# When Does Yards After Catch Stabilize for Running Backs?

So far in this series on the reliability of Yards After Catch (YAC), I’ve found that it stabilizes in 22 games for TEs and 31 games for WRs. For both positions, that’s slower than it takes for Average Depth of Target (aDOT) to stabilize. In this post, we’ll find out whether or not RB YAC stabilizes faster than for the other two positions; and if so, if that’s also faster than RB aDOT.

### Methods

As always, below is the standard procedure I use to determine a stat’s stabilization point; this time applied to RB YAC:

1. I collected YAC data for all RBs that played at least 8 games for the same team from 2006 to 2016.
2. Starting with RBs that played 8+ games for the same team, I randomly selected two sets of 4 games for each RB and calculated their YAC in both sets.
3. I calculated the split-half correlation (r) between the two randomly-selected sets of games.
4. I performed 25 iterations of Steps 2 and 3 so that r converged.
5. I repeated Steps 2-4 in 8-game intervals, from 16+ games all the way to 72+ games.
6. For each “games played” interval, I calculated
1. the number of games at which the YAC variance explained, R2, would mathematically equal 0.5.1
2. the “true” YAC for a hypothetical RB that’s had an observed performance of 9.00 YAC through that number of games.2
7. I calculated a weighted average of the results from Step 6.3

### Results and Discussion

Below is the stability table for RB YAC:

GamesnrR2 = 0.50Avg YACObs 9.00 YAC
Wtd Average477.948.47
45480.08467.718.36
83380.16427.888.44
122400.23417.988.49
161730.28428.058.52
201350.29498.168.58
24900.28628.258.63
28570.35538.268.63
32310.28838.498.74
36270.30838.518.76

As the “Wtd Average” row shows, RB YAC takes 47 games to stabilize, which translates to 80 receptions, 101 targets, and 613 routes run. Here’s how that compares to the RB receiving stats I’ve analyzed previously:

So the good news is that YAC isn’t the least reliable RB receiving stat. The bad news is that it’s third-worst, as well as the fact that two other yardage-related stats (i.e., YPRR and aDOT) stabilize in half as many games or fewer.

But let’s step back for a moment and put the various YAC stabilization points (i.e., games, receptions, targets, and routes) into proper perspective. Because statistics suggests that offensive football is a top-down game, with player performance relying heavily on schemes and play calls devised and implemented higher up the food chain, my reliability analyses look at player-team tenures (e.g., stats from Matt Forte’s tenure with the Bears are analyzed separately from stats from his tenure with the Jets). My data contains 935 such RB tenures, but only 128 — or 13.6 percent — of them involve at least 47 games, 80 receptions, 101 targets, or 613 routes run.4

What’s more, if we look only at active tenures heading into 2017, then only 18 RBs have amassed at least 47 games, 80 receptions, 101 targets, or 613 routes run with their current team. In alphabetical order, they are Le’Veon Bell, Giovani Bernard, Isaiah Crowell, Andre Ellington, Devonta Freeman, Jeremy Hill, Mark Ingram, Duke Johnson, Doug Martin, LeSean McCoy, Jerick McKinnon, Bilal Powell, Theo Riddick, Darren Sproles, Jonathan Stewart, Chris Thompson, James White, and T.J. Yeldon.

And just to be crystal clear as to why the above is important, remember that what I’ve just detailed are the base rate and current count of RB tenures for which Actual YAC represents at least 50% skill. In other words, for about 86 percent of all RB tenures and for all but 18 current RB tenures, the YAC we’ve observed to date is at least 50% luck — and a vast majority of the time a far higher percentage than that.

Speaking of which, after converting Actual YAC into True YAC for 2016, below are the standings for all RBs that played at least 25 percent of snaps per Pro Football Focus:

