Size Matters. The Numbers Don’t Lie.

There’s a rumor that “size doesn’t matter.” Perhaps size isn’t supposed to matter in the grand scheme of things, but we all know it does. Life just isn’t fair that way. In reality, inadequate size severely limits your likelihood of playing NBA hoops or even becoming president. Conversely, no wrestlers cut from the Sumo squad will ever be competitive divers, horse jockeys or fighter pilots. Basic anthropometrics (e.g., human measurements like height, reach, hand dominance, etc.) will follow you everywhere, and influence your life in subtle, but significant ways. And the surrounding world gently pushes us towards pursuits that favor our genetic disposition like an Invisible Hand pushing good ideas towards success. Studies have even correlated income levels and leadership prowess with height, though I’m still trying to figure out if blondes really do have more fun…anything in the name of science.

So size matters much more than just whether or not you can ride a roller coaster (sorry Urijah, come back next year). It can influence how different people perform, succeed, or fail at a variety of physical tasks. But we want to know what it means for MMA. Specifically, how much can we know about knockout rates in fights based only on the weight class? Let’s settle that question once and for all.

Starting with a fundamental hypothesis, that bigger guys score more knockouts, we arranged the vast database of UFC fights provided by FightMetric (the keepers of all UFC statistics) by how each fight ended, and by weight class. For simplicity, we filtered out fights that were disqualifications or later overturned, as well as catch weight fights, which were all small in number. We also isolated only fights occurring in the UFC, spanning a recent period of almost three years for consistency of matchmaking in the “modern age of MMA.” Note, I’ve left out featherweights and bantamweights because when the analysis was run, they had not yet joined the UFC, nor would we have a full sample size if we reran the analysis today.

Size Matters: Fight Outcomes by Weight Class

Fight Outcomes by Weight Class

 

Heavy Hands

 

Clearly, heavyweights do score more knockouts, almost three times the rate of lightweight fighters. A simple linear regression using only the weigh-in value to predict the rate of finish by strikes accounts for ~90% of the variability in this model, indicating the extremely high importance of this key metric. It also reveals the specific value of size: ~1/3% finish rate per pound. That is, for each additional pound of fighter weight, striking finishes go up by that much. For each full percentage point increase, that’s one additional fight in a hundred that could put someone to sleep in devastating fashion, or add to the highlight reels. It may not seem like a lot, but scaling up 10 lbs. boosts the TKO rate by 3%, leading to an overall increase of 33% spanning the 110 lb. difference between heavyweights and lightweights. Now that is a lot. That means you can bet that a heavyweight fight will end in strikes and be correct slightly more than half that time, while betting against a lightweight fight ending in strikes will get you paid four out of every five times. Analysis of Strikeforce competitions revealed very similar patterns, despite a variety of slight differences in environmental variables. Size matters. Period.

KO/TKO Rate by Weight Class

One obvious reason for is this relationship is the idea of knockout power. Some guys have it, and some don’t – but lot of that has to do with how much muscle mass is available to drive the key strike. Shane Carwin once contemplated his powerful striking, saying that when his hands touch people, “they go to sleep.” More accurately, he’s loaded with muscle mass all over his body, making his heavy hands that much more powerful. Muscles are like engines; they burn energy and convert chemical potential energy into mechanical kinetic energy. The more muscle you have, the more work you can do. The men currently populating the upper echelons of the UFC heavyweight division have a lot of muscle mass, and their fists can do some serious work. The result is that we see a lot of heavyweight fights ending in strikes.

 

A contributing factor for heavyweight knockout artists is their higher overall power striking accuracy. In the critical metric for knockout blows – the power strike to the head – heavyweights beat all other weight classes in accuracy. This is likely due to lower agility and quickness in heavier fighters, and their inability to move fast enough on defense. More mass is great once you get it moving, but it takes a lot of effort and it doesn’t happen on a dime (the concept of inertia). The higher striking accuracy rule for heavyweights is even more pronounced in the clinch and on the ground. Conversely, the quick and elusive lightweights have the lowest power striking accuracy of any weight class standing, in the clinch and on the ground. Are they poor strikers? Probably not, their opponents are just fast.

