Lucky & Unlucky WRs Through Week 8
5I’ve been mulling around this idea for awhile and was hoping to get some feedback on the concept. I think it’s close but could use some refinements.
The idea is that some WRs are “luckier” than others. Luck is a strong component of the NFL. While it’s hard to put an exact number on luck in the NFL, from a win-loss perspective, it seems to range somewhere between 24%-52% [source]. If luck plays such a large part of a win or loss, luck must play an impact in individual performance.
Thought Process
WRs receive targets throughout a game. There are a wide variety of reasons that a target doesn’t result in a completion (drops, interceptions, poor throws, knockdowns, injuries, etc). My first approach was to analyze the average Catch Rate, look at the standard deviation for each WR – examine who was above/below 1 standard deviation from the average to determine which WRs have been “Unlucky” or “Lucky”.
Faulty Logic
I realized that this logic did not take into account the skill of a WR’s QB. Peyton Manning’s WRs had excellent Catch Rates, while Blaine Gabbert’s WRs had poor Catch Rates. This resulted in a bias for players with good/bad QBs.
New & Improved Logic
Each QB has an average Completion Rate. Each WR has an average Catch Rate. What if we examined how a WR’s avg Catch Rate compares to their QBs avg Completion Rate? Would this determine “luck”?
Unlucky WRs
Name | Catch Rate | QB Cmp % | CR – QB % | FPTS ↓ |
---|---|---|---|---|
18-A.Green | 54.8 | 65.6 | -10.8 | 149.4 |
83-V.Jackson | 45.6 | 58.6 | -13.0 | 127.3 |
88-E.Sanders | 54.4 | 66.2 | -11.8 | 82.6 |
13-T.Hilton | 50.9 | 61.1 | -10.2 | 81.5 |
85-K.Thompkins | 41.8 | 55.7 | -13.9 | 80.4 |
11-M.Wallace | 46.9 | 59.4 | -12.5 | 75.8 |
10-R.Woods | 42.3 | 58.5 | -16.2 | 64.5 |
84-S.Hill | 50.0 | 59.3 | -9.3 | 63.0 |
13-C.Givens | 48.9 | 60.7 | -11.8 | 57.4 |
18-S.Rice | 42.9 | 61.0 | -18.1 | 56.1 |
18-G.Little | 38.9 | 56.1 | -17.2 | 48.8 |
81-D.Heyward-Bey | 51.4 | 61.1 | -9.7 | 43.0 |
Lucky WRs
Name | Catch Rate | QB Cmp % | CR – QB % | FPTS ↓ |
---|---|---|---|---|
87-J.Nelson | 72.2 | 67.1 | 5.1 | 145.9 |
10-D.Jackson | 60.8 | 56.2 | 4.6 | 142.3 |
80-V.Cruz | 61 | 55.7 | 5.3 | 138.7 |
84-A.Brown | 76.7 | 66.2 | 10.5 | 131 |
11-J.Edelman | 68.6 | 55.7 | 12.9 | 106.2 |
82-M.Jones | 70.6 | 65.6 | 5 | 102.9 |
87-R.Wayne | 66.1 | 61.1 | 5 | 101.8 |
13-K.Wright | 69 | 57.2 | 11.8 | 89.3 |
10-B.Gibson | 69.8 | 59.4 | 10.4 | 80.6 |
14-J.Blackmon | 60.4 | 54.7 | 5.7 | 76.5 |
11-J.Kerley | 65.9 | 59.3 | 6.6 | 73.6 |
89-D.Baldwin | 76.7 | 61 | 15.7 | 66.2 |
Issues With New Logic / Feedback Please!
I still don’t think my model is quite there. I can’t seem to wrap my mind around how to change this. I would love to hear any feedback you may have.
How Do I Use This?
I focus most of my time analyzing WRs, as they tend to have the biggest upside for me (PPR). I recommend using this in a few ways:
#1 Use to identify “unlucky” WRs to try to trade for them, as theyshould bounce back to at least the average.
#2 Use to identify “lucky” WRs on your team, who you may want to try to trade before they fall back to the average.
#3 Pick up “unlucky” WRs on the waiver wire, wait until they return to the average for better performance.
Just from a quick look at your lucky/unlucky lists I’m guessing you might need to adjust for the depth of target. AJ Green and Mike Wallace are probably getting more deep balls that are long shots anyway while others on their team (Tyler Eifert/Jermaine Gresham and Brian Hartline) pick up the shorter, high percentage throws.
Thanks for the feedback Steve. On Reddit, a user suggested to take into account WR1/2/3 status. Really appreciate the feedback.
I think Steve got most of what I was going to say, but that having been said, it may also be beneficial to take a look at the relationship between “hurried” quarterbacks (maybe even key O-line injuries?) and the depth of the ball thrown. Chances are, hurried quarterbacks won’t be throwing the ball 50 yards, thus making targets whom are likely to be less deep (our Wes Welkers and, for now, Jarret Boykins) more likely to get targets over the deep threat likes of James Jones and Dez Bryant. Good luck man, I’ve thought about doing something like this as well, but not for Fantasy or NFL even, just for shits on college basketball. I feel like a lot of statistics aren’t being utilized the way they should be.
Are you working with R? That sounds fun.
It would seem pulling in historic catch rates for the WR would be a better determinate of luck. What if the wide receiver just isn’t as good as the rest of the WRs on their team?