PlayerTmGRecActual YACRkTrue YACRk
Ezekiel ElliottDAL153212.329.11
Tevin ColemanATL133112.329.12
Spencer WareKC143311.759.03
Jordan HowardCHI152911.168.74
Jamize OlawaleOAK131214.318.75
Melvin GordonSD134110.288.66
Paul PerkinsNYG131511.948.57
Le'Veon BellPIT12759.1158.48
Zach ZennerDET111811.168.49
Matt ForteNYJ14309.4118.310
Tim HightowerNO15229.6108.211
Fozzy WhittakerCAR16259.4118.212
James WhiteNE16608.7208.213
Damien WilliamsMIA15239.4118.214
Derrick HenryTEN141310.198.115
Chris IvoryJAX11209.3148.116
Bobby RaineyNYG8209.1158.117
Isaiah CrowellCLV16408.5228.118
James StarksGB9198.9188.019
Latavius MurrayOAK14338.5228.020
Matt AsiataMIN16328.4268.022
Devontae BookerDEN16318.4268.023
Duke JohnsonCLV16538.2288.024
Darren SprolesPHI15528.2288.025
Todd GurleyLA16438.2288.026
Ryan MathewsPHI13138.6217.927
Jonathan C. StewartCAR1389177.928
Robert TurbinIND15268.2287.929
David A. JohnsonARZ16808327.930
Mike TolbertCAR15108.5227.931
LeSean McCoyBUF15508327.932
Wendell SmallwoodPHI1268.5227.933
Jeremy HillCIN15218327.934
Devonta FreemanATL16547.9357.935
Frank GoreIND16387.9357.936
Kenyan DrakeMIA1397.7387.838
Doug MartinTB8147.6397.839
Dwayne WashingtonDET11107.4417.840
Andy JanovichDEN756.4577.741
Jacquizz RodgersTB10137.2467.742
C.J. AndersonDEN7167.3437.743
Ty MontgomeryGB10347.5407.744
Jeremy LangfordCHI12197.3437.745
Jerome FeltonBUF1166.2607.746
LeGarrette BlountNE1676.1637.747
Thomas RawlsSEA9136.8537.748
DeAndre WashingtonOAK13177517.749
Terrance WestBLT16347.3437.750
Charcandrick WestKC15287.2467.751
T.J. YeldonJAX15507.4417.752
Rob KelleyWAS14126.3597.653
DeAngelo WilliamsPIT8186.7547.654
Jonathan GrimesHST8136.2607.655
Mark IngramNO16467.2467.656
Theo RiddickDET10537.2467.657
Bilal PowellNYJ16587.2467.658
Patrick DiMarcoATL974.4747.659
Mike GillisleeBUF1595727.560
Shaun DraughnSF14296.6567.561
Giovani BernardCIN10396.7547.562
DeMarco MurrayTEN16536.9527.563
Kenneth DixonBLT12306.2607.464
Christine MichaelSEA9205.6667.465
Aaron RipkowskiGB1093.4757.466
Benny CunninghamLA10165.1707.467
Jerick McKinnonMIN15436.4577.368
Lamar MillerHST14315.9657.369
Kyle JuszczykBLT16376.1637.370
Carlos HydeSF13275.5687.271
Jalen RichardOAK16295.6667.272
Alfred BlueHST14122.3777.173
Jay AjayiMIA15274.9737.174
John KuhnNO13163.2767.175
Chris ThompsonWAS16495.1706.877

In contrast to what we saw with True WR YAC vis-a-vis True WR aDOT, the top of the True RB YAC standings doesn’t exhibit a clear distinction between YAC-specializing, short-zone route runners and other types of receiving backs. That’s, of course, because all RBs are predominately YAC-specializing, short-zone route runners. This lack of distinction — coupled with RB aDOT being far more reliable — means you should primarily rely on aDOT when trying to predict a RB’s receiving future.

Nevertheless, for the sake of completeness, I’ll do the same thing I did for WRs, identifying players that meet the following criteria heading into 2017:

• Top-36 True aDOT in 2016
• Top-36 True YAC in 2016
• On the same team as 2016
• Top-3 on the current depth chart for 2017

The initial list contains 11 RBs, but 4 of them experienced a change at head coach and/or offensive coordinator (Tevin Coleman, Devonta Freeman, Devontae Booker, and Todd Gurley), and 2 of them saw competition brought into the fold (Spencer Ware and Jamize Olawale). That leaves five RBs who truly — pun intended — fit the bill (in True YAC order): Jordan Howard, James White, Damien Williams, Duke Johnson, and David Johnson.

### DT : IR :: TL : DR

The present reliability analysis tells us that a RB must log 47 games (or 80 receptions or 101 targets or 613 routes run) before his Actual Yards After Catch (YAC) reflects less than 50% luck. In terms of yardage stats for the position, while this is faster than (the awful) Yards per Target (YPT), it’s slower than Yards per Route Run (YPRR) and Average Depth of Target (aDOT). To put this in perspective, and based on my dataset, the probability that a given RB will log enough games (or receptions or targets or routes) with a given team to overcome YAC being mostly luck is a mere 13.6%.

00

1. The formula is (Games/2)*[(1-r)/r]

2. The formula is [(Observed Performance * Games) + (League-Average Performance * Stabilization Point)] / (Observations + Stabilization Point)

3. Weighted by group size.

4. I’m fully aware that the “true” percentage is almost certainly higher because some inactive RBs would have reached one of the thresholds if I had data from prior to 2006 and some active RBs will reach one of the thresholds in the coming years.