Interview with Reed Kuhn (MMA Analytics)
Reed Kuhn is the leading figure in MMA analytics. Reed's work can be seen on his own website, televised UFC events, other various MMA websites and talk shows, and in his upcoming book, “Fightnomics: The Hidden Numbers and Science of Mixed Martial Arts.”
Can you briefly describe your education and career background and how it has helped you in the analytics industry?
My undergrad work was in Physics at Washington and Lee University. Going to a liberal arts school to study physics wasn’t going to keep me on the hard core science path for long. I was literally the only pure physics major in my graduating class. I immediately went into consulting thinking I’d never have any reason to go back to school.
After three years of consulting, however, having transitioned into more scientific and analytical roles, I decided on a part-time master’s program at the University of Virginia in Systems Engineering. I fell in love with systems analysis. I looked at the world a different way after that, wanting to break everything down into diagrams and numbers so I could think of how to “optimize the system.” Facing the difficult prospect of pursuing a PhD and likely leaving the working world for a very long time, I instead went to business school at Duke University for an MBA. From there, I was immediately drawn to management consulting, partly out of ignorance of the financial sector, but also because of the fast project environment and problem solving emphasis that it provided.
After a few years of strategy consulting post-MBA, all my prior academic and professional experiences converged. I became a guy that was analytically numerate and experienced in looking at data and extracting meaningful conclusions. When I stumbled upon a database of MMA data, I simply did what came naturally. Getting to play with information as a hobbyist, but bringing the full power of academic and professional training to bear on the subject was a blast. Like a NASA engineer helping out with a kid’s science project, it was hard not to go overboard. In no time I was presenting 30-page consulting decks to UFC fighters and their trainers, graphs on every page and numbers everywhere. I looked at being a fighter and each matchup as a system to be quantified, understood, and then optimized towards some objective. From physics to strategy consulting, every bit of my background is relevant.
There hasn’t been much done with regards to MMA and analytics. What inspired you to start looking at that industry?
I’ve always been an MMA fan, but I had no idea that there was data for the sport until I began searching for it online. I was switching consulting firms in 2009 and took a few months off. A good friend of mine, Nick Palmisciano, a classmate from Duke and founder of Ranger Up apparel, took me to my first UFC event when he was sponsoring a fighter. I was in a new place that I knew nothing about and so I did what came naturally. I asked a lot of questions of fighters, trainers, and even the producers of the event, but didn’t really get satisfying answers. No one knew the basic finish rate of fights (what percentage end by KO/TKO/Submission), or how the dynamics of divisions, or size differences really played out. No one had real answers, and if they did there certainly weren’t any numbers attached. So I simply assumed that if I went digging, I could be the first to provide them.
Not knowing that there was data out there, my first hypotheses about the sport were focused on behavioral economics and loss aversion. But in attempting to find some organized data, I stumbled onto FightMetric, who has just recently launched and completed quantifying their historical UFC database. They were looking for research fellows, and I applied. The robustness of that database changed the type of things I wanted to analyze, and opened the aperture of possibility immensely.
What are the advantages and disadvantages you face working with a sport where there is not much research done before you?
The obvious advantage is the greenfield nature of the landscape. No one has addressed MMA from an analytical perspective, so there’s opportunity to innovate and pioneer everywhere. That’s also the downside. Few people are prepared to see numbers for “cagefighting,” and many are not interested in turning a visceral and exciting sport into another domain for stats nerds. Fortunately, the opportunity tends to outweigh the naysayers, as it did in many sports before MMA.
There’s also the statistical literacy barrier. FightMetric created a system of quantifying MMA fights that is not widely known, so even when I boil things down to simple metrics, it’s still foreign to many fans, even those who watch the sport regularly. Slowly but surely this is changing. The UFC is putting statistical bullet points into their fight previews, and even fighters themselves have been using some of the statistical vocabulary when they talk about past fights. The first big hurdle in this process was simply taking what happened in a sport and turning into organized data. The analysis is the fun and easy part. Mainstreaming and popularizing all of that will be the next step.
What are some of the MMA things you look at analytically? Can you briefly describe some of the metrics you use for your analysis?
I started with the obvious macro-trends: how fights go down. The history of the UFC is very short by mainstream sports standards, but the evolution has been rapid. Looking at fight trends in terms of how they end tells an interesting story of how fighting styles have evolved under competitive pressure. The initial foundation of the UFC backed by the Gracies was supposed to demonstrate the dominance of Brazilian Jiu Jitsu as a fighting system. But their victories caused a reaction by the market, and the competition not only got a lot better, but forced MMA into new directions. The data reflects all of this, and also shows that while the sport has recently stabilized, there are subtle changes still at work.
There’s also the basic Tale of the Tape that MMA inherited from Boxing. Like many other things that were inherited (e.g., judging and scoring systems), it’s not perfect for this sport. I’ve analyzed every data category on the Tale of the Tape to understand which ones mean something, which ones don’t, and then the larger question of what’s missing from it. That leads to a lot of interesting questions like: is the Southpaw advantage real, and if so, how exactly does it work inside the cage? Some of what I would find might be obvious, and others counter-intuitive. But without the numbers to prove the case, no really “knew” anything, they only suspected or had a hunch. That’s the benefit of the scientific method; it keeps us from perpetuating false misperceptions and helps us figure things out with greater precision and certainty.
