≡ Menu

Adam Steele on Negative Yards per Attempt

Adam Steele is back for another guest post. You can view all of Adam’s posts here. As always, we thank him for contributing.


On Monday, I updated my ever-evolving Positive Yards Per Attempt metric. Today’s post will serve as an introduction to its contra metric, Negative Yards Per Attempt (NegY/A). The very simple formula is as follows:

NegY/A = ( – sack yards – INT * 45) / dropbacks

The result will always be either zero or negative, but less negative (i.e., closer to zero) numbers are better. I chose to exclude fumbles because I want to maintain an apples to apples comparison with PY/A, so NegY/A covers passing plays only. I want to be very clear – NegY/A is NOT intended to be a comprehensive measure of QB play and should never be cited on its own. Its primary purpose, as with PY/A, is to estimate the relative importance of the different components of the passing game.

I won’t bore you with more words, so lets get straight to the numbers. Similar to the PY/A table, NegY/A is presented as both value over average and relative to league average on a per play basis. I wanted to cover the same timeframe as the previous article, so this includes all QB seasons since 1992 of at least 224 dropbacks (n = 829).

Brady was happy to see his ranking in NegY/A

We have a mixed bag of results. At the very top are a plethora of MVP caliber seasons; this makes sense because a true MVP must be dominant at moving the ball and adept at avoiding mistakes. However, as you scroll further down the leaderboard, you’ll start to see names that invoke mediocrity rather than greatness. Many of these dubious seasons were driven by either fluky low interception rates or purposefully conservative playing styles. If you click Value to invert the sorting, you’ll see that the worst NegY/A seasons follow a similar pattern. [1]And the two worst seasons came from the same city in the same year. There are a handful of truly awful QB’s who consistently hurt their teams with mistakes, but there are also a number of serviceable and even HOF-caliber players lurking at the bottom of the NegY/A pile. The bottom 20 seasons in Value led to a not-so-bad 0.44 win percentage, which goes to show that there is far more to winning than merely avoiding bad plays.

As previously mentioned, the correlation between Relative PY/A and wins is 0.56. The correlation between Relative NegY/A and wins is 0.41. While the strength of those correlations appear fairly close at first glance, this is a case where correlation does not imply causation. As most of you probably know already, game context has a massive effect on INT%, which in turn has a massive effect on NegY/A. Quarterbacks throw far more picks when trailing late in the game than they do while leading. Thus, interceptions are often a byproduct of losing rather than the cause of losing. In the table below, I’ve parsed out the components of both metrics to show how much (or little) they change depending on game script. To keep it simple, I looked at teams leading in the fourth quarter and compared them to teams trailing in the final period. The data is from the last five seasons, but I feel confident that the general trend would hold going back several decades. The most important column is Effect, which is the proportional difference between the higher average and the lower.

Every aspect of passing offense is better while leading, except for Sack% which declines slightly. This jives with common sense, as leading teams can choose to pass under optimal circumstances and don’t have to take many risks. However, aside from TD%, the benefit of leading is relatively small. None of the other components shift by even 10%. The gigantic outlier is INT%, as trailing teams throw 81% more picks than leading teams. That is, of course, a huge difference, and it’s far greater than the variance among QB’s in “true” proclivity for tossing it to the wrong team. Despite the slight boost to Sack% for trailing teams, that bloated INT% warps their average NegY/A. In fact, NegY/A is 53% better among front runners than chasers. Meanwhile, PY/A dips only 9% among trailing teams. This data prompts me to conclude that PY/A is relatively situation-neutral, therefore lending legitimacy to its fairly strong correlation with winning. On the other hand, NegY/A is highly situation-dependent, which leads me to believe that the causation arrows is backwards when it comes to the correlation with winning. Just as teams rack up rushing yards after they’ve already salted the game away, quarterbacks pile up interceptions when their team has little hope of winning.

For a more visual and colorful representation of the first table, I’ve included some more bubble graphs. Graph #1 plots Relative NegY/A against Win %, Graph #2 shows NegY/A Value compared to Win %, Graph #3 reveals each quarterbacks’ progression of career Value over time, and Graph #4 displays career Relative NegY/A compared to Win %.

As always, please leave any thoughts or contributions in the comments.

References

References
1 And the two worst seasons came from the same city in the same year.
{ 3 comments }