When Do Receptions and Touchdowns per Route Run Stabilize for Tight Ends?

Through Week 10, the Top 3 tight ends (TEs) in Receptions per Route Run (RPRR) were Rob Gronkowski (20.5%), Jimmy Graham (20.1%), and Travis Kelce (20.0%). Gronkowski (3.35%) also ranked third in Touchdowns per Route Run (TDPRR), behind only Julius Thomas (4.35%) and Antonio Gates (3.47%). ((Routes run stats provided by Pro Football Focus.)) In that same time frame, the league averages for these stats were 11.8% RPRR and 1.22% TDPRR. Given this context, which Top 3 so far has the least sustainable outlying performance? For that, we need to answer the question, “How long does it take before RPRR and TDPRR stabilize for TEs?“

Back when I ran this analysis for wide receivers (WRs), I found that RPRR stabilizes in slightly less than one-fifth of the time it takes for TDPRR to stabilize. However, WRs are smaller, faster, and more elusive than TEs in general, so the positions tend to require different skill sets and experience different usage patterns. In short, one can reasonably argue that, because TEs see a higher proportion of their targets in the red zone than WRs, TDPRR may very well be a better “skill” indicator than for WRs, while RPRR may very well be a worse “skill” indicator than for WRs.

Let’s find out.

Methods

Hate to be a broken record, but it’s important:

1. I collected data for all TEs that had at least 8 games played in PFF’s historical database, which currently runs from 2007 to 2013.
2. To control for team effects, I included only those TEs that played 8+ games for the same team.
3. Starting with TEs that played 8+ games, I randomly selected two sets of 4 games for each WR, and calculated their RPRR and TDPRR in both sets.
4. For both of these metrics, I calculated its split-half correlation (r) between the two randomly-selected sets of games.
5. I performed 25 iterations of Step 4 so that r converged.
6. I repeated Steps 3-5, increasing the TE inclusion criteria in 8-game intervals, from 16+ games all the way to 72+ games.
7. For each “games played” group, I calculated the number of games at which the variance explained in each metric, R², would mathematically equal 0.5. ((The formula is (Observations/2)*[(1-r)/r].))
8. I calculated the “true” RPRR and TDPRR for a hypothetical TE that’s had an observed performance of 15.0% RPRR and 1.25% TDPRR through X number of games. ((The formula is [(Observed Performance * Observations) + (League-Average Performance * Stabilization Point)] / (Observations + Stabilization Point) ))
9. I calculated a weighted average of the results from Steps 7 and 8. ((Weighted by group size.))

Results

Below is the table for RPRR:

Games	n	r	R2 = 0.50	Avg RPRR	Obs 15.0% RPRR
4	296	0.18	19	12.1%	12.6%
8	200	0.34	15	12.2%	13.1%
12	136	0.48	13	12.4%	13.6%
16	98	0.59	11	12.6%	14.0%
20	67	0.65	9	12.6%	14.2%
24	48	0.72	7	12.8%	14.4%
28	35	0.80	8	12.9%	14.6%
32	22	0.81	14	12.9%	14.6%
36	12	0.72	14	13.3%	14.5%
Wtd Average			14	12.3%	13.7%

New readers can click here for an explanation of how to read the table, so let’s proceed directly to the “Wtd Average” row, which tells us that RPRR takes 14 games to stabilize. The average TE in my sample ran 18.2 routes per game, so 14 games translates to 262 routes run. Using this information alongside the sample’s weighted average RPRR of 12.3%, we can determine that a TE with 15.0% RPRR after 14 games (or 262 routes run) has a True RPRR of 13.7%, which is the midpoint between their observed performance and the league average.

And now, here’s how the same analysis played out for TDPRR:

Games	n	r	R2 = 0.50	Avg TDPRR	Obs 1.25% TDPRR
4	296	0.18	19	1.02%	1.06%
8	200	0.13	54	1.03%	1.06%
12	136	0.17	58	1.04%	1.08%
16	98	0.29	39	1.08%	1.13%
20	67	0.30	46	1.09%	1.14%
24	48	0.48	26	1.13%	1.19%
28	35	0.38	46	1.13%	1.18%
32	22	0.25	96	1.12%	1.15%
36	12	0.29	89	1.20%	1.22%
Wtd Average			41	1.05%	1.15%

The “Wtd Average” row says  TDPRR takes 41 games (or 740 routes run) to stabilize. It also says that a TE’s 1.25% observed TDPRR at that point suggests a True TDPRR of 1.15%.

Discussion

Before providing my thoughts about the bottom-line results, it’s worth highlighting a couple of results that might fly below your radar otherwise. First, the split-specific sample sizes suggest ((Now that’s alliteration, folks!)) that a TE’s tenure with the same team is far shorter than for WRs: Whereas 29 WRs lasted at least 72 games (i.e., “n” in the “36+” row), only 12 TEs achieved that amount of longevity from 2007 to 2013.

Second, the 1.05% weighted average TDPRR for TEs is, practically speaking, identical to the 1.04% for WRs. That said, except for the “32+” group, TEs averaged a higher TDPRR than WRs at every team-specific career length, which is a secondary finding that lends some credence to the skill-set/usage distinction I hinted at in the intro.

Pivoting to the main event, it turns out that, while RPRR takes twice as many games to stabilize for TEs than it does for WRs (14 vs. 7), the proportional gap is far smaller for TDPRR: 41 games for TEs, as compared to 33 games for WRs. Furthermore, given their relative disparity in routes run per game (18.2 vs. 27.0), an average TE’s TDPRR stabilizes in fewer routes run than an average WR’s (740  vs. 882), and the proportional gap for RPRR isn’t as large as 2-games-to-1 (262 vs. 188).

DT : IR :: TL :DR

Of Rob Gronkowski’s league-leading 20.5% RPRR and his third-ranked 3.50% TDPRR, which is more of an indicator of his “true” TE ability? The short answer — as it was for WRs — is his RPRR: It takes 262 routes run for a TE’s RPRR to stabilize, whereas it takes 740 routes run for his TDPRR to stabilize. The long answer, however, is that Gronkowski’s RPRR is only slightly less sustainable than Antonio Brown’s league-leading 19.6% RPRR for WRs and more sustainable than Martavis Bryant’s league-leading 7.06% TDPRR. This is a difference with a distinction because of the disparate theoretical skill sets required by the two positions.