I do not watch sports and the only gambling I do is on Wall Street. However, I am going to share with you a technique you can use to increase the odds that you will win a college football bet.
For the past 15 years I have been following the computer models designed by Jeff Sagarin. I do it because I like numbers and have a minor interest in how Ohio State does. His work is published on the USA Today website. Jeff is an MIT graduate that was an early pioneer in computer modeling of sports. My observation is that after about week 3 in a season, his models tend to be more accurate than the Vegas line. Well at least they have been for the games that I have followed, which is primarily Ohio State.
A month ago when Ohio State was ranked #1 in the country, Sagarin’s model disagreed. He didn’t even have them in the Top 10. When I added up the numbers, I knew it was very likely that #19 ranked Wisconsin would win. And they did.
Here is how you use the Sagarin model to determine how you should bet on college football. Note that Vegas sets the odds in a way that half the betters will lose every time. I would only use the model below when Sagarin disagrees with the Vegas line. By how much? That depends upon your risk tolerance. I personally don’t bet, so this is all academic for me.
For this example, I will use this Saturday’s Ohio State vs Iowa game.
Check With Vegas
Yahoo! Sports Odds shows that #8 Ohio State is favored by 3 points against #21 Iowa.
Check With Sagarin
Go to Jeff Sagarin computer ratings page on USAToday.com and select the College Football NCAA rating by team for the current season. Now you need 3 numbers: Home team rating, visitor rating and the home field advantage number.
Now add it up to see the what the computer ratings model believes the line should be.
Ohio State 85.96 – Iowa 82.87 = 3.09
The game is at Iowa, so we subtract 2.94, which gives Ohio State a 0.15 point advantage over Iowa.
Sagarin believes the game is even and that Ohio State should not have a 3 point advantage.
On an even bet, you should go with Ohio State. But on points, you should go with Iowa.
Is this method right every time? Absolutely not. This is gambling. The best time to use the Sagarin vs Vegas method is when the difference is significant. I wish I had saved the data from just prior to the Ohio State vs Wisconsin game. It was a slam dunk to bet on Wisconsin.
Sagarin also does modeling for other sports. I have not followed those numbers.
UPDATE: Ohio State won by exactly 3 points. Vegas got this game right, but using Sagarin’s ratings any OSU bettor would have paused to make that bet.
eric
Dec 2, 2011 — 2:43 pm
I think you subtracted the wrong way. It’s at Iowa, so it would be 85.96- (82.87-2.94)= 6.03, not 0.15.
Using that logic you would assume OSU should be favored by 6, so OSU -3 would be a good bet.
MAS
Dec 2, 2011 — 2:48 pm
@Eric – No, I think It added correctly. Iowa was the home team, so the 2.94 home field advantage would offset OSU’s Sagarin advantage. Had the game been played in Columbus, it would have have been a 6 point advantage.
Nick
Dec 2, 2011 — 3:22 pm
@Eric and MAS.
Mas is actually right, though the math is a little confusing.
Eric, since Iowa is at home, you should actually add that number to their rating before doing the math. Teams are usually considered to have about a 3 point advantage when playing at home.
Or, as MAS did, subtract it from the resultant spread (which is the same as subtracting it from Ohio State’s rating before doing the comparison).
And MAS, despite the outcome, Iowa would have been the better bet for sure. For one, the game actually pushed, so either bet would have just been returned to the better.
Also, I’m not sure why there are two sets of numbers, but going by the first set, Iowa was also the cheaper bet. To bet on Ohio State, you would have had to bet $115 to win $100, verus $105 to win $100 on Iowa.
Those betting odds often move ever so slightly to fine tune the line, before they books resort to adjusting the point spread.
Unless, of course, those were the odds on the money line (straight win or lose) bet. Hard to tell with a 3 point spread. 🙂