Research Challenge -- the low-hanging fruit

Terrence Chan — Tue, 08 Mar 2011 18:45:03 -0000

I love spreadsheets. I also love doing stuff that makes me look like I know things about statistics, even though I'd probably fail first-year college stats right now if you put the exam in front of me. So the FightMetric Research Challenge is pretty much right up my alley. So, here's my attempt to make some sense out of the raw data from these 1486 fights.

The average winner has a reach advantage of .23755 inches. Or 0.33% of total length. The average reach of all fighters in the data is 72.778 inches. The longer fighter has a record of 688-641, or wins 51.8% of the time.

When there is a difference of five inches or more, the longer fighter is 150-106 (58.6%).

The average winner has a reach advantage of .23755 inches. Or 0.33% of total length. The average reach of all fighters in the data is 72.778 inches. The longer fighter has a record of 688-641, or wins 51.8% of the time.

When there is a difference of five inches or more, the longer fighter is 150-106 (58.6%).

The average length of a fighter who wins by KO/TKO (referee or doctor stoppage) is 0.58" longer than that of his opponent. The longer fighter is 253-199 (56.0%) to win. Looking at fights where the reach differential is 5 inches or more, the longer fighter holds an impressive 62-28 (68.9%) lead in the (T)KO department. We can draw a significant conclusion that being longer correlates to winning by knockout.

The average fighter who wins by submission, however, is only 0.27" longer (at this point my use of the word “longer” without identifying the context is starting to make my inner 8-year-old snicker). The longer fighter holds a record of 203-174 (53.8%) when someone taps. Fighters with a five inch or greater reach however are 44-29 (60.2%) in the submission department, so if you’re a submission grappler, it does seem helpful to be much longer than your opponent, but not of much help to be just a little longer. More on the submission stuff below.

In a perhaps surprising result, the shorter (er, less-lengthy?) fighters are more likely to win a decision. The longer fighters here are 231-265 (46.6%). Maybe the shorter fighters tend more to be wrestlers who are getting on top and grinding out more decision wins. Or it could be the other way around: fighters who know they are giving up reach realize they need to get inside and take the fight to the ground.

A further breakdown of the submission data

While we know that in the standup department longer fighters are superior with ranged attacks like jabs, crosses and roundhouse kicks and shorter fighters will utilize more powerful hooks, uppercuts and elbows on the inside, KO data is not typically broken down in MMA. But happily, we do get submission breakdowns so we can test the theories of whether certain body types are better at certain submissions. So, grappling nerds, let's take a look at the sub data.

When you think about guys with long limbs, you think about triangles. And the data bear this out. The longer guy has triangled the shorter guy 28 times, while the shorter guy has only wrapped his legs around the longer guy’s neck 17 times (62.2%). The reach advantage for the typical winner by triangle is a full inch (1.05" to be precise).
At the other end, submissions that tend to be associated with shorter, thicker guys are the arm triangle and guillotine. The arm triangle holds true, with the shorter guys holding an 18-13 (58.1%) advantage. But the guillotine is still slightly the realm of the longer guy, by a margin of 44-37 (54.3%).
The good ol’ rear-naked is of course the most common submission in MMA, if not all of grappling, and here the long guys still hold the edge, 63-49 (56.3%).
Shorter guys have a tremendous record of 13-1 in the leglock (ankle lock and kneebar are the only lower-body submissions listed) game, but I think just tells you that Rousimar Palhares is shorter than everyone he’s fought.

So it does seem that longer is a little better. On the other hand, it is more important how you use it. Or that's what we'll keep telling ourselves, anyway.

SBNation.com User Blog: Terrence Chan

Research Challenge -- the low-hanging fruit

A further breakdown of the submission data