Statistical Handicapping in the NFL: Part 1, Sides
October 10, 2006
Using statistics is a very important part of getting an edge in betting the NFL. Starting this week, I will be adding other stats to the Point Spread Power Ratings
(PSPR) that I make available on my site.
After I have weighed all relevant motivational factors in a game, I analyze these stats. The best bets are those in which you have both an emotional and a statistical edge. I sometimes make bets when one team has a big motivational edge and the stats are even. Or, when emotions are even but one team has a big statistical advantage. Games where one team is up on emotion and the other ahead on stats mean a pass.
Let me explain the stats given in the "Sides" part of the spreadsheet by working across the columns left-to-right:
PSPR: This is the line (all are with respect to the Home team) based on a summation of a number of stats that 20 years of handicapping experience have taught me are relevant to covering the spread. As described in more detail in my book, I keep pts per 100 yds, rushing attempts, yds per pass and sacks; all for and against, to name just a few. The final line that appears in the column factors in the appropriate Home Field Advantage (HFA), Away Field Disadvantage (AFD) and strength of schedule for each team.
"Fair": Before I look at the opening lines each week, I play "bookie" and make a line as if I ran my own offshore sportsbook. I ask myself, "what would the line have to be in order to evenly split the betting action amongst my customers?". I never pay more than that line to bet on either team.
"MC": This stands for Monday Consensus. It is another "price check". I never pay more than the consensus line that has developed by late afternoon or early evening the day after the opening lines are posted (I use DonBest.com's free odds). So if Team A was -7 on Monday evening, I would not lay more than that on them nor take less than +7 on Team B.
"Total": is the "MC" total (Over/Under) for the game.
I use it to make the "Push Score" for the game. That is the final score that would push both the side and total of the game. I round up or down to likely numbers; for example, 22-19 would become 23-20 for a line of -3, 41.
Predicted Score: This is a statistically created score based on each team's offensive and defensive yards per game and points per 100 yards multiplied together. I am looking for discrepancies between this score and the pointspread and the linesmaker's push score.
Turnovers: This last column is a three week moving average of each team's net turnover differential. I look for an absolute difference of at least three (e.g. +2 vs -1) and look to bet on the negative team. Short term aberrations in turnovers tend to even out, so I am looking for a team that has been extra lucky (to bet against) or one that has been extra unlucky (to bet on) or both.
Reed Hogben M.D.
Questions? Contact me.