Last week, I looked at a variety of defense and special teams statistics and their relation to fantasy value. Today, we’re going to do the same thing with quarterbacks. The metric that I’m correlating stats to is Plus/Minus, which is actual points minus expected points based on salary. This is super important – we’re looking at the correlation of stats to Plus/Minus, not to fantasy production.
The reason the distinction is important is because it includes site pricing. Rather than just looking at the most important statistics to quarterbacks scoring fantasy points (it will obviously correlate highly with touchdown rate), it’s more important – in my opinion, of course – to see how sites price players. Knowing quarterbacks who throw a lot of touchdowns is nice (and basic knowledge), but knowing how DraftKings factors their touchdown rate into their pricing is more useful.
So, here are a variety of statistics that I’ve compiled and their respective correlations:
In case you don’t already know, ANY/A is adjusted net yards per attempt, aDOT is adjusted depth of target, and aC% is depth-adjusted completion percentage – see a good explanation of the last two in this excellent Pro Football Focus article.
Before we go over specific stats and observations, I think it’s interesting to note that of the top 40 quarterbacks I looked at last season, 30 of them had a positive Plus/Minus. That suggests that DK underprices QB’s in general — anecdotally, I think it might have to do with customer satisfaction. If DK priced someone like Aaron Rodgers where he probably should be, a new user coming to the site might be turned off by not being able to roster him and still get other good players in their lineup.
The thing that stands out the most to me is the lack of correlation between Plus/Minus and the Vegas numbers: average spread and average implied points (based on the over/under and moneyline). Again, this doesn’t mean that those statistics aren’t important to quarterbacks; they absolutely are. However, this data shows us that last year DraftKings priced this heavily into QB salaries. The Vegas numbers are important as a baseline, but don’t expect to find hidden value here.
Interestingly, the stats that have interceptions as a part of them correlate highly with Plus/Minus. This is a little surprising to me, as DraftKings doesn’t penalize QBs for interceptions as much as season-long leagues – they’re minus-1 on DK and generally minus-2 in season-longs. However, maybe this is a bit overstated and interceptions are a little more important than I initially thought.
ANY/A is also correlated highly with Plus/Minus and also has an interception component in the calculation:
The last one we’ll talk about is passing yards per game, which correlates more strongly than the other touchdown/passing yard stats. Most people (rightfully) focus on touchdowns – they’re what you need in bunches if you want to take down a GPP – but it looks like passing yards aren’t factored into salaries as much. I’m not sure how to use this, to be honest, as you still can’t get away from trying to get touchdowns from your QB. I think perhaps it can be used as knowledge to try to find a QB that can do both in a tournament: throw multiple TDs and rack of a bunch of yardage.
Last quick note: I wrote this article because I had people request it after reading the defense one, and I thank you for that. I’m planning on doing the same for the other positions – RBs, WRs, and TEs – but please always let me know what articles you want to read. It makes us both better DFS players.
