AI in MMA could become huge in the coming years. MMA, much like most other sports takes time to prepare for and, more than other sports is highly anxiety-inducing, especially for the fighters themselves.
Although all fighters, coaches and teams will enter a fight with a good idea as to how their opponent will fight, for example, someone like Khabib Nurmagomedov always looked to wrestle during his fights. Alternatively, someone like Israel Adesanya is highly unlikely to commit to a grappling-heavy approach due to his toolkit.
But how could AI in MMA affect the way that fighters approach a fight? We take a look at the possibility of AI in MMA, the various ways in which it could work and realistically, how successful it could (or couldn’t) be.
AI in MMA: Fight Prep
One way in which AI in MMA could help is with preparing a fighter for a fight. More often than not, a head coach, especially one who is head coach of a huge team, take Mike Brown for example, will have a huge number of fighters, all competing at various different points throughout the year, across the globe, against different types of opponents. He will be cornering and coaching fighters of all different styles and abilities.
Someone who is competing in their first MMA fight isn’t going to be able to tailor their own personal training plan and game plan than someone like Dustin Poirier, who has 40 MMA fights.
To say that a lot of Brown’s time will be taken up with fight plans and prep and that he must spread himself thinly is an understatement. If only there was a tool that Brown could input data into which would formulate a game plan for him, against a specific type of opponent…
This is where AI in MMA could come in. AI relies upon computers to learn and perform tasks to solve problems. It leans on a huge amount of information, simulating billions of variables and identifying future patterns that are likely to occur.
There are some techniques in MMA that are far more likely to occur than others. For example, the emergence of the leg kick in recent years is something that AI would have picked up on, having become far more prominent in the past few years.
In theory, a head coach could formulate a game plan for an upcoming opponent based upon what AI has told them. Let’s say that Dustin Poirier is set to rematch Conor McGregor for a fourth time. AI read all the data already out there on McGregor, his 28 professional MMA fights, any training footage that is out there on the Irishman, his boxing match with Floyd Mayweather and any other relevant information.
There is plenty of data out there on fighting styles, tendencies, preferred techniques, weaknesses and likely outcomes, allowing coaches to formulate a game plan around the information provided.
Again, let’s take McGregor vs. Poirier 4 as an example. Brown could run a simulation of this fight 1000 times, focussing on the most likely outcome across all 1000 fights. If, for example, McGregor wins multiple times via knockout after Poirier throws an extended combination, this is something Brown could try and train out of his fighter. Likewise, if Poirier wins over 300 times after taking McGregor down, this is something that he should look to work into his training regime.
AI in MMA: Judging
How many times have we seen wrong, controversial or extremely questionable decisions in the world of MMA? AI could certainly aid this. We saw the introduction of an AI judge for the first time in a major fight when Tyson Fury and Oleksandr Usyk rematched in December. The AI judge scored this fight in a questionable manner, having Usyk winning 118-112, scoring two rounds 10-10, something that is rarely seen in the sport.
For the first time ever, an AI-powered judge will monitor the fight 🤖 Free from bias and human error brought to you by The Ring. This groundbreaking experiment, which won’t impact the official results, debuts during the biggest fight of the century, #Usyk2Fury, on December 21 🥊… pic.twitter.com/RdslVlBLRC
— TURKI ALALSHIKH (@Turki_alalshikh) December 17, 2024
Although there is far more data in boxing, AI in MMA could see judging be used artificially in years to come. What the Fury/Usyk judge showed, however, is that there is certainly work to be done.
AI in MMA: Rankings
This is something that UFC president, Dana White has commented on multiple times and something that he’s clearly passionate about. AI in MMA could aid with fighter rankings. Taking the UFC as an example, there are clear issues. Let’s look at women’s bantamweight. Chelsea Chandler is ranked at 135lbs, despite never winning at the weight. She has two wins in the UFC, one at 140lbs and one where she missed weight.
Michael Chandler holds the #7 spot at lightweight. This is in spite of him not having won a fight since 2022 and having never put back to back wins together in the UFC. His most recent win was against Tony Ferguson, who, at the time was ranked #9. Should he really be ranked at #7 over someone like Renato Moicano, who is ranked #10 and has put four back-to-back wins together, featuring three finishes and against very respected opposition?
The pound-for-pound rankings, especially in the UFC have been an issue for a number of years. Is it Jon Jones or Islam Makhachev? Should activity be a factor? What about the strength of the opposition? Should history be taken into account? AI could potentially fix this.
AI in MMA is certainly going to happen in some way, shape or form. It’s important for the sport, which is still fairly new, to move with the times. With Dana White now a member of the META board, we expect big steps to be taken with regard to AI in MMA in the coming years.
