Error: Twitter did not respond. Please wait a few minutes and refresh this page.
Intelligent, unique MMA analysis
I’ve just updated the top 25 ELO rankings for MMA fighters. You can check out the new rankings by clicking any of the links on the left-hand column of this blog. If you want to know where Alistair Overeem and Rashad Evans rank after their surprising losses at UFC 156, just click on the heavyweight or light-heavyweight rank, respectively.
One thing I have to note is that I’ve decided to remove the ratings penalty for fighters outside the UFC. Even though having that penalty did make my rankings conform to the standard better, I just felt it was unfair to fighters competing outside the UFC. A fighter like Michael Chandler was way lower in the rankings than he should have been.
The downside is that there are now oddities like Pat Healy being ranked #6 at lightweight. The main flaw of ELO is that it tends to overrate fighters who go on long winning streaks against merely decent opposition, as is the case with Healy (and Ben Askren, who is now the #4 welterweight).
If I submit my rankings to the Bloody Elbow consensus (a distinct possibility), I’ll probably make some manual adjustments to where the fighters rank based on a little bit of subjectivity. The one fighter that drives me crazy the most is Anthony Johnson. Johnson built up a very good rating at welterweight, before weight cutting issues forced him all the way to light-heavyweight. In ELO, Johnson keeps the rating he built at 170 pounds, which makes him the #6 light-heavyweight. If I was to submit a ranking to the consensus, I’d probably drop Johnson to somewhere around #15 because of his relatively unproven status at light-heavyweight.
The best thing I can say is: take the ratings with a grain of salt, and understand that they are simply the ELO output, no more, no less.
I’ve also added ELO ratings for all UFC fighters to the links on the left-hand column. These rankings are a little different than the top 25 rankings; the ratings themselves are the same, but the fighters are placed in the weight class they’re going to compete in in the future, instead of the weight class they last competed in. These ratings are meant to look at how relevant future fights are. Let’s take Danny Castillo vs. Paul Sass as an example. If we go to the page showing all UFC lightweights, we see that Sass is ranked #26 among UFC lightweights, while Castillo is ranked #37. Based on those rankings, it’s probably safe to say that whoever wins between Sass and Castillo will move into the top 20.
To summarize, the UFC rankings are mostly for fun, to play matchmaker a little bit, and to look at some upcoming fights and see just how important they are.
Listed on the left-hand column of this blog are links to my top 25 rankings for each weight class in mixed martial arts. To rank fighters, I have decided to use a modified ELO rating system. ELO works by taking the rating of two fighters, estimating each fighter’s probability of winning based on those ratings, and then adjusting each fighter’s rating based on the outcome. The more “surprising” a result is to ELO, the more the fighters’ ratings are adjusted. Click here if you’d like a detailed explanation of what ELO is and how it works.
For this ranking, I have decided to use a K-factor of 250. That’s because, from what I’ve seen, this K-factor is the best model for how the sport of mixed martial arts works. K-factors that are too small tend to reward quantity over quality; if the K-factor was 20, the top-ranked fighters would be Travis Fulton, Dan Severn, and Jeremy Horn. These are not the best fighters in the world, they just have the most wins. If the K-factor is too large, a top five fighter can be jettisoned from the rankings altogether with one untimely loss. By setting the K-factor at 250, I’ve calibrated the system to reflect the best balance between rewarding quality wins, without punishing a fighter too much for one loss.
A question that has to be asked is: what are MMA rankings? What are we ranking? If we rank fighters alphabetically by first name, then Aaron Anspach is #1. Obviously that’s not what I’m doing, but I would argue that MMA rankings don’t rank fighters based on who the “best in the world” is either. Instead, they rank fighters by relevance; specifically, how deserving a fighter is of a title shot.
Because I’m ranking fighters based on relevance, fighters who do not compete in the UFC will get a penalty in my system. That might seem like me hating on fighters in other organizations, but the fact of the matter is that fighters outside the UFC are generally less relevant than fighters in the UFC. Fighters whose last fight was in Strikeforce or Bellator will get a small penalty, and fighters in any other organization will get a large penalty.
Trust me when I say that these modifications are for the best. If I didn’t make these adjustments, these rankings would be full of fighters who broke into the top 25 based on a winning streak against mediocre competition in a regional promotion.
Because this is a mathematical system, there will be some strange/unusual rankings. For the most part, these rankings go along with conventional wisdom, but there are a few exceptions. While I want these rankings to be credible, I also don’t want to conform to group-think. If one of my rankings is a lot different from the norm, it doesn’t make my rankings right or wrong. If anything, it provides an opportunity for discussion about the fighter who is ranked higher or lower than the standard.
Here are the top 10 rankings for each weight class:
|4||Junior dos Santos||2715|
|10||Rafael dos Anjos||2663|
|7||Jussier da Silva||2191|
After the UFC’s decision to have Chael Sonnen fight Jon Jones in April, Forrest Griffin was left without an opponent for UFC 155 in December. To replace Sonnen, the UFC chose Phil Davis, who overwhelmed Wagner Prado en route to a second-round submission victory at UFC 153. In my opinion, this fight will be the end of Forrest Griffin as an upper-tier light-heavyweight in the UFC.
ELO: Forrest Griffin 2735 Phil Davis 2731
After making the switch to ELO, I sometimes get the (probably wrong) feeling that I’ve stumbled on what the UFC uses for matchmaking. One example: their insistence on having Rory MacDonald take on B.J. Penn. I currently have MacDonald ranked #10 at welterweight with a 2854 rating, while Penn is ranked #11 at welterweight with a 2840 rating. Forrest Griffin vs. Phil Davis represents another case of two fighters having similar ELO ratings.
But I don’t think the fight between Davis and Griffin is really that close. In fact, I think it’s going to be a long night for Forrest Griffin.
