Customizing David Wilson’s GLSP Projection
At the beginning of the season there will be a few players that will be difficult to forecast using the GLSP app (for an explanation of GLSP see this post) because the usage we’re about to see won’t be anything like the usage the player has seen in the past. It would have been possible to build in additional customization options into the app but I didn’t because a simple way to set your lineup at the beginning of the season is by starting guys in the order they were drafted. But even if the GLSP apps offer a limited amount of customization and some players fall through the cracks (new starters, rookies, etc) I can create custom projections in my system so I thought I would do that for David Wilson.
My sense is that Wilson will be a back very similar to a Chris Johnson (although maybe not 2009 Chris Johnson), DeAngelo Williams, Reggie Bush 2011-2012 or even Ahmad Bradshaw. That doesn’t mean that he’s going to share a running style with those players, but I do think he could have a similar usage profile. Those players are also smaller in terms of the size spectrum for feature backs. I created a hypothetical Wilson and then ran GLSP projections for his matchup with the Cowboys and the following table shows the 25 similar games that make up the projection.
You can see that several of the games are actually against the Cowboys 2012 defense which is important. Those games should at least inform our expectation of what we might get from Wilson this weekend. Also notable is that the top two most similar games are for a Giants running back against that DAL defense. If you assume PPR scoring then the median projection comes out to be about 17 fantasy points which will be just fine for your RB2.
It’s probably worth discussing the validity of the exercise that I’ve just gone through and whether it has value from a prediction standpoint. I think it does and the reason I think it does is because it’s essentially just a more formalized version of the same thing you might do mentally. You might try to think about what kind of RB Wilson is, and then you might try to think about how an RB like Wilson would do against the defense he’s facing. All I’ve done is dug up actual game results for similar players against similar defenses. All of the matchups might differ slightly from David Wilson vs. The Cowboys although in composite they will be very close. I haven’t posted the player averages in order to save on space but if I had you would be able to see how close these defenses are to the 2012 version of the DAL defense and the same would also be true of the players when compared to my hypothetical Wilson (I assumed that hypothetical Wilson would average about 17 carries for 80 yards and a touchdown every other game, and then also assumed about 15 receiving yards). I’ve also leaned on the magic of automation in coming to these results as it would have been very difficult to do the same thing without an algorithm that scores games by similarity.
Like I said, I left additional customization options out of the GLSP app because I think in a few weeks we won’t need them. Until that time a reasonable strategy is to just start your studs (of which David Wilson is definitely one).