Vegas vs KenPom Week 10

flyer its asking for permission to view,If you click on access to anyone with the link it should open

R = Road; H = Home; F = Favorite; D = Dog;

A = Line moving away from KP; T = Line moving toward KP 

I wanted to go back to something I brought up a few days ago about the same teams showing up under the system and in those cases where they are getting smacked again and again. The conclusion that I've come up with is these teams have no depth to compete for the full game against the better teams in their conference hence the failure to cover even though they may have covered the 1H. E. Carolina, Rice, SMU, Towson, W&M, Citadel keep showing up as plays and they are killing us. How to filter out the crap teams that just can't compete even though Vegas gives them the benefit of the doubt?  My answer is to use the KP Thrill  Score.  Any game that is below 30 is eliminated from play. That prevents us from playing a game between 2 crap teams where we shouldn't be laying or taking points and prevents us from playing a team that just can't compete on a regular basis. They will have their flashes of brilliance but I'm not interested in being a hero by picking these teams. An added benefit is reducing the number of plays. If you're playing a decent amount of money on each of these plays you know what 23 games on the same day feels like. I've run the numbers for the last 14 days and I'm a little burnt out but here's what I've found. If anyone wants to do other weeks with this filter and report back, great, but I've seen enough.

Last 14 days starting with games of Thurs. 1/12:

Regular plays per the thread 58-56 with the last week being an awful 22-34

Filtered plays with 30+ Thrill Score  40-21 with the last week being 17-10

If you really wanted to get cute and reverse the plays for the thrill scores under 30..... 75-39  GL

Wow...didn't realize the system was making that many plays from those same teams...

29 out of 56 in the last week were games filtered out...

That's a lot of non thrill games the system is wagering on...lol

Definitely something to keep an eye on for future games...

that's good stuff kirby...

Kelly Criterion Method
Last Night: 5-7-1* -$167.80
Since 1/13/2012: 51% or 47-45-1
Running Total: -$87.18

Week 10 Inputs updated for 1/26
Starting bankroll: 500.00 - 87.18 = $412.82
Home, Favorite, Away (HFA)56.3% or 34
Home, Dog, Away (HDA)55.7% or 29
Road, Fav, Away (RFA) 56.7 % or 37
Road, Dog, Away (RDA)55.9% or 32
Home, Fav, Towards (HFT) 53.3% or 8
Home, Dog, Towards (HDT) 53.2% or 7
Road, Fav, Towards (RFT) 53.2% or 7
Road, Dog, Towards RDT) 53.1% or 6

*Note, I got Rhode Island +6 for the push. 

KenPom fan match.

I updated this last night it has been 5-0 since I started posting the spreadsheet.

 https://docs.google.com/spreadsheet/ccc?key=0AjP1VdOEnMobdHVmdWNZaWY1TkVjaFZhZEMtZ2hqd1E&hl=en_US#gid=2

flyerlax on your spread sheet is the prediction score the fan match score?  Also what is the "spots" column. It's the first I have seen that mentioned.

Also the "thrill" how would you recognize it if the prediction score is the fan match.

This is the out of the box thinking and in depth stuff I have been talking about.

Note to pitchblack....... I never said that someone needed a line service to make selections based on a system.

I'd rather get the best line possible and not have to log in to all of my accounts and manually hunt for the best line. The line service along with the updated lineups/injury report and other tools it offers is what I am referring to.  I think that clarifies my position.

Kirby, that is an interesting idea. 

Paraphrasing one of the better, more succinct descriptions I've heard of Beaver's system:

"When Vegas goes out on a limb, side with Vegas"

Currently, one proxy for how far out on a limb Vegas has gone is Beaver's intervals for the delta between KenPom's spread and the opening line of Vegas.  It sounds plausible that, another variable to include in that proxy is the "thrill score."  Reason being, there is a positive correlation with the grandeur of the event and the sum of money Vegas takes in on the event.  One might assume, events with high thrill scores and big deltas equate to Vegas really going out on a limb.

As with anything, more data is needed.

BWD --

I am looking strictly for a line difference of 4 or greater. In the case of the first selection Virginia.  Vegas has them - 17. The sagarin ratings have VA by 23.4 so the difference is 6.4 and thus meets the criteria for a play.

One of the things you'll note is the vast difference between team rank as well. VA is shown ranked #23 while BC is #259.

To illustrate yesterday St Louis +3 was a play vs Xavier. St Louis was ranked 24 while Xavier 54. Sagarin will usually come up with a play of a higher ranked team as a dog. It's pretty spot on.

I tend to look at this after conference play starts because the numbers start to fall in line once you can get a better gauge of how the teams match up. When you have the cup cake games in the mix early in the season you don't get a good gauge of things to come in conference unless it's a big time match up.

I don't look at added games at all, because frankly it's a crapshoot no matter how "good" one team looks against another. I'd rather play on the better known teams than say the IUPUI's and Citadels out there. To me that's forced betting and frankly how many people can name a starter from the IUPUI squad without looking or being a homer. I'd say none unless it was some phenom we all heard of.

Thanks for your question.

IPFW 29.5
Pacific 21.4
UC Riverside 18.9
Pepperdine 25.1
Montana St.
CSNR 20.4
Sac St. 19.9
Portland 17.5

So... Your philosophy is the opposite.  You side with a predictive computer instead of Vegas when there is a large delta.  

I would say yes he is.  He is taking Sagarin over Vegas.  Ive tried playing systems like this off of Sagarin before and never had any luck.  Gonna watch this,

BWD