I spotted a prop yesterday that seemed simple to analyze and almost too good to be true. I ended up winning but the skeptic in me says that it's likely I missed something. I'd appreciate a critique of my reasoning on the following. If I got it right, well, cool -- others might find this post useful.
The book was offering this: "First team to receive a penalty: Alabama -115 / LSU -115". This immediately caught my eye because it seemed very unlikely that both teams commit penalties at the exact same rate. I checked their season records and it turned out that Alabama averaged something like 3.9 penalties per game while LSU averaged 6.25. I threw out the handful of non-FBS opponents and the ratio was basically the same. I didn't compute the variance but it seemed small, and their head-to-head matchup penalties of Alabama=6, LSU=7 didn't seem far out of line (and more importantly, it preserved the ordering).
This told me that if both teams commit their average number of penalties, about 62% will belong to LSU. I found the effective fair odds of the favorite by looking at the ratio, 6.25/3.9 = 1.6, then by doing the usual conversion to get -160. A -115 payout for fair odds of -160 seemed terrific. Too good, actually.
Am I missing anything?
The book was offering this: "First team to receive a penalty: Alabama -115 / LSU -115". This immediately caught my eye because it seemed very unlikely that both teams commit penalties at the exact same rate. I checked their season records and it turned out that Alabama averaged something like 3.9 penalties per game while LSU averaged 6.25. I threw out the handful of non-FBS opponents and the ratio was basically the same. I didn't compute the variance but it seemed small, and their head-to-head matchup penalties of Alabama=6, LSU=7 didn't seem far out of line (and more importantly, it preserved the ordering).
This told me that if both teams commit their average number of penalties, about 62% will belong to LSU. I found the effective fair odds of the favorite by looking at the ratio, 6.25/3.9 = 1.6, then by doing the usual conversion to get -160. A -115 payout for fair odds of -160 seemed terrific. Too good, actually.
Am I missing anything?