Fading KenPom College Hoops System

[onthewhat]

Many of you probably use his site, KenPom.com, very good site with lots of stats and numbers, and he also has a model where he predicts games.

The lines on most of the games falls within a few points of the KenPom number. The line on a handful are 5 or more points "off", and so it would seem the easy play there would be to follow his model and play the "off" number, but not so fast, the bookmakers know very much more than Mr. Pomeroy, so we are FADING his selections where his number is a reasonable amount different than the bookmakers number.

For example, today KenPom had Old Dominion -12 @ George Mason. Bookmakers opened the number at George Mason by 3, but it got bet to Old Dominion -6. Easy play there is to play Old Dominion +3, since it's 12 points better than Pomeroy's model. But wait, the Bookmakers are pretty smart and George Mason wins easily.

There are many more examples, but that's the basis of it. Many people check his model. The Bookmakers know about it. If they hang a number way different than this model, but 90% of the games have lines within a point or two of the model, then the bookmakers know something. That's what I'm banking on, anyways.

Will post Sunday's irregularities later, then we will decide which ones are plays.

[SamsNCharge99]

never heard of it, but good luck

[JsmooveFTW]

Yeah, go for it. You and Patty. I love that site though. It and NBA stuffer are the best. Ask my firefox.

[JsmooveFTW]

FYI, last year that realtimerpi site's prediction had some tail-worthy totals predictions. Stick to your systems

[Crash]

yea me either

[durito]

Many of you probably use his site, KenPom.com, very good site with lots of stats and numbers, and he also has a model where he predicts games.

The lines on most of the games falls within a few points of the KenPom number. The line on a handful are 5 or more points "off", and so it would seem the easy play there would be to follow his model and play the "off" number, but not so fast, the bookmakers know very much more than Mr. Pomeroy, so we are FADING his selections where his number is a reasonable amount different than the bookmakers number.

For example, today KenPom had Old Dominion -12 @ George Mason. Bookmakers opened the number at George Mason by 3, but it got bet to Old Dominion -6. Easy play there is to play Old Dominion +3, since it's 12 points better than Pomeroy's model. But wait, the Bookmakers are pretty smart and George Mason wins easily.

There are many more examples, but that's the basis of it. Many people check his model. The Bookmakers know about it. If they hang a number way different than this model, but 90% of the games have lines within a point or two of the model, then the bookmakers know something. That's what I'm banking on, anyways.

Will post Sunday's irregularities later, then we will decide which ones are plays.

lol, he's a good first approx, but his model isn't designed for betting, you have to take it further... fading it isn't gonna do shit though

never heard of it, but good luck

lol

[pokernut9999]

Could you say 1 game is a small sample size ?

[hhsilver]

have you used his site in other years. Have you found that the line usually moves toward his number when the difference is more than 1 or 2?

[onthewhat]

lol, he's a good first approx, but his model isn't designed for betting, you have to take it further... fading it isn't gonna do shit though

ok...but how come the majority of the games are less than 2 points difference from his model and the lines?

Could you say 1 game is a small sample size ?

i was giving an example...

[Masu485]

i think every stat model in existence had old Dominion as a large fav and deviated immensely from the vegas lines.

[onthewhat]

Sunday January 3rd

KenPom Line: Miami FL -17 @ Pepperdine
Betting Line: Miami FL -12.5

KenPom Line: Ohio State -5 @ Michigan
Betting Line: Michigan -1

KenPom Line: Wisconsin -7 @ Penn State
Betting Line: Wisconsin -4.5

KenPom Line: Iowa State -11 vs. Houston
Betting Line: Iowa State -5.5

KenPom Line: St Johns -8 vs Providence
Betting Line: St Johns -4.5

KenPom Line: Creighton -1 @ Evansville
Betting Line: Creighton -4.5

[onthewhat]

How many points do you all think we need for the lines to differ by to be a play? I'm pretty comfortable having Houston +5.5 and Pepperdine +12.5 as plays for sure. Then maybe Michigan -1.

Providence +4.5, and Creighton -4.5 are only 3.5 points off, so unlikely to list those as plays. Don't even know why I'm putting Wisconsin/Penn State up there, the line is only 2.5 points off. Pretty much everything else is exactly dead on for tomorrow's card.

[onthewhat]

Official Plays:

Pepperdine +12.5 -110
Michigan -1 -110
Creighton -4.5 -110

Waiting on Houston as I think line may move in our favor. Putting Creighton on there because they are a decent sized road favorite and the predictor only had them winning by 1.

[hhsilver]

can you answer my question? thanks

[onthewhat]

have you used his site in other years. Have you found that the line usually moves toward his number when the difference is more than 1 or 2?

really haven't paid that much attention in terms of line moves. I have looked at the site in past years though.

[pavyracer]

Are you down to the last 2 dimes from the 25K unsecured loan onner?

