At a recent Chip Kelly press conference, Kelly noted that he liked the 3-4 better because of the philosophical advantage of having more linebackers on the roster than defensive linemen, for the purpose of special teams:
“What direction we end up ultimately heading in, I like the 3-4 better. When I first started at Oregon, I think from a special teams standpoint philosophically, if you carry more linebackers on your roster than you do defensive linemen, you help your team from a special teams standpoint, but you just didn’t do that in a day.”
The logic makes sense. Linebackers are typically going to be faster than defensive linemen, and more likely to tackle a speedy returner in the open field. I wondered how the logic actually matched up with the statistics, and so, I aggregated all the kick and punt coverage numbers for each team over the last 5 years, and separated them by 4-3 teams and 3-4 teams. I omitted 4 teams (the Patriots, Saints, Seahawks, and Cardinals), who either run some kind of hybrid defense, or just had some other gray area in terms of their base defense.
First, kick coverage. Over the last 5 years, 3-4 teams held a slight advantage in terms of yards per return, but gave up the home run with significantly more frequency:
Punt coverage was the exact reverse of kick returns. 4-3 teams were better in terms of yards per attempt, but gave up the home run with significantly more frequency:
I also did a separate metric for teams that switched from a 4-3 to a 3-4 or vice versa over the last 5 years. Those teams, by my count, include the Browns, Redskins, Dolphins, Packers, Bills, Colts, Texans, Chiefs, and Broncos. Below are the numbers for each of those teams combined when they were running a 4-3, and when they ran a 3-4.
First, kick coverage. This was the same as the overall kick coverage metric. 4-3 teams gave up slightly more yards per return, but gave up TDs with less frequency:
For punt coverage, again, same as above. 4-3 teams gave up fewer yards per return, but a higher frequency of TDs. The interesting thing for teams that changed their base defense is that their numbers are worse for both kick and punt returns than the teams that did not change their base defense:
So what do the above numbers tell us about whether or not 3-4 is better than a 4-3 in terms of special teams coverage? Logically, it makes sense, but statistically, I’m not sure a clear argument can be made for either side.