Advice

Gabriel and Hogan Topped 2016’s Receiving Efficiency Results

Last offseason I introduced my metrics YPCOE and YPTOE, which stand for Yards Per Carry/Target Over Expectation respectively. I used them to identify season-long regression candidates for rushing and receiving efficiency, and it successfully identified Davante Adams and Dez Bryant as bounce-back candidates, while also correctly identifying Tyler Lockett and Sammy Watkins as players likely to regress. I’m going to do that again this year — for both rushing and receiving — but I’m going to expand upon what I did last year, because I’ve made significant improvements to the methodology, which has led to new insights.

Methodological Changes

In last year’s article, I simply regressed receiving TD rate (reTDRT) against receiving fantasy points over expectation per attempt (reFPOEPA). I did this to remove the TD part of the fantasy points over expectation so that we’re solely looking at yards over expectation per target. This gave some wonky results, which were both field position and attempt dependent. Instead, I built a neural network model to regress targets, TDs, and field position against receiving fantasy points over expectation (reFPOE), removing the per-attempt designation, since I’m now also including targets as a factor in the model. Yes, reFPOE includes field position in it, but I found the TD adjustment part of the reFPOE calculation was actually under-fit (which isn’t a bad thing, it’s better to be conservative), so my model is a bit more aggressive on the field position part. The other thing I did, which will be for part two of this series, is to see if YPTOE is depth dependent as well (it is). Using Josh Hermsmeyer’s air yards database, I depth adjusted the YPTOE results, to get depth-adjusted YPTOE (da.YPTOE — someone help me think of a better name for this metric). Finally, I noticed there are certain players whose YPTOE and reTDRT numbers are somewhat “sticky,” i.e. somewhat repeatable from year to year. I will dive into this more at the end of this series. I believe (but can’t be sure yet), by depth adjusting YPTOE, plus making one other adjustment, we’ll have an efficiency metric that is somewhat stable from year to year. Anyway, let’s dive into the 2016 results for the non-depth-adjusted version of this metric.

2016 Receiving Efficiency Results

Subscribe to the best value in fantasy sports

You're all out of free reads for now and subscribing is the only way to make sure you don't ever miss an article.

By RotoDoc | @RotoDoc | Archive

Comments   Add comment