LA Rams defensive tackle Aaron Donald described as ‘just average run-defender’ starts a Twitter tornado
The LA Rams are proud of the effort of defensive lineman Aaron Donald. Why not? They have a right to be, as Donald has been one of the most effective defensive linemen of the modern NFL. Well, it depends on who you talk to, I suppose. You know the saying about liars. There are liars, damned liars, and statisticians.
Using video evidence is not something that should be performed casually. In a football game, each defender is given a specific assignment. Sometimes the player wins the battle, sometimes a player loses the battle. But the third outcome, where a player redirects an offensive play into the path of his teammates, is another outcome that most defensive coordinators would be quite happy with.
He is ‘Just averager for a defensive tackle’?
But not in this analysis. To be fair, the analysis did not seem to be supported by hard data, frequency tables, or the like. When asked if the data was weighted for double-teaming, the response was that he was not double-teamed particularly often.
Aaron Donald is just average for a defensive tackle in run stop win rate. I took a look at our new metric’s most surprising result: https://t.co/Q7ihSYOLYD
— Seth Walder (@SethWalder) September 8, 2020
Curiously, 2019 data recorded that Donald was double-teamed 319 occasions. That is against 926 defensive snaps. Later in the dialogue, that count was named as 410 double teams against 926 defensive snaps. Aaron Donald was double-teamed over 44 percent of his defensive snaps. I do not possess a Ph.D. in statistical analysis, but that seems to be a fair amount more than “not particularly often”.
Stats say what you want them to say
Statistical analysis is not cherry-picking to prove out a foregone conclusion. The use of video evidence to support a controversial conclusion is not evidence at all. It’s an edited version of “see here it is”, which is the basis of TMZ reporting. Statistics must use and rely upon gross data organized into data tables, comparing discovered values to the normal values. For example, here Pro Football Focus cites Aaron Donald’s performance over the past two years as the best in the NFL
Well, the now-viral opinion that Aaron Donald is “just average” against the run seems to be an opinion that is hard to justify. After all, the conclusion that Rams defensive lineman is “just average” infers that there are numerous defenders who are better than Donald against the run. I would be hard-pressed to name them.
Statistics measure nothing if they do not measure results. In football, that is sacks, tackles, and tackles for losses for a defensive line position. What about TFLs?
Most Tackles For Loss – Last 4 Seasons
Aaron Donald 77
Chandler Jones 67
Cameron Jordan 67
Jadeveon Clowney 60 pic.twitter.com/BTci6khIAP
— CBS Sports HQ (@CBSSportsHQ) September 8, 2020
In the end, there is a new metric and it needs to be reworked. Aaron Donald is not an average run defender.
I’m not trying to take anyone to task here. But if I try to sell a cure-all elixir, then I need to have clinical trials and proof of any claims that I make. That also applies to statistics. Aaron Donald has recorded the second-most number of sacks since 2015 with 63.0. He has recorded 312 tackles in his career.
I understand the concept. The analytics have led to the conclusion that Donald penetrates too far into the offense, allowing running backs to rush in his wake. The conclusion is that a savvy offense can set up mini-draw plays to rush the lane where Donald used to stand. Perhaps the most hilarious response on social media occurred when Donald’s former defensive coordinator Wade Phillips weighed in on ESPN’s analytics scorecard.
But as we know, Aaron Donald works his best magic when he blows up plays in the backfield. Those runs for positive yardage do happen, yes. But there are the tackles for losses as well. So measure away. I feel confident that however measures are taken, NFL coaches will want Aaron Donald’s version of defense over other alternatives. Average or not.