Mean Regression Record to Date
Sides: 41-30, 57.75%, +8.0u
Totals: 31-30, 50.82%, -1.7u
OVERALL: 72-60, 54.55%, +6.37u
I did a fair bit of analysis on the past performance this morning, concentrating mostly on the totals since they are bringing down the win rate. However, I think there has been a bit of bad luck: The average winning total has won by 4.5 points more than the average losing total has lost. Another way to look at this is if you give or take 3 points from all of those games, a worst case scenario would only be 29-32, whereas a best case scenario would be 37-24. Given that info, our expected record is 33-28, for a 54.1% win rate. Essentially, volatility has not only not been on our side, it's actually been uncharacteristically unfavorable.
The "losing" record did give me a chance to review a number of variables, and, I believe, improve on the accuracy of the plays to some degree, eliminating some of the 50/50 and 52/48 games, and focusing on those that truly are not only likely to win, but likely to win at a rate that makes us profitable.
I added two new factors into the mix, both performing much the same as the rest of the factors - identifying historical statistical outliers that are due to chance and randomness, and not due to a team's true ability. That is, it is incorrect and inefficient to use these variables as predictors of future performance, and by correcting for these, we see which lines are most susceptible to our strategy.
Good luck tonight. A large card, and 23 plays here. 13-10 or better would be nice. Or 23-0.
Duke O139' Louisville O131 Stanford O130 USC O116' Wisc-Mil O125 Bama O134 Eastern Illinois U118' Weber U148 UCLA U123' IUPUI U144' StMarys U138 Pepperdine U135
USC +14 NoColorado +5' FloridaAtlantic -6' Troy +6' Detroit +3' Pacific +3 Iowa +5 SIU-Ed +9' Tenn-Chat +6 Idaho +1' San Diego +10