Ramblings from the Ultimate Sports Sentimentalist

Well, the college football season has come to a close, and it’s time to look back to see how accurate my preseason predictions were compared to the experts.  There are many categories/systems that can be used to compare the accuracy of a prognosticator’s picks, and there are three such methods which I implore.  The first, and most important in my opinion,  is by looking at each team and seeing which expert came closest to correctly predicting that team’s ranking in the final conference standings.  The second method I employ is the Stassen Survey, which has been popularized by fellow preseason prognosticator Phil Steele.  This second system looks at the predicted conference finishing position for each team and then compares that with their actual postseason conference rank.  The difference between those ranks is then recorded for each team and added together.  For example, in 2009, I predicted Michigan to finish 9^th last year in the Big 10 and Phil Steele had them at 6.  They ended up finishing 10^th so my score would now be 1 (10 minus 9) and Phil’s would be 4 (10 minus 6).  You then continue this process for all teams.  In addition, I also compare my preseason rankings of the top 120 teams in the country, and those of the experts, with a final top 120 ranking.

Overall, this is the fifth season I’ve conducted this full, in-depth analysis of my preseason picks and have had mixed results in doing so.  I ended up being more accurate than the experts in both 2007 and 2009, less accurate in 2008, and about the same in 2010. (see full history here)   This year I had some highlights and lowlights, but overall my predictions were slightly worse than Phil Steele and most other preseason prognosticators.  With that being said, I still had some ingenious predictions which I will rehash shortly.  Before I explain that though, let me go ahead and give you a breakdown of my performance against the experts in all of the statistical categories mentioned above.

Preseason ranking of all 120 teams: In the preseason, I ranked all teams 1-120 and also took an average of the full 1-120 preseason rankings from several different preseason publications.  The experts’ average picks more accurately predicted the final postseason outcome of 62 teams.  I was more accurate in my analysis of 54 teams and there were four ties between us.  Believe it or not, I actually tied the experts in total ranking variations for each team.  If you add up the disparity between my preseason rankings and the final rankings, as well as those from the experts, you will find that we both scored exactly 2,484 points.  It’s unbelievable that such a complex calculation could produce a tied result.  We also tied in the number of teams we predicted exactly right.  The experts correctly picked Purdue to finish 85th in the country, and I correctly pegged New Mexico to finish at # 119.  (Note: I used Pre Snap Read’s Final Top 120 Ranking as the final poll used to compare my preseason and postseason predictions).  In the past, I have used the “CBS 120″ for this purpose, but this season CBS moved to a computer ranking system which does not accurately reflect the general consensus of where each team finished.)

The Sport Sentiment Survey (My scoring system): Using my desired scoring system, where one looks at each team to determine which expert came closest to correctly predicting that team’s final conference ranking, Phil Steele beat me in 3 out of the 4 main categories.  He more closely picked the finishing position of more BCS teams (18-14) and also more total Division 1-A teams (31-28).  Further, he also more accurately predicted the final standings of the BCS conferences than I did by a 3-2 margin.  I was more accurate in my predictions of the SEC and Big 12 but he beat me in the Big 10, Big East, and Pac-12.  We tied with our ACC predictions.).  In regard to total conferences, we tied at five with me being more accurate with my preseason predictions for the Big 12, SEC, MWC, MAC, and Sun Belt and him being better with the Big 10, Big East, Pac-12, C-USA, and WAC (as mentioned previously we tied in the ACC.).  See chart below for full results.

Stassen Survey: This is Phil Steele’s suggested scoring system which is described in detail above.  Phil beat me using this scoring system as well.  See chart below for conference-by–conference breakdown.

Bowl Games: Phil beat me here again.  I correctly predicted 7 bowl slots and 0 total bowls, while Phil got 9 bowl slots and 1 full bowl.  While normally such an across-the-board loss to Phil Steele would make me quite disappointed with my preseason picks, it must be stated that Phil had a really good year handicapping college football.  I am sure Phil bested almost all preseason publications, in addition to mine, this season.

Surprise Picks: Two of my predicted surprise teams (Rutgers and Georgia) ended up surprising whereas the other six did not.  I also ended up with a negative number in the sliding scale analysis I use that takes into account the confidence/significance intervals I provide for each surprise team. (See surprise picks entry for more information about that)

Disappointment Picks: If you are looking for a bright spot in my preseason predictions, this is it.  A whopping five of my predicted disappointment teams ended up falling short of their preseason expectations.  I also ended up with a positive number in the sliding scale analysis I use that takes into account the confidence/significance intervals I provide for each disappointment team. (See disappointment picks entry for more information about that)

So after taking all this into account, I must conclude that I overall performed a little below average with my preseason predictions.  With that being said, I still hit the jackpot on a couple of key matters.  First of all, I correctly predicted Notre Dame to underachieve (why do experts keep overrating these guys every single season!  Notre Dame is the one team in the country that I have evaluated more accurately than the experts in each of the past five seasons!).  I also correctly picked Georgia to win the SEC East when the whole world picked South Carolina to take it (including arch rival Andrew Hogan).  Furthermore, my miracle “Rutgers to the BCS” pick almost came true, as the Scarlet Knights fell short of the Big East crown by a single game!

I will now methodically rehash my preseason’s rankings, as well as the experts’, to show you our brilliance as well as our stupidity.