 

Physics is Merciless

 

Isaac Newton determined that force is equal to mass times acceleration (F = MA). Physics is merciless. There’s no magic in this kind of Force, and there’s no hologram bracelet you can wear to break these laws or prevent their inexorable, unrelenting truths. The monstrous meat paws of a heavyweight, with correspondingly massive arms, back, hips, and legs propelling their hands forward, are capable of inducing rapid acceleration of a human head upon contact. This brings us to another important equation in collisions: p = mv. That is, momentum equals mass times velocity. When the point of impact occurs between fist and head, we’re ultimately concerned with how much energy is transferred from the hand into the formerly stationary head, and collisions like this are governed by momentum. Momentum is also conserved; it can neither be created nor destroyed. A direct strike means imparting as much energy as possible into the acceleration of the target’s (less massive) head. Both equations rely on mass.

 

Looking at the other variable in the equation leads us to believe that throwing faster punches should also equate to knockouts and since lighter weight fighters have quicker hands, they should have more KO’s right? No. The much larger mass of a heavy fighter seems to dominate any increase in quickness of lighter hands. Larger fighters also tend to be taller, with longer arms and reach. While it certainly takes more energy and time to get a big hand accelerated and moving at high velocity (inertia again, thanks Newton), long arms also have a longer runway to accelerate, before they run out room, quickly decelerate, then finally stop at their maximum reach distance. The shorter arms of a lightweight may snap into movement quickly, by they can’t accelerate for very long. Additional analysis could even separate true knockouts from fights stopped due to a barrage of strikes from a dominant position. Whether or not most lightweight fights ending in TKO are from volume, as opposed to single power shots might further show that the fight strategy for a lightweight should account for the reality of their size. We’ll have to tackle that analysis another time.

 

What Have We Learned?

 

Knowing that their chances of scoring a knockout at lightweight are drastically reduced, smaller fighters have attempted to win on the ground far more often than heavyweights. The physics of the situation has changed the way fights go down. Lightweights attempt 72% more takedowns per fight than heavyweights, and attempt 95% more submissions. A compounding bonus for the lightweight submission game is that the percentage of submission attempts that are locked in tightly is more than twice that of heavyweights. The intermediate weight classes line up in these metrics as we’d expect, and thus the performance differences between fighters of varying size goes on and on. Suffice to say, if a BJJ submission expert is going to get a chance to wield their skills in an MMA fight, they’re going to need some time to get position and work their game. For lightweights, their increased chances of surviving a few direct strikes along the way means their chances of lasting long enough to develop a finishing position is much greater than for heavyweights, who may never get that chance. When was the last time you saw a heavyweight willing to absorb blows while repeatedly shooting for takedowns? Ultimately, we’ll see better BJJ practitioners among the quicker, elusive, lighter weights where their skills may shine than at heavyweight, where they’re only one crashing XXXL glove away from some Octagon naptime.

 

What to Remember

  • Over half of all heavyweight fights end via strikes, but only one in seven by submission
  • Almost half of all lightweight fights will go the distance, and only one in five end via strikes
  • Only lightweights finish more fights by submission than by strikes; it’s still mostly a striker’s game
  • The more muscle you have, the more you’ll be able to accelerate your opponents head, which is bad (for them)
  • Fighters of a given weight class will adjust their strategies to account for their size and attempt to finish on the ground
  • Size matters, and physics is merciless

 

 

Notes:

Source data was provided by FightMetric, and analysis performed by Reed Kuhn. An abridged version of this article appeared in FIGHT! Magazine 2011 as part of the “Fight Science” feature.

 

 

Be Sociable, Share!
468 ad

5 Comments

  1. JB says:

    I’m currently working on a Statistics project for a college course and I’m looking for raw data that is split by year, weight class and fight outcomes. Any help or point in the right direction would be awesome

    Follow fight fan, JB

    • Reed Kuhn says:

      JB, raw data that is already organized is hard to come by in any field. I have a fellowship that allows me some access. But Wikipedia has each UFC event stored as a single page listing the names and weight class of every fight along with the outcome. If you set up a spreadsheet to allow you to quickly tally weight classes and outcomes, you can click through each UFC event and capture as much data as you want for later analysis in a fairly short period of time. Hope that helps!

