Even with the Internet easing the post-season pain and withdrawal from the gridiron action, we still often revel in the pre-season magazines. They’re in some way a more tangible tie-in to the great game and a gleeful reminder that the long awaited new campaign is right around the corner.
With that reality, the next is to ask which of these publications—beyond the joy of any kind of information and illustrations they may provide—are more worthy of our time and money from a forecasting perspective? Many will claim, but few are "chosen."
A tremendous site for tracking such things is Stassen.com (http://preseason.stassen.com/), which covers a lot of ground in comparing pre-season polls and publications. Though titled with "magazines," it lists various computer-based polls as well. It cloaks the accuracy of conference predictions (http://preseason.stassen.com/prediction-accuracy/2009.html), but we’re going to explore how effective these entities are in forecasting our beloved top 25, top 10, and national champion. In this era of the BCS, we see more of an emphasis on the polls than once was the case.
Everyone seems to want to know what most recently happened, so we’ll briefly explore the 2009 prediction results, but the bigger focus will be on three and five year accuracy. Those give us a broader and more meaningful view while still emphasizing recent history.
Stassen, noted above, saved me a lot of time with his tracking data. From it, I noted the degree of variation for each pick and summed them. For example, Athlon’s chose Florida No. 1 last fall, with the Gators finishing No. 3 in the AP, costing Athlon two points (three minus one) for that selection, and on down the line. The lower the total score, the better.
For ’09, an outfit called "CompughterRatings.com" (http://www.compughterratings.com/2009PreseasonRankings.php) took the top rank among 16 competitors in predicting the nation’s best 25 (AP final poll). To clarify, it’s not a pre-season magazine. CPA Rank, also without a mag, came in second, while Lindy’s followed (therefore No. 1 among actual newsstand publications), with Southern College Sports (site only) fourth, Athlon (mag) at No. 5, and Sports Illustrated (SI) in sixth. Phil Steele, with his cup overflowing of information, came in just 9th for last year’s top 25 forecast. Football Outsiders took the disdained last-place designation, with handicapper Jim Feist surviving just a hair ahead.
But that’s just one season, and a single fortunate or unfortunate choice can lift or drag one outfit’s accuracy score beyond reason. If a salivating pass rusher pummels a quarterback into next year, current year predictions go up in flames.
In the three-year comparison (which leaves 2009’s newbie winner out of the mix), we find 12 contestants. CPA Rank, 2009 runner-up, ties with GamePlan for first. GamePlan—just seventh of 16 predictors in ‘09—shows how important it is to take the blinders off of the "what have you done for me a minute ago?" phenomenon.
Phil Steele follows these two co-winners at No. 3, pursued by Southern College Sports and SI to round out the top five. The well-known Sporting News (TSN) sags into 10th position among the dozen, with CollegeFootballNews.com in dead last.
Some may consider a five-year outlook even more meaningful, and while that leaves just eight outfits available for comparison, there’s a lot of value in checking.
In this half-decade, the grizzled vet GamePlan ranks No. 1 in picking the top 25, followed by the more "present-day" CPA Rank, with Phil Steele again rating third. Lindy’s, another long-time and well-known pre-season mag, comes in fourth, while the handicapper’s Gold Sheet unwillingly claiming eighth, or last place.
Do you wonder, though, through this top 25 selection process if maybe some entities are more accurate at picking the "cream of the crop" while faltering when it comes to the less critical teams that languish near the bottom of the polls? [Here, I tabulated like the top 25, but also included the variation when schools finished in the AP top 10 but not in the publications’. For instance, Iowa finished No. 7 last year, yet wasn’t chosen that high by anyone. Yet, some selected the Hawkeyes in the top 25, so those pubs shouldn’t be penalized as many points as those who failed to include Iowa at all.]
Generally, as expected, the predictors rank roughly the same in choosing their top 10 as they do the top 25, though there are notable exceptions. Phil Steele, the supposed stalwart of soothsayers, slumps from third for the top 25 to seventh (next to last) in picking the top 10 over a five-year period. Conversely, the Gold Sheet vaults from last (eighth) to fifth. More importantly, Lindy’s surges from fourth to second.
Once again, though, as with the top 25, GamePlan reigns over the five-year throne for the top 10.
Choosing the national champion seems largely to be a lazy exercise in groupthink, with few trend breakers. No one accurately forecast the national champion in 2005, 2006, 2007, or 2009. However, Phil Steele, while missing the mark on Florida in ’06 by the biggest margin (he chose the Gators No. 20) of any publication, shared the honor with Athlon as the only one getting the Gators right in ’08.
Taking into account the top 25, top ten, and national champion selections over the three-and five-year frames, we find this overall ranking among the top pre-season publications:
- CPA Rank
- (tie) Phil Steele and Lindy’s
GamePlan (http://www.gameplanmagazines.com/) was established in 1970 and, apparently, still does a "pretty good" job of analyzing and forecasting. After buying it a number of times in earlier years, I haven’t done so the last several. That may change after this investigation!