[onthewhat]

Paver Strike 1

[jjgold]

Good stuff Herman

that is difference with a pro like you and us action guys. I see how you look for edges with different sportsbetting models.

[onthewhat]

Official Plays:

Pepperdine +12.5 -110
Michigan -1 -110
Creighton -4.5 -110

Waiting on Houston as I think line may move in our favor. Putting Creighton on there because they are a decent sized road favorite and the predictor only had them winning by 1.

Houston moved our way. Adding Houston +6.5 -110

[jellobiafra]

The biggest discrepancies from the site I use are Iowa St, Ohio St, Evansville, and Duke. I decided to play ISU, OSU, and Duke. Probably a fairly similar model. I've been very happy with it.

http://accstats.blogspot.com/

[onthewhat]

So you're tailing those public models? What's your record doing so in pretty noticeable discrepancies?

I don't think tailing KenPom in discrepancies is profitable, so we will see if fading him is.

[onthewhat]

[jgray]

I looked at this before a few years ago and found that home dogs getting 4 or more extra points has been marginally profitable. I haven't looked at that angle in a while though.

[jellobiafra]

So you're tailing those public models? What's your record doing so in pretty noticeable discrepancies?

I don't think tailing KenPom in discrepancies is profitable, so we will see if fading him is.

I've been looking for certain spots where the discrepancy makes sense to me. So it's part gut and then I use that to to support my choice, or actually more vise versa...if that makes sense. I pick about 2-4 a day that I like off of that. I don't play every substantial discrepancy, but every play I make has at least a 4 or 5 point discrepancy.

I don't know if you'd call the one I linked a "public" model. It's a friend of mine from another board (non-wagering). There probably aren't more than a 100 or so people who are even aware of it. I personally think it's better then KenPom.

[onthewhat]

KenPom is obviously way more popular than the one you posted. KenPom ranks in the 179,000's while your site is in the 9 millions in terms of traffic ranking.

The reason I'm fading the KenPom discrepancies is because so many of the lines match, while a few are quite a bit off, and the model is so public so it makes sense to fade the public in those situations.

[jellobiafra]

Well one thing that can be said for these models is it eliminates the need to waste a bunch of time "handicapping". No matter how much you crunch numbers or how many different things you factor into your number crunching, you're not going to come up with anything more intricate or accurate than what these guys have put into these models.

In the end it comes down to whether or not you trust the predictions ...be they your own pencil and paper bullshit or these sophisticated models. So if you are generally a gut player such as myself, these are just tools to use as a justification for your plays.

[MrMonkey]

Going with you on Mich -2. Relying on stats and gut! Ohio St hot shooting and Mich cold shooting shouldn't continue? Harris and Sims have to tack the reins. Beilein has to get them to D up! OhioSt has won 8 of last 9. Home team hoping gets it down! Good luck to all!

[onthewhat]

Houston/Iowa State just went to 2OT. I have +6.5, hopefully it holds up.

Creighton is up 11 in the 2nd half. I have -4.5

[onthewhat]

kok suckers lost by 7 in 2OT, unreal. Had the right side, shit happens.

0-1, -1.10

[onthewhat]

Creighton covers.

1-1, -0.10

[Chi_archie]

while you are at it look into the totals...... add the predicted score up and see how if there is anything there....


you also may want to check out http://www.realtimerpi.com/

I use Pommer, real time, and Sagarin as a guide, to Strength of Schedule and to just compare quality wins and losses of teams on a game before I bet them.

I don't necesarily go with any predictions, I take them as an opinion but I like the different stats like defense ratings, tempo ect.

[onthewhat]

I am just going to stick to KenPom's predictor

Michigan covers

2-1, +0.90

Pepperdine tonight

[Chi_archie]

keep it goin

[onthewhat]

Going to for rest of season

[onthewhat]

Monday January 4th

KenPom Line: VA Commonwealth -10 vs. Northeastern
Betting Line: VA Commonwealth -6

KenPom Line: Cincinnati -1 vs. Pittsburgh
Betting Line: Cincinnati -4

KenPom Line: Old Dominion -18 @ Towson
Betting Line: Old Dominion -12

KenPom Line: George Mason -2 vs. Hofstra
Betting Line: George Mason -5.5

KenPom Line: James Madison -6 vs. Deleware
Betting Line: James Madison -10

KenPom Line: Harvard -15 @ Santa Clara
Betting Line: Harvard -6

KenPom Line: Nevada -8 @ New Mexico St.
Betting Line: Nevada -4

KenPom Line: Davidson -8 vs. Samford
Betting Line: Davidson -11.5

KenPom Line: Murray State -30 vs. Tennessee Martin
Betting Line: Murray State -22

KenPom Line: Morehead State -10 vs. Tennessee State
Betting Line: Morehead State -14