Both Chase Stuart and I have shown that per-route metrics are more reliable indicators of a wide receiver’s (WR) “true” yardage ability than per-target metrics (or per-reception metrics, for that matter). With respect to my research, I found that Yards per Route Run (YPRR) stabilizes after 14 games, or about 350 routes run, while Yards per Target (YPT) takes nearly three times as long. With this in mind, I endeavored to find out how long it takes for two other fantasy football-relevant WR stats to stabilize: Receptions per Route Run (RPRR) and Touchdowns per Route Run (TDPRR).
Once again, here’s a brief outline of how this reliability analysis works:
- I collected data for all WRs that had at least 16 games played in PFF’s historical database, which currently runs from 2007 to 2013.
- To control for team effects, I included only those WRs that played 16+ games for the same team.
- Starting with WRs that played 16+ games, I randomly selected two sets of 8 games for each WR, and calculated their RPRR and TDPRR in both sets.
- For both of these metrics, I calculated its split-half correlation (r) between the two randomly-selected sets of games.
- I performed 25 iterations of Step 4 so that r converged.
- I repeated Steps 3-5, increasing the WR inclusion criteria in 8-game intervals, from 24+ games all the way to 80+ games.
- For each “games played” group, I calculated the number of games at which the variance explained in each metric, R2, would mathematically equal 0.5.1
- I calculated the “true” RPRR and TDPRR for a hypothetical WR that’s had an observed performance of 15.0% RPRR and 1.50% TDPRR through X number of games.2
- I calculated a weighted average of the results from Steps 7 and 8.3
Let’s begin with RPRR:
|Games||n||r||R2 = 0.50||Avg RPRR||Obs 15.0% RPRR|
Here’s an example of how to read the table. There were 223 WRs that had (at least) two sets of 12 games for the same team, and those WRs averaged 12.3 receptions for every 100 routes run. The split-half correlation for this group was 0.66, which translates to a stabilization point of 6 games. Armed with this information, we can estimate that a WR with 15.0% RPRR after 12 games actually has a “true” RPRR of 13.6%.
The “Wtd Average” row combines the results of all nine groups, and tells us that it takes 7 games for RPRR to stabilize. Using the weighted average number of routes run per game (27.0), we can estimate that 7 games translates to 188 routes run. The “Wtd Average” row also tells us that a WR with 15.0% RPRR after 7 games has a “true” RPRR of 13.7%, which is the exact midpoint between his observed performance and that of a league-average WR (12.5% RPRR).
Moving on, below is the results table for TDPRR:
|Games||n||r||R2 = 0.50||Avg TDPRR||Obs 1.50% TDPRR|
The bottom line here is that TDPRR takes 33 games to stabilize, which translates to 882 routes run. And again, a WR with 1.50% TDPRR after 33 games actually has a “true” TDPRR of 1.27%, which is 50% of the distance between observed performance and league-average performance.
A couple of things jump out at me about these results. First, you might recall from my previous article that Targets per Route Run (TPRR) stabilized in 7 games. Well, now we come to find out that RPRR also stabilizes in just 7 games. This suggests that targets are no more reliable than receptions once we control for routes run. To me, that’s another indication that target data isn’t the great beacon of information that it’s come to be perceived in recent years.
Second, the 33-game result for TDPRR suggest that it takes over twice as long as the aforementioned 14 games for YPRR (or 882 vs. 351 routes run). This mimics recent research I did on QBs, which showed that it takes twice as many throws for touchdown rate to stabilize than does yards per attempt (1,059 vs. 396 attempts). I don’t know if there’s any hidden wisdom in the similarities between these two sets of results, but at the very least it further supports the idea that, in the passing game, yardage is more skill-based than touchdowns.
DT : IR :: TL : DR
Running with the idea that the number of routes run by a WR provides more reliable information about his “true” ability than does his number of targets, I calculated the stabilization points for RPRR and TDPRR. It turns out that the former is half-skill/half-luck after 7 games, or 188 routes run, while the latter is half-skill/half-luck after 33 games, or 882 routes run. These findings further support previous research showing that (a) targets data isn’t as valuable as you think, and (b) yardage performance is “stickier” than touchdown performance.