## Research Challenge: Reach and Results

#### Who Wins by Which Method

I see a lot of work has already been done on breaking down the data statistically.  This post will concentrate on the question: "Given reach data alone, what can we expect from a fight?"  One prominent (and obvious) conclusion is that Long guys tend to win/lose with different methods and frequencies than Short guys.

Here's a table with some win-distribution info.  To keep things simple, along one axis are just two reach categories:  Long reach (if one guy has a longer reach than the other) and Short reach.  The other axis is the method by which the fight ended: Decision, TKO, or Submission.  For each fight where there was a difference in reach, these are the probabilities it ended in a particular way:

 Win Freqs:  Reach x Method Method Long wins Short wins sum diff Decision 17.4% 20.0% 37.4% -2.6% TKO 19.1% 15.0% 34.1% 4.1% Submission 15.3% 13.1% 28.5% 2.2% Total 51.8% 48.2% 100.0% 3.7%

So Long has a 52%-48% overall advantage (+3.7%) over Short to win, but the single most probable outcome (20.0%) is for Short to win a decision.  Long has advantages to TKO and Sub, and a disadvantage to decisions.  Here are the most likely outcomes, in order:

 Winner Method Prob Short Decision 20.0% Long TKO 19.1% Long Decision 17.4% Long Submission 15.3% Short TKO 15.0% Short Submission 13.1%

I haven't calculated the confidence intervals, just stating the frequencies of outcomes.  The fighters' reaches are equal in 10.6% of the fights.  Here are the frequencies at which fights end by various methods, given the winner's reach is Long, Short or Equal:

 Method Freq: given Reach-Winner Method Long Wins Equal Reach Short Wins Decision 33.6% 46.5% 41.5% TKO 36.8% 24.8% 31.2% Submission 29.5% 28.7% 27.3% Total 100.0% 100.0% 100.0%

Onto the reach-difference distributions {(R_win_i - R_lose_i)} for fights ending by a particular method:

 Method Mean(inch) Stdev(inch) Decision -0.07 3.26 TKO 0.58 3.48 Submission 0.28 3.43

There is also a lot of suggestive stuff in the specific methods of submissions.  It looks like Long guys are susceptible to ankle locks, for example.  I see that others have gone into that in much greater depth.

To take this further, I'd probably want to increase the number of reach categories, but I'm not sure what the next step might be, statistically.  One approach to quantifying a reach advantage is fitting trendlines, as I've seen others do.  Another approach  might be to compare the relative probabilities of a player's reach difference including him in each method's population (from the distributions above).

Hope this has been worth reading!  Thanks.

(First post!)

\The FanPosts are solely the subjective opinions of Bloody Elbow readers and do not necessarily reflect the views of Bloody Elbow editors or staff.

## Trending Discussions

forgot?

As part of the new SB Nation launch, prior users will need to choose a permanent username, along with a new password.

I already have a Vox Media account!

### Verify Vox Media account

As part of the new SB Nation launch, prior MT authors will need to choose a new username and password.

We'll email you a reset link.

Try another email?

### Almost done,

By becoming a registered user, you are also agreeing to our Terms and confirming that you have read our Privacy Policy.

### Join Bloody Elbow

You must be a member of Bloody Elbow to participate.

We have our own Community Guidelines at Bloody Elbow. You should read them.

### Join Bloody Elbow

You must be a member of Bloody Elbow to participate.

We have our own Community Guidelines at Bloody Elbow. You should read them.