College Football Has Kicked Off But More Data Is Necessary For Actual Picks
College Football Kickoff
Aug 30 marked the kickoff weekend of the 2014 college football season. My cursory impressions of the weekend in the rearview mirror: An outstanding performance by Texas A&M on both sides of the ball was balanced by disappointing efforts by both Alabama and Florida State. Anyone reading this blog may have felt it curious that there were no picks available on this site. However, it takes three weeks of data from games played for my handicapping model to begin to caibrate itself. The first week of games that picks will be made available will be Sep 17-20 game week. It takes over 200 plays for a team's performance variance to be statistically significant and of some predictive value.
Some may view the fact that I "sit out" the first three games of the year from handicapping as a big disadvantage. The other side of that coin is that my handicapping model will be more accurate with each snap that a team takes. In my eyes it is much better to keep my discipline and make picks only on games that I have a statisitical and analytic edge. Until there is enough data from the current season, any picks from me would be no more than an educated guess, and that is NOT good enough for me. Remember I'm trying to achieve a minimum 68% accuracy against the line for the cumulative season.
What To Expect In The Near Future
1st Pick Week
Sometime before Sep 18 I will issue my first game picks for the college football season on this site. At that time, however, not every team will have taken 200 offensive and defensive snaps. The net result is that the match-up list of forecastable games will be the most sparse of the regular season. Further, the larger FBS schools may still have non-FBS teams on the schedule and I exclude any games with Vegas lines in excess of 28.5 points. As a result I'm anticipating that there will be approximately eight games with enough of a statistical edge to make picks on.
2nd Pick Week & Thereafter
Sometime before Sep 25 (remember there are mid-week games on ESPN and FOX) virtually all FBS schools will have taken at least 200 snaps on both sides of the ball. At that time I anticipate a full complement of weekly matchups that will have a significant statistical edge. From the second pick week until the end of November I anticipate that there will be 13-23 games each week with a satisfactory statisitical edge.
For some, the prospect of playing 13-23 each week will seem prohibitive. As a result each game will be ranked in descending order from the largest staticial edge to the smallest. I further anticipate that the "pick mix" will be about 60% favorites and 40% underdogs, so be prepared to play both sides of the line often. I personally view the high number of available games to play as an advantage. I view the large number of potential picks as a systemic strength. It provides both diversification as well as an opportunity to introduce leverage into your playbook.
A bit of a warning: Don't automatically expect the largest statisitical edge to be the games with the easiest covers; my handicapping model does not ALWAYS work that way. These are merely the games that the model flags as having the highest probability of covering the line, possibly by a scant point. That being said, the average "beat" against the line for the entire 13-23 weekly portfolio should be approximately 13-16 points.
Introducing Leverage Into The Equation
I will also make some parlay suggestions in order to take advantage of leverage. In my back-testing efforts from the 2013 season, the weekly variance of the accuracy of my game forecasts was remarkably low while the cumulative accuracy was quite high. If I can come close to duplicating the results from the 2013 back-test, then introducing leverage via parlay suggestions should be quite profitable. I will eat my hat if I am not profitable/successful in my parlays.
Given that parlays are typically viewed as a sucker's bet by most astute sports players, I can hear the collective groans reverberating through the web right now. That being said, I'm calling my "shot" now and I will own it if I'm wrong at the end of the season.
The Suspense Is Killing Me
I'm the most anxious for the first three weeks of college football to get in the books that I've ever been in my entire adult life. A mountain of homework has been done over the past three months in order to get to this point so I'm incredibly interested to see if the handicapping season pans out as I hope. My plan is to publish the picks for free each week through the end of October. If, however, the picks are 68+% successful then only a partial list of games, perhaps only one game will be published on a weekly basis.
More To Come In A Couple Of Weeks - Jack
