[AdamStock1983]

I hear this all the time in college hoops. Handicappers talk about having to adjust your power ratings. In college hoops many people talk about using Jeff Sagarin or Ken Pomeroy power ratings.

At the risk of sounding like an idiot how do you use power ratings in your handicapping? Ken Pom has his set of ratings. How do you take the ratings and apply them to handicapping? What do you look for when comparing two teams? What numbers do you use? It has to be something more complex than a team rated 4 playing a team rated 39 so the higher rated team is probable to win. Is their a formula to use? I want to know how to apply power ratings to handicapping.

Any input or feedback would be appreciated

[JohnAnthony]

You don't sound like an idiot, you just sound innocent (don't know which is worse in this industry). No one with real specific knowledge is going to share it with you just like that on a public forum. However, wiser gentlemen (I'm pretty shi*t @ College BB) here may throw some useful hints.

[Wrecktangle]

Sagarin's ratings are in points, use them directly.

It won't do you much good though.

[LT Profits]

Sagarin's ratings are in points, use them directly.

It won't do you much good though.

So are Pomeroy's, just look at his predicted scores

[AdamStock1983]

I been handicapping for years. Had my best NCAAB season ever however I have never used power ratings. I did not know how to apply power ratings to handicapping was such a secret. Everyone talks about it openly. I am just trying to become a better handicapper, innocent or not I have no ego.

[BOA12]

I use them in teaser and money line plays where there is 10+ point differential with the line

[TomG]

power ratings are fine you just need a way to convert power ratings into win percentages and/or point spreads

the book justin7 linked to here a few weeks ago about rankings is actually pretty good but a bit heavy in math for casual readers

[LT Profits]

I use them in teaser and money line plays where there is 10+ point differential with the line

Generally speaking, power ratings that vary by that much from real lines are flawed.

[genghis518]

Sagarin's ratings are in points, use them directly.

It won't do you much good though.

I wouldn't say that. I use them for leans against the spread. After factoring in Home field advantage and a few of my other tweaks... i see how much of an edge i can get on the current line. When i see a major deviation... it usually leads to a nice moneyline play. Ex. Pittsburgh.. even with HF Add... I only make them a 2 pt fave... its worth it for me to not only take the points... but also go with the +260

[LT Profits]

I wouldn't say that. I use them for leans against the spread. After factoring in Home field advantage and a few of my other tweaks... i see how much of an edge i can get on the current line. When i see a major deviation... it usually leads to a nice moneyline play. Ex. Pittsburgh.. even with HF Add... I only make them a 2 pt fave... its worth it for me to not only take the points... but also go with the +260

Princeton seems like a popular pick tonight, although not to your extreme. Pomeroy has Pitt -4, Sagarin Pitt -5.5.

[genghis518]

Thats where the tweaks come in. +/- for a few different factors... FT, Rebounds, etc. Everyone has their own system that works best for them.

[Brooklyn Dick]

Princeton seems like a popular pick tonight, although not to your extreme. Pomeroy has Pitt -4, Sagarin Pitt -5.5.

Prlnceton should get killed on the boards tonight.

[genghis518]

and then again.. every time you really like a game... it BLOWS UP in your face. lol

ugliest first half EVER!

[_bester_tuff]

power ratings are fine you just need a way to convert power ratings into win percentages and/or point spreads

the book justin7 linked to here a few weeks ago about rankings is actually pretty good but a bit heavy in math for casual readers

where it is please?

[LT Profits]

where it is please?

http://www.sportsbookreview.com/forum/handicappe...k-ratings.html

[CHUBNUT]

Power ratings only tell you what the line should be based on past results. betting on them is like catching a train and using a timetable from a year ago, it could be there at the same time but chances are its leaving at a different time. thats why latest information is critical. i suggest that the more people know about a sport the less they should use power ratings as theyre a distraction but otherwise theyre good for orientating you as to where the line is going or come from.

[HuskerExpat]

I use power ratings as follows. I set a power rating for the teams that I follow (I follow a couple of conferences). The difference in the power ratings is the hypothetical spread on a neutral court. Then I apply an adjustment for home court advantage (or other factors) and come up with my proposed point spread. When there is a variance between my line and the actual line, then I make a bet.

As noted by others, power ratings you find on the web may or may not be useful. The difference is that my power rating is how I project that the team will play in the next game, whereas power ratings on the web (Sagarin, etc) are more typically a reflection of how the team has played in the season through the present. The two approaches often result in a similar power rating, but not always.

[brewers7]

I have found that most people who use power ratings instead of handicapping don't win...In fact, they lose their ass...And most people who use power ratings do so because they don't know how to handicap...

Now I say "most" because there are exceptions to every rule...

Power ratings can only be used in conjunction with handicapping in a very small informational way...Very small...But I do have one column designated (out of about 1100) in my database for a team's power rating and for me, it's just OFF EFF minus DEF EFF to get the NET EFF, and Net Eff is basically my "power rating"...

[TomG]

maybe people use the term "power rating" differently but i am referring to a power rating as basically a synonym for a statistic (http://en.wikipedia.org/wiki/Statistic): A statistic (singular) is a single measure of some attribute of a sample (e.g. its arithmetic mean value). It is calculated by applying a function (statistical algorithm) to the values of the items comprising the sample which are known together as a set of data.

not sure how one could handicap without using some form of power rating even if it's just subjective from watching games

[WILLIE]

maybe people use the term "power rating" differently but i am referring to a power rating as basically a synonym for a statistic (http://en.wikipedia.org/wiki/Statistic): A statistic (singular) is a single measure of some attribute of a sample (e.g. its arithmetic mean value). It is calculated by applying a function (statistical algorithm) to the values of the items comprising the sample which are known together as a set of data.

not sure how one could handicap without using some form of power rating even if it's just subjective from watching games

You got to have at least a starting point of some sort...i e a power rating of some kind.

