FightMetric: Where MMA Meets Statistics

Fight Metric: Where MMA Meets Statistics

by Mark DellaGrotte

The UFC is the world’s premier mixed martial arts organization. Its events are broadcasted worldwide, drawing thousands of fans to watch fighters compete in combat sports such as boxing, kickboxing and jiu-jitsu.

There have been many changes over the years in how these fights are organized and presented, but one thing remains constant – the fight itself.

UFC events feature two or three rounds of action, with the winner determined by knockout, submission or decision. While there are several ways to measure performance, one way is through statistics.

These include things like takedowns landed per minute (TK/min), strikes landed per minute (ST/min) and time spent striking per round (CT/round). Other metrics include number of significant strikes landed per minute (SC/min), average strike speed (AS/mph) and total minutes fought (MCH).

These statistics provide insight into the skill level of each fighter, which is useful when comparing them to one another. However, they do not tell us much about the actual outcome of a bout.

What happens after a punch lands? How long does it take for someone to fall down? How long do they stay down? How many punches are needed to cause a fighter to fall? Do fighters with more endurance take longer for punches to affect them?

These questions and more can be answered through data analytics.

Data Analytics

FightMetric was created in 2007 to provide new statistics and analytics of fighters based on official UFC events. The organization’s goal is to collect and present MMA data in a meaningful way, ultimately providing a better understanding of the sport.

The system was implemented into the UFC’s official scoring criteria in 2012. Officials collect data from each fight and submit it to the FightMetric system, where it is then added to the database.

How It’s Done

FightMetric: Where MMA Meets Statistics - Picture

The system uses scales of measurement to analyze each fighter’s style, ability and result of each fight. After a match is complete, each judge scores each fighter on a ten point must system – with ten being a ten perfect round and one being a losing round.

This is used to determine who the winner of each round was, with three rounds equaling a complete fight.

Scales include Aggression, Defense, Knockdowns, Range, Effective Attack Time and Effectiveness. The system works through an algorithm which assigns points based on which fighter wins each category.

Only after this process does the round get officially scored.

All of the collected data is used to analyze fighters and their styles. The main factors are listed below (number of factors in brackets).

The higher the number of factors favoring a fighter, the greater their odds are of winning future matches.

The system has proven to be an accurate and reliable way to compare fighters using statistics and data rather than opinion and argument, ultimately providing more insight into mixed martial arts as a whole.

A list of all the factors involved in the algorithm can be found here.

The Mathematics Of The Fight

FightMetric has collected data from almost 4000 fights since it was implemented into the UFC official scoring criteria in 2012. Using these statistics, as well as other databases such as Wikipedia and Sherdog, I have written a computer program to analyze the statistical information of each fighter compared to one another.

The program works through an algorithm to determine the probability of X outcome occurring, finding the likelihood of Y happening over Z.

Unfortunately, there are still some limitations with the data found in official UFC fights. For example, not all the statistics are recorded on every single fighter during every round, so it is impossible to calculate effective striking distance and accuracy.

FightMetric: Where MMA Meets Statistics - Picture

It also does not differentiate between head and body strikes, and only takes into account the number of strikes thrown and landed, not the impact or damage caused by each.

The program is far from perfect, but it is an incredibly useful tool to analyze the statistics involved in a mixed martial arts fight and how they affect the overall outcome.

There are four factors that can affect the outcome of a fight: Striking, Wrestling, Grappling and Health.

Sources & references used in this article:

Performance evaluation and favoritism: evidence from mixed martial arts by P Gift – Journal of Sports Economics, 2018 – journals.sagepub.com

The Southpaw Advantage?-Lateral Preference in Mixed Martial Arts by J Baker, J Schorer – Plos One, 2013 – journals.plos.org

Application of the matching law to Mixed Martial Arts by HA Seniuk, JP Vu, MR Nosik – Journal of Applied Behavior …, 2020 – Wiley Online Library

Effect of rule changes on technical-tactical actions correlated with injury incidence in Professional Mixed Martial Arts by JR Fernandes, F Dal Bello, MADB Duarte… – Journal of Physical …, 2018 – researchgate.net

Effective Aggressiveness and Inconsistencies in the Bijuridical Treatment of Aggressive Behaviour: Mixed Martial Arts, Bullying, and Sociolegal Quandaries by SG Ross – UNBLJ, 2015 – HeinOnline

A winning smile? Smile intensity, physical dominance, and fighter performance. by MW Kraus, TWD Chen – Emotion, 2013 – psycnet.apa.org

He Hit First-An Analysis of Mixed Martial Arts Matches to Determine The Significance of the First Strike by N Gullo – 2013 – FENN-M&S

Fighting fandom: How fan identity influences self-reported and physiological arousal during exposure to violent sports imagery by J Kylmälä – 2020 – helda.helsinki.fi

MMA Mastery: Ground and Pound by M Devlin, AC Billings, J Leeper – Journal of Sports Media, 2016 – muse.jhu.edu