In response to the recent post about the comparative success of short and deep passes a few readers asked for a break out of the results by down. Some suggested the advantage of deep passes might only due to unsuccessful check downs on 3rd down, which is certainly plausible. The numbers are going to be slightly different here because I neglected to exclude red zone attempts in the original post. (Please read the original post for definitions and caveats.)
Here is the break out by down. It appears that the advantage exists on all downs, and the advantage is not significantly greater for 3rd downs than for 1st or 2nd downs. But that doesn't necessarily rule out the check-down effect, as they are a common tactic on all downs.
Down | 1 | 2 | 3 | 4 |
Deep | 0.43 | 0.50 | 0.53 | 0.71 |
Short | 0.10 | 0.09 | 0.12 | 0.59 |
Difference | 0.33 | 0.41 | 0.41 | 0.12 |
But not all 3rd downs are the same. Here is the same break out of EPA by down, but limited only to 10 yards to go. This isolates the 3rd and long situation from other 3rd down attempts, but just because the to-go distance is the same doesn't mean the situations are the same. The difference between deep and short attempts is larger on 3rd and 10, but it's nearly as large on 2nd and 10.
Down | 1 | 2 | 3 | 4 |
Deep | 0.43 | 0.60 | 0.63 | -1.26 |
Short | 0.10 | 0.14 | 0.16 | 1.65 |
Difference | 0.32 | 0.46 | 0.48 | -2.91 |
And again for only 5 yards to go. Keep in mind 1st and 5 situations are relatively scarce. For that matter, fourth down attempts for all to go distances are relatively scarce too, as the data are restricted to 'normal' situations--1st and 3rd periods with the score within 10 points.
Down | 1 | 2 | 3 | 4 |
Deep | 0.47 | 0.25 | 0.62 | -0.65 |
Short | -0.14 | 0.08 | 0.15 | 0.25 |
Difference | 0.61 | 0.16 | 0.48 | -0.90 |
Brian, with the proliferation of the "West-Coast offense", could you run the SUCCESS RATE for these numbers? (Maybe you did in the first post, & I forgot.) Also--are these ALL passes, or just completions?
I read your site and Football Outsiders. I've always felt FO undervalued the big play. I think these two articles on pass depth lend support to that belief.
When I played APBA football as a kid I figured out really quickly that deep passing was the key to winning my league. Given that the game was based on the stats generated in a given NFL season it seems I simply stumbled on to the apparently statistical advantage of deep throws.
I like the analysis, but it seems intuitive that long passes yield a higher EPA. As you said in your previous post, surely every team would rather throw deep than short... my guess is that QBs generally check the coverage and throw deep if it's available, and if it's not there, they throw short. I'd hesitate to draw the conclusion that teams should pass deep more, because obviously the more that happens the more predictable and less effective it becomes.
How are throwaways denoted in the pbp data? Does that count as an "aborted" play or is it an incomplete "short pass"?
Forgive me if this is naive, but why should we expect there to be a difference depending on the down? Shouldn't your performance on all downs be about equal in the long run?
There was suspicion that short passes on third and long, usually ending the drive, would significantly affect the overall EPA of short passes.
James, you're right, sorry for the brain fart. We're talking about EPA of short passes, NOT a stat like completion percentage, etc.