  2. Steve @ doyoubjj.com says:

    Awesome write-up man. This is the sort of analysis about mma that I love, even though I’d be a lightweight, lol. It only hurts my feelings a little. Also, look at the submission rate and the decision rate of the hw’s so there.

    I just got into a semi-heated discussion on an mma forum precisely about what you wrote about, the equations of force and such. I was trying to provide an overall understanding of the f=ma equation, but of course people on the other side were saying you can’t apply it wholly. I knew that but didn’t want to get into energy transfer and all that. I did learn something though so it was all good.

    I’d love it if you let me drop a link to my bjj website that looks at bjj and mma kind of from a similar perspective, one of engineering and physics. I think it is great to be able to use simple machines like levers to submit people. This link goes to my main bjj reading/learning/discussion page. Thanks man!

  3. Daniel B says:

    Just came across this via another article. I like your analyses but, I think there may be another factor at play with the heavyweight T/KO’s. The power to weight ratio (which favor’s smaller fighters) means that it is a lot harder for heavyweights to get out from underneath their opponent on the ground. Here I think you are looking at a squared factor (strength=area) vs a cubed factor (weight). This would tend to lead to more ground and pound TKO’s rather than full KO’s so more detailed data would probably help separate this. For the lighter weight classes it seems that a dominant top position such as mount seldom lasts for long and the worst damage done is usually a transition to the rear. But for a heavyweight mount is a big problem often leading to more striking damage. I can imagine at near zero gravity it would almost be an advantage to be ‘on the bottom’ as you would have more control. There’s probably a lot of factors at play such as

    - Wide difference in weights at heavyweight
    - Rotational strength of the neck muscles compared to momentum of punch
    - Larger head radius leading to more rotational leverage of punches
    - Heavier weights classes tiring out more quickly

    • Reed Kuhn says:

      These factors may be true as well. But punch for punch, heavyweights score more knockdowns from a standing position than any weight class, and the general trend line through divisions is a positive correlation between size and knockdown rate. Because knockouts are collisions, the physics in question is less about Force (F=ma), and more about kinetic energy and momentum. A fist can actually be DEcelerating when it hits its target but it may be maximizing the momentum of the mass of the arm/body that is behind it.

      A new article “Deconstructing the Greatest UFC Striker” will be posted soon that shows the graph of knockdowns/per landed head strike – that should clear up any questions about size and power in the UFC divisions.

      Also – very interesting suggestion on examining the time spent in mount by weight class. I have that data and can test that hypothesis. Expect to see it in the book, and thanks for the great comments!

Trackbacks/Pingbacks

  1. The Finish Line - Fightnomics | Fightnomics - [...] demonstrated in a previous Fight Science (FIGHT! Magazine, March 2011), TKO rates correspond directly to body mass. Bigger fighters ...
  2. How UFC Fights End, by Weight Class and Method - Fightnomics | Fightnomics - [...] know that size matters in MMA.  Bigger fighters finish more fights.  But bigger fighters primarily finish by strikes, so ...
  3. Theme of the Week: Compassion | Westside AKF and Crossfit Hillsdale - [...] a side note, here’s a really cool article about weight classes in the UFC and how different weight classes ...
  4. Striking Assessment: UFC Flyweight Division - Fightnomics | Fightnomics - [...] has landed 3 knockdowns in his 26 minutes of UFC action, finishing two fights by TKO. Given that size ...
  5. UFC Bantamweight Striking Assessment - Fightnomics | Fightnomics - [...] of the 27 bantamweight fighters graphed have yet to score a knockdown in their Zuffa appearances, not an unusual ...
  6. How All UFC Fights End – May 2013 Edition - Fightnomics | Fightnomics - [...] in MMA is that Size Matters. There’s basic physical and scientific reasons for this which you can read more ...
  7. Deconstructing the UFC’s Pound for Pound Most Dangerous Striker - Fightnomics | Fightnomics - [...] or high kick will have very different outcomes depending on who threw it, and who absorbed it. The reality ...

Leave a Reply

Your email address will not be published. Required fields are marked *

*

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>