The most specific type of analysis uses in-fight statistics to look for patterns of performance, and fighter benchmarks within the context of their weight class. I can look at every aspect of fighting (strikes, takedowns, grappling advances or submissions) and consider them in the context of pace and success rates, overall and by sub-category, and how all these things change over time, whether throughout the history of the UFC or just during the course of a single fight. Some of the conclusions from these analyses could influence how fighters change their strategy both before and during fights. So really, that should convince anyone that analytics has a place in this sport. Like all competitive environments, everyone is looking for an edge. Analytics has been providing an edge in baseball for a long time. More recently football, basketball, and even hockey are employing analysts at the team level to maximize their performance. Fight camps can too.
You have a new book, Fightnomics, coming out soon. Can you talk about this? What inspired you to write the book and what was the process like?
I left the world of strategy consulting at the beginning of the year and gave myself some time to focus exclusively on writing a book. Up until that point, analyzing MMA stats was just a nights and weekends kind of hobby, and writing articles for magazines or working with the occasional fighter was something that was fun, but also a distraction. Along the way my Fightnomics.com blog also gained traction, and more often my research and analysis was being linked to and quoted by larger MMA media outlets. Still, taking the leap and severing my income was a huge risk, and I have yet to see if it will ever pay off. But I can confirm the ride has been fun! I’m looking at this as a sabbatical to do some science, and make a contribution to a certain field, however much a niche one.
While I’ve also ramped up other projects in MMA and increased the number of clients I work with, research for the book has run parallel to all of it. It’s almost done, and will hopefully be available on Amazon by the end of November. “Fightnomics” the book covers a lot of the basics and should serve as a definitive introduction to MMA statistics and analytics, but there’s also room in there to explore some more advanced and controversial topics. It’s like “Moneyball and “Freakonomics” were dropped into a cage fight. Having already set a likely page count of 300 pages (with a LOT of graphs), I think I may have to save some analysis for a later date in the interest of getting to market sooner. From there, I’m also considering some very analytically oriented eBooks that will essentially be a visual and quantitative summary of my research.
Given the niche market for MMA content, we’re going the self-publishing route. I have the help of New York Times Best-Selling Author Kelly Crigger, who is a personal friend but also brings the experience of having published books in MMA and other subjects. Fortunately, it’s easier to go this route than ever before, and I don’t think the lack of a “real” publishing house will hurt our distribution at all. MMA fans all digest content from a finite list of providers, and by plugging into those outlets we’re confident we can reach all the fans that are likely to be interested in a book like this. At that point, we hope word of mouth and social media kicks in to help spread the word, as those are powerful allies in today’s online world.
What is your favorite thing about working with sports and analytics?
It’s the combination of a hobby and real world skills. I enjoyed analyzing numbers for a sport because it was fun, but also interesting to learn something no one had ever learned before. I also brought DayJob skills to the table. The trouble with hobbyists is that they are constrained by time and money, and they don’t give the hobby the same effort that they can give other projects within their true expertise. So “going all in” on this project has truly been fun. I get to turn on the TV and sometimes see my insights at work, and hear it in the quotes of the athletes or broadcast analysts. I’d be watching these things if I wasn’t doing this project, but now I’m actually connected to them.
What advice would you give to students and young professionals interested in breaking into the industry?
For me it was mastering skills that were versatile enough to be brought to a new subject. Analytics as an expertise and consulting as a professional are perfect for that. An interest in sports is not enough. There are so many fans, and die-hard fans will trump you on dedication and historical knowledge. Differentiating yourself with a tool kit of analytical capabilities is critical. Knowing information is not enough either. You have to understand and wield the scientific method to discover new things that haven’t been found before so that you can maintain learning momentum. You have to know how to ask the right questions, and then have the skills to go and answer them. Being able to do this eventually requires information and data, which is a barrier to entry for most sports. However, the proliferation of sports data that is publicly available now should enable a lot of hobbyists to contribute to the larger knowledge base.
The final challenge is the hardest of all, and that’s proving your value to industry. I don’t think I’ve cracked that riddle yet. The MMA world is fairly open, so I’ve attended a lot of events and networked extensively, which is definitely a big advantage over trying to have an impact from a disconnected ivory tower. A workaround to hoping you get noticed is to force industry to notice you. Become an expert on a small area, gain credibility in that small space, and become the person that is sought out for that subject. Then slowly expand your range of expertise. If you truly do good work, eventually competitive forces won’t be able to ignore the potential advantage that you can provide.
Many thanks to Reed for his time and insights. As a friendly reminder, don't forget to visit Reed's website and follow him on Twitter. Sign up here to receive a notification when the book is ready and be eligible to win an advanced copy signed by the authors.