t know. That said, some spreads varied widely, and on these instances, where both the bookmakers and kenpom were on the same side as to the favorite, kenpom seemed to get the score more accurately. One glaring example was PSU vs. Purdue, the game where the out of bounds play was called wrong, in the last minute of play. Bodog had Purdue by 10-11 points, kempom by 6. While anecdotal, IMO, these discrepancies bear watching closely. kenpom may help in contests like this. I think this may have to do with a game not heavily bet, where smart money, or big money, stays away. Without a lot of action, maybe kenpom’s predictions are more useful, in that the spread is not right due to lack of betting action. There could be opportunity in these wide spread games. BTW, I hear bodog may have to move offshore, and like fulltilt, maybe money left there could be at risk if betting online. Caveat emptor.

It seemed easier to beat kenpom late in the season. Maybe his work doesn’t emphasize more recent play, or looks at overall stats rather than specific matchups. He had NU over PSU at Evanston. Dumb pick, IMO. That was a PSU lock.


#3

Eyeballs for me :wink:

BUT, if you can’t eyeball I would rather review stats and perform my own analizing…with a large grain of salt. I’m not a “program stat guy” at all!!!


#4

When you call this a “great time waster”, are you saying it’s fun to waste time messing around with it? Or that it’s a waste of time, so don’t bother?


#5

It’s amazing to me that for all Kid has to say about kenpom’s work he has yet to take the 3 minutes necessary to actually figure out what all his numbers mean and how he “comes up with” them. Although it’s not surprising.

When you call this a "great time waster", are you saying it's fun to waste time messing around with it? Or that it's a waste of time, so don't bother?

I think you’re confusing two completely different terms here. A time waster is something you do to pass the time, a waste of time is something that isn’t worth doing. Clearly I wouldn’t be sharing something with you I didn’t think was worth doing.


#6

[quote=“tundra, post:3, topic:2642”]Eyeballs for me :wink:

BUT, if you can’t eyeball I would rather review stats and perform my own analizing…with a large grain of salt. I’m not a “program stat guy” at all!!![/quote]

Freudian? I hope. Sounds like it would hurt. ;D


#7

I think you’re confusing two completely different terms here. A time waster is something you do to pass the time, a waste of time is something that isn’t worth doing. Clearly I wouldn’t be sharing something with you I didn’t think was worth doing.[/quote]

I didn’t have it confused. I just wanted to make you weren’t confusing the terms - I didn’t know where you stood in the great kenpom debate. And reading kids post right after yours probably didn’t help.


#8
Eyeballs for me ;)

BUT, if you can’t eyeball I would rather review stats and perform my own analizing…with a large grain of salt. I’m not a “program stat guy” at all!!!

Freudian? I hope. Sounds like it would hurt. ;D

Ouch! :smiley: