Posted:
#1
After reading Moneyball by Lewis and Trading Bases by Peta, I decided to formulate my own version of sabermetrics in an attempt to beat the spread. My system downloads daily game, line and spread data from covers.com, then spiders baseball-reference.com to get detailed team and individual stats. For each game, lineup information is used to calculate the pythagorean probability of each team winning. A second pythagorean probability is then calculated weighing the starting and available bullpen pitchers. These probabilities are compared to the implied probabilities from the moneyline and bets are taken where there is an edge of several percentage points or more in both the team and the pitcher pythagorean probabilities.
All plays risk 2% of a starting bankroll of $1000.
Sample output of the system:
------------------------
7:05 PM ET Cleveland [Z. McAllister] (138) at LA Angels [C. Wilson] (-150) 8.5
CLE Pythag: 26.58 | Implied: 42.02 | Diff: -15.44
LAA Pythag: 73.42 | Implied: 60.00 | Diff: 13.42
CLE PitchP: 28.90 | Implied: 42.02 | Diff: -13.12
LAA PitchP: 71.10 | Implied: 60.00 | Diff: 11.10
LA Angels (-150) Bet 20 to win 13.33
------------------------
Here the Angels have a 73.42% chance of winning based on team/lineup stats and a 71.10% chance of winning based on starting and available pitchers. At -150, they have an implied probability of 60%. Since both the pythagorean probabilities for a win are much greater than the implied, the Angels are a play at -150.
0
Hot Forum Topics
