As I said a number times last week, our projections are based mostly on statistical analysis. They are also very aggressive. Few prognosticators ranked Eric Decker in the top 10 or put Cam Newton in the top 3.
The statistics said Newton would be the No. 2 quarterback and Decker would be the No. 6 wide receiver. Both players matched those projections exactly. A collection of miscues were right there with the good rankings, but that can be expected, right?
With that question in mind, I pondered further: “Are the hours spent working on the projections really worth it for our readers?” Based on a rudimentary comparison I ran today I can say, “yes.”
I compared our projections-based rankings to the actual final statistics then compared a leading premium fantasy football site’s projections against the final statistics.
Difference between actual and projected rankings was the rudimentary statistic used in this test. Examples from the wide receiver ranks: We had Eric Decker at No. 6. He finished sixth. The difference was 0. We had Pierre Garcon 44th. He finished first, a difference of 43. The differences for the top 50 players at each of the big four positions were then averaged to get per-player differences.
Then I filtered it further, looking at starters and, in the case of running backs and receivers, all of the No. 1s (RB1-RB14, etc.). The results are below. Again, these are the average differences between actual and projected rankings per player. Note how close the differences are between the two sites (in this comparison 1.00 is one ranking spot above or below the actual ranking, so a difference of less than 1.00 means the rankings were correct/wrong a near identical number of times):
|Running backs||Ninjas||Expert site|
|Wide receivers||Ninjas||Expert site|
|Tight ends||Ninjas||Expert site|
|All players (200)||Ninjas||Expert site|
Take away the tight end rankings and the overall starter numbers (70 players) are within one ranking (16.96-15.90).
Obviously, there is some work to be done (especially in our tight end rankings) but considering how close we are to the experts who do this for a living, we’re feeling pretty good. Our projections will retain their aggressiveness.
When Pierre Garcon can be the top wide receiver any week, no projections are safe or perfect. But I can confidently say our projections, as unconventional as they are, compare favorably to the best in the business.