Is Reach Difference a Clear Factor In Who Wins?

Taking a look at reach differences in MMA fights and whether it appears related to fight outcome.

Introduction:

When looking at the “Tale Of The Tape” in MMA fights, you’ll see a few familiar statistics: the fighters’ ages, heights, weights, and reaches. Aside from weight—which is controlled for with weight classes—these graphics beg the question of if a quick look at these statistics tells you anything about who may win the fight. In other words, do these numbers even matter?

In this article, I will take a cursory look at one of these statistics, reach, which is a measure of a fighter’s wingspan. The general consensus is that fighters with longer reach have, all else being equal, an advantage in striking. These fighters can hit an opponent from the outside with a lower chance of being hit themselves.

I performed an exploratory analysis using Python with a fantastic historical UFC dataset found here.

Findings:

First, let’s take a look at typical differences in reach in past fights. Below is a histogram that shows the differences in reach across all UFC fights:

The average difference in reach between fighters is 6.5 centimeters (or ~2.5 inches). The histogram above shows that the vast majority of fights (around 75%) are held with reach differences under 10 centimeters (or ~4 inches). How does this differ across weight classes?

Differences in reach for non-title bouts are fairly consistent across weight classes. On the other hand, there seems to be far more variability for title bouts. Additionally, title fights have a lower average difference in reach. One notable exception is the reach differences in the Light Heavyweight division, due in large part to Jon Jone’s long title reign and his massive wingspan (84.5 inches or ~215 cm)! So, now that we know that there are sizable reach differences across weight classes on average, let’s take a look below at how often the fighter with the longer reach wins:

At first glance, it seems that the title fight column is giving us more noise than signal. Because there are far fewer title fights than non-title fights, our sample size is smaller. And, because some fighters have had long title reigns, how their reach stacks up against their opponents skews the data further. The column on the right seems to suggest that fighters with longer reaches are actually at a slight disadvantage. However, I’m skeptical that the difference in success is solely due to differences in reach. In the future, it could be interesting to control for fighting style and other factors to see if these rates change significantly.

Thoughts and Further Work:

This (very) quick exploratory analysis lends some support to the idea that differences in reach between the fighters can impact the fight, but not in the direction expected. To investigate this further, I could remove outliers—for example, Jon Jones, as a successful fighter that always has a reach advantage—and control for other factors. I could also build a Machine Learning model to analyze the feature importances of the variables used to build the model. This would tell us how important differences in reach are relative to other factors like age, record, and fighting style, in predicting who wins a given fight.