FanPost

Research Challenge - Better Later than Never

Any Chance for Partial Credit?

 

While I’m fully aware that this entry is much too late to be considered for promotion to the main page, I had originally planned to put this together in a timely fashion.  Unfortunately, other obligations got in the way and pushed this back.  So hopefully others will gain something from my thoughts and I can receive some feedback from my peers.

To lead off, I’m sure there are many more talented mathematicians than myself taking part in this challenge and perhaps they can draw some keen insight that I have clumsily overlooked or point out areas where my conclusions lack sufficient basis; however in my experience as a day-time data analyst I’ve found that the most useful analysis isn’t always the most mathematically rigorous (though a certain degree is always necessary) but rather those that take on uncommon perspectives and present the data in the most digestible fashion.  For this reason I’m going to forgo most math and statistical jargon and try to present my data graphically.

 

The question that has been put in front of us for analysis is whether or not reach matters in MMA.  This is an extremely open ended question that could take many   I think a few of the posts preceeding entries in this contest have adequately answered that there is a mathematical certainty that reach advantage does show a strong correlation with win percentage.  I have little to add here but some graphical representation of what has already been proven. 

Victorybyreachadvantage_medium

 

Data

 

Reach Advantage

Matches

Sum of Win%

<-6

30

34.1%

-6--4

89

44.9%

-4--2

193

47.1%

-2-0

329

52.0%

0-2

355

49.7%

2-4

258

49.1%

4-6

174

55.1%

>6

58

65.9%

Grand Total

1486

50.0%

 

 

What could be added here is that not only does reach advantage matter, but the nominal reach (meaning, the reach of a fighter regardless of the reach of his opponent) sheds some light on a few insights and may get away from the standard view of the research question of whether being lengthier than your opponent is advantageous.

Finishbynominalreach_medium

 

Data

 

Reach

# of Matches

             Finish%

<67

127

43.3%

67-69

239

46.4%

69-71

499

57.7%

71-73

571

59.7%

73-75

625

61.6%

75-77

499

64.7%

77-79

305

68.2%

>79

107

68.2%

Grand Total

2972

60.0%

 

 

We can plainly see that finish rate of a fighter increases significantly and dramatically along with the nominal reach of a fighter.  While this is interesting, it has been previously touched on and is slightly inadequate.  All finishes are not created equal and we should not be content to presume that they are.  Fortunately the data provided allows us to go a few layers deeper.  For the below graphic I excluded submission due to injury.  This is the distribution of method of victory graphed along nominal reach.

Victorysmethodnominalreach_medium

Sum of Match

Category

 

 

 

Reach

Decision

Sub

KO/TKO

Grand Total

<67

38

14

7

59

67-69

66

30

19

115

69-71

101

74

70

245

71-73

103

84

79

266

73-75

116

87

105

308

75-77

77

75

92

244

77-79

54

40

68

162

>79

14

11

30

55

Grand Total

569

415

470

1454

 

 

We’ve already established that rangier fighters end fights in decisions at a slower clip.  The extra wrinkle that we can add here is that those lost decisions are being made up in KO/TKOs but not submissions.  The % of victories that occur due to submission does not show a strong relationship with the fighter’s reach while % of victories from TKO/KO and Decision (though in the opposite direction) obviously do.

The logical question here is why?  I have two theories, neither of which can be tested using the data provided. 

<!--[if !supportLists]-->-          -  <!--[endif]-->The first theory is that the lengthier fighters are more likely in larger weight classes and carry more power and less stamina to defend themselves for the entire duration of a bout.  This could be easily verified by looking at the method of victory spread across weight classes.  My hypothesis would be that as you progress up through the weight classes you will see that decisions are less frequent, KO/TKO are more frequent, and submission wins hold relatively constant.

<!--[if !supportLists]-->-          -  <!--[endif]-->Next would be theory that fighting strategies and single discipline backgrounds vary across fighters of different lengths.  Do fighters from wrestling backgrounds tend to be stockier than those from striking backgrounds?  Do lengthier fighters tend to sprawl and brawl to use their reach rather than work offensive wrestling in search of a takedown?  This is harder to answer and could only be addressed using in-fight data such as the rate of takedown attempts performed by fighters of different reaches.

In closing I’m extremely excited about the premise of this challenge and only wished that I was able to submit a more timely entry into the discussion.

 

PS – I have one more graph that I found interesting but couldn’t think of how to work it into the main portion of the entry.  I grouped all submissions together into smaller groups to look at whether longer guys were in more danger of being caught in an armbar or leglock.  I didn’t see anything to suggest that was true, but what I did find surprised me.  Simply put, when longer fighters are submitted they are much less likely to get choked.  

Losersubspread_medium

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

X
Log In Sign Up

forgot?
Log In Sign Up

Forgot password?

We'll email you a reset link.

If you signed up using a 3rd party account like Facebook or Twitter, please login with it instead.

Forgot password?

Try another email?

Almost done,

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.

Spinner

Authenticating

Great!

Choose an available username to complete sign up.

In order to provide our users with a better overall experience, we ask for more information from Facebook when using it to login so that we can learn more about our audience and provide you with the best possible experience. We do not store specific user data and the sharing of it is not required to login with Facebook.

tracking_pixel_5349_tracker