The two major sports magazines in the country, Sports Illustrated and ESPN the Mag, are always vying for readership. What helps get more readers than being right? Plus, if you want People to buy your magazine (I mean patrons, not the other magazine) then you have to be funny or right. And right is the best way.
I’m looking at which magazine provided the better NFL Preview issue last season, via accuracy of their picks. The 2014 season previews just arrived in my mailbox last week, so hopefully it’ll help us extrapolate results ahead, giving us a better idea of what these teams records could be. And, to assure you, I am not pointing back, in retrospect, and ridiculing their picks because they’re wrong, merely making observations. I had the Houston Texans as a Super Bowl contender last season (along with the magazines) but then they went 2-14.
The first troubling thing is that the two do not cover football in the same manner. ESPN the Magazine is the one I prefer, because it gives projected records, schedule breakdowns, best/worst case scenarios, playoff chances, and interesting side-bars. Last year they had “Non-QB MVP” and this season it’s a segment called “Secret Superstar” which can help familiarize a casual fan with the NFL’s unsung heroes.
So, to determine how well ESPN the Mag did for predictions, I calculated out the Predicted Wins and used the absolute-value difference between that and Actual Wins to figure out how many wins their projection was off by. Out of all 32 teams, ESPN picked five records correctly, which doesn’t sound impressive, but is considering the variables in a 16-game season with injuries, suspensions, morale, etc. They picked the Cincinnati Bengals with 11 wins, the Denver Broncos racking up 13 wins, the hapless Oakland Raiders managing four wins, the St. Louis Rams an impressive seven, and – my personal favorite – the Dallas Cowboys, once again, finishing a mediocre 8-8. They predicted nine more teams within one win of their actual record. That makes for 14 of 32 to be predicted within one game of their actual finish, a 44% clip.
The numbers also varied by division. For example, the super-competitive NFC South had an absolute value deviation of 5.5, meaning that the predicted record was, on average, off by nearly six wins – a gigantic sum. Conversely, the easy-to-predict AFC North finished with an absolute value deviation of 0.75, a minuscule number. Essentially, they nailed it.
Here’s the divisional breakdown of how well ESPN did.
The average difference between ESPN projections and actual finish was 2.59 games, a modest finish. That number would be much lower if it wasn’t for the vastly under-performing Texans, Redskins, and Falcons squads and the large disparity between the 3-win projected team of the Arizona Cardinals winning 10 games.
Here’s the full chart.
Sports Illustrated actually fluctuated from year to year. In 2013, they essentially gave power rankings of the entire conferences, but gave no records predictions. This season they gave divisional predictions, but again without the record. It’ll be tough to compare them year to year as they tinker with coverage. Maybe I’m just bitter that their method doesn’t give me numbers to compare, but it does seem like the win predictions are useful in analysis. There are positives in that, just ranking the best teams flat-out without a record, but simultaneously it’s always nice, as an analyst, to see their breakdown and their why. Yes, SI does that with their write-ups (all done by one staff member) and they re-cap high-impact additions and subtractions, but their analysis is lacking compared to ESPN the Mag and their ink is used on summary of the off-season.
Nevertheless, we can take the power rankings, assign each team a number and then contrast that with their ending result for the year.
Sports Illustrated did not pick the Miami Dolphins to win the AFC East like ESPN the Mag did, but they had some highly-erroneous picks as well, highlighted by the article bashing the Carolina Panthers everything, ranking them last, only to see the blue-and-black capture second-place.
Since “power-ranking” of teams is significantly easier than picking a win-loss record (it is so based on volume and opportunity for certain things to go wrong), then SI gets less sympathy. They did well in the AFC, picking the top five teams in generally the correct order, but the one team they were really down on – the San Diego Chargers – burned them in the long run.
Also, I recognize that this may not mean much because it is only a one-year study, but I hope this can become an annual study and see if trends develop.
So while this study is far from having a hard-line conclusion, it appears that, if you’re a betting person, ESPN the Magazine is the advice you should heed.