[AdamStock1983]

I watch plenty of games but I am almost asking how you can apply power ratings from an advanced metrics standpoint like people do with stats in baseball

[HuskerExpat]

Someone PM'd me, but because I can't respond yet (not enough posts), here is my response:

At the start of the season I'll take a look at Sagarin Rankings, but that's the only time I look at other power rankings as part of the process. Still, my power rankings are often times very similar to Sagarin, but just a point or two difference makes a big difference when setting a line and deciding whether to bet. I'm not really focusing on any one stat but more the overall performance of the team, etc. That's not to say that I ignore stats, but just that it is not part of a formal process of coming up with the power ranking. For reference, my final rankings for Big Ten teams were as follows:

1. Michigan State [92]
2. Ohio State [92]
3. Michigan [85]
4. Indiana [85]
5. Wisconsin [85]
6. Purdue [84]
7. Minnesota [80]
8. Northwestern [79]
9. Illinois [79]
10. Iowa [77]
11. Penn St. [74]
12. Nebraska [71]

Those numbers would give me a basis for the spread between two teams on a neutral court. The home court advantage in the Big Ten is enormous, anywhere from 5 to 7 points, depending on the venue. A key injury is more likely to cause me to drop the power rating of the team a point or two (depends who, maybe as high as 3 or 4 points depending on replacement, etc). Another factor that might go into the additional points might be a bad matchup. For instance, Michigan is a good team, but they get dominated by a team with a physical inside game. In those sorts of games, I'll add another point or two, depending on the matchup.


I don't bet NBA, so I can't really comment on using this for those games. I'm sure the basics are the same, but the home court advantage and things like that are likely different.


A big key to this is my keeping detailed records of every game comparing my line vs. the actual line, and compare that to results. That helps me continually adjust the power rankings and how I make adjustments in specific match ups, for home court advantage, etc.

[AdamStock1983]

Husker what are the numbers on the right of the teams? How did you come up with them. What stats do you use to make your own power ratings or is it simply the eye ball test?

I guess that is where I get confused. How the hell do you come up with your own power rankings? If someone could fill me in on when constructing your own power rankings what goes into it

[HuskerExpat]

The numbers to the right of the teams are my end of regular season power rankings for those teams (they are different today). It is closer to an eyeball test than anything. Most of the hard statistics I use to actually adjust the power rankings come from comparing my line vs. actual line vs. actual results. For instance, for the first half of this season Northwestern was consistently beating my line because I had them ranked too low (off the top of my head I think I had them at low to mid 70s early in the season). So that helped me see that my power ranking for Northwestern needed to come up. As for actual basketball statistics, I use those more for making adjustments to my line in specific matchups.

Keeping records of my line vs. actual line vs actual result also helped me just simply avoid Purdue games. I had very good results for the 11 other teams in conference games, but Purdue was just so inconsistent or I just couldn't figure them out, so I stopped betting their games at all until the very end of the season when they seemed to improve quite a bit and play consistently.

[PatrickBateman]

Does anyone know how the sagarin power ratings are calculated?

[WILLIE]

Does anyone know how the sagarin power ratings are calculated?

Sagarin, like the developers of many other sports rating systems, does not divulge the exact methodology behind his system. He offers two rating systems, each of which gives each team a certain number of points. One system, "Elo chess," is presumably based on the Elo rating system used internationally to rank chess players. This system uses only wins and losses, with no reference to the victory margin. The other system, "Predictor," takes victory margin into account. For that system, the difference in two teams' rating scores is meant to predict the margin of victory for the stronger team at a neutral venue. For both systems, teams gain higher ratings within the Sagarin system by winning games against stronger opponents, factoring in such things as home-venue advantage. For the Predictor system, margin of victory (or defeat) factors in also, but a law of diminishing returns is applied. Therefore, a football team that wins a game by a margin of 7-6 is rewarded less than a team that defeats the same opponent under the same circumstances 21-7, but a team that wins a game by a margin of 35-0 receives similar ratings to a team that defeats the same opponent 70-0. This characteristic has the effect of recognizing "comfortable" victories, while limiting the reward for running up the score.

[AdamStock1983]

What about Ken Pom as well. I look at the ratings and besides the listing of teams I have no clue what they mean and how to apply to handicapping

[WILLIE]

What about Ken Pom as well. I look at the ratings and besides the listing of teams I have no clue what they mean and how to apply to handicapping

Here's three explanations from Mr. Pomeroy himself. Hope it helps answer your questions.

kenpom.com/blog/index.php/weblog/pomeroy_ratings_faq/

kenpom.com/blog/index.php/weblog/introducing_the_kpoy/

kenpom.com/blog/index.php/weblog/free_tim_williams/

[cutchemist42]

power ratings are fine you just need a way to convert power ratings into win percentages and/or point spreads

the book justin7 linked to here a few weeks ago about rankings is actually pretty good but a bit heavy in math for casual readers

Which book is that?

[LT Profits]

Which book is that?

I posted link in Post #15.

[cutchemist42]

Thanks! Missed that!