One area Griffin has a clear, distinct advantage in is the striking game. Davis has consistently shown a disdain for striking in his fights – not once has Davis won a fight because of what he did as a striker. Instead, Davis wins fights by taking his opponents down and dominating them with superior grappling. If this fight is inexplicably a striking match, Griffin should have a clear edge. Griffin isn’t a super-talented striker, and doesn’t have serious knockout power, but what he does bring to the table is aggression and conditioning. Griffin often wins by simply putting pressure on his opponents and wearing them out. Case in point: his upset victory over Mauricio “Shogun” Rua at UFC 76.
Unfortunately for Griffin, it’s extremely difficult to see him keeping this fight standing. While Griffin manages to defend takedowns against some fighters by virtue of being a huge light-heavyweight with decent takedown defense, that won’t be enough against the prototypical athlete in Phil Davis. As Davis has shown in fights against Prado, Antonio Rogerio Nogueira, Tim Boetsch, and Brian Stann, he will relentlessly pursue takedowns. To date, the only man to out-wrestle Davis was Rashad Evans, an elite light-heavyweight fighter with excellent wrestling himself, to go along with dangerous striking.
Phil Davis is going to look to take Forrest Griffin down early and often, and he’s going to succeed at it. To make matters worse for Griffin, Davis is an excellent grappler, who won’t be in any danger of falling into the trap of a triangle choke, armbar, or any other submission attempt Griffin might try from bottom position. Griffin at least had the submission threat as a path to victory against Sonnen. Against Davis, it’s just not going to work.
For Griffin, that leaves conditioning, and that’s not going to work either. Griffin may be a cardio machine, but so is Davis, a man who fought five hard, grueling rounds against Rashad Evans. And unlike the Evans fight, I feel Davis will succeed in imposing his game on Griffin – taking the fight to the floor, and engaging in ground and pound and submission threats.
I don’t see Davis stopping Griffin in this fight. Griffin is a mentally tough fighter, and Davis isn’t punishing enough with his strikes to win by knockout or TKO. Griffin also has very good jiu-jitsu, and has never been submitted in his career. I don’t think Davis will be the first. But I do see Davis getting this fight to the ground, and forcing Griffin to effectively win the fight off his back.
Davis doesn’t have good striking, and against a fighter with good striking and excellent takedown defense – at light-heavyweight, this pretty much narrows it down to Jon Jones and Rashad Evans – he will struggle. Griffin has the striking, but doesn’t have the takedown defense. I see Davis winning by decision and placing himself into serious title contention. For Griffin, a fighter whose motivation has been questioned in recent years, I see a demoralizing loss which will send him out of top ten lists and into possible talk of retirement.
As promised, here are lists of the top ten rated opponents of Fedor Emelianenko, Jon Jones, Anderson Silva, and Georges St-Pierre. Why look at each fighter’s top ten opponents? Because champions are defined by who their toughest opponents were. Fedor is not a legend because of his fights against Hong Man Choi and Zuluzinho.
The highest average is a virtual tie between Anderson Silva and Georges St-Pierre. Fedor Emelianenko and Jon Jones are well behind. However, Fedor has two unique disadvantages in this study. One is that the heavyweight division is not as deep as middleweight or welterweight, and therefore, elite heavyweights tend to have lower ratings than elite middleweights or welterweights. The other is that Fedor’s prime was earlier than Silva or St-Pierre, meaning that Silva and St-Pierre are beneficiaries of ratings inflation.
As for Jon Jones, he just hasn’t been a champion long enough yet to have built the same kind of resume as the other fighters on this list. But it’s noteworthy that Jones has faced the most opponents with a 3000 rating or higher. All Jones needs is time.
As I said in yesterday’s post, I can throw all the numbers at you I want, and I can argue for St-Pierre as having the toughest opposition (and I would). But if you’re trying to determine who the greatest fighter of all time is, there is no number that can make that argument. It’s up to you.
I’ve been getting requests for a while to look at the relative competition of the three men widely regarded as the most dominant champions in the UFC – Anderson Silva, Georges St-Pierre, and Jon Jones. In the past, with SILVA, doing this would have required many hours of work. With ELO… let’s just say it’s quite a bit easier to do.
For this post, I’m going to look at only the championship fights of each man. It’s an unfair comparison if I include non-title fights, because Silva fought for the title in only his second UFC fight, while St-Pierre and Jones needed a few more fights before they became champions.
(Two notes. First, Dan Henderson’s rating may seem lower than it should be. This is because Henderson was in the rare position of fighting for the title despite losing his last fight. Henderson was representing PRIDE Fighting Championships as their champion in two weight classes. Henderson’s fight against Silva was his second UFC championship fight, after losing to Quinton Jackson for the light-heavyweight championship. Second, Travis Lutter is included because that fight was supposed to be for a championship, before Lutter missed weight.)
(I’ve excluded the third Matt Hughes fight because it was an interim championship fight. Hughes’s ELO rating in that fight was 2656.)
At first glance, it seems that Jones has had the toughest competition, but he’s also had a much shorter championship run than Silva or St-Pierre. The other thing working in Jones’s favor is what is known as “ratings inflation.” Over time, the ELO ratings of the best fighters in MMA tend to rise, because as the sport grows, the distribution of ratings gets wider. The highest-rated fighters go higher, and the lowest-rated fighters go lower. That’s why Matt Hughes had only a 2539 rating when he fought Georges St-Pierre the first time. If I look at just the last two years, the competition of Jones is very comparable to the competition of Silva (St-Pierre has been inactive for too long to make that a meaningful comparison).
Meanwhile, Anderson Silva and Georges St-Pierre are in a virtual tie. St-Pierre has the slightly higher average, but he also lost once; ironically, to the lowest-rated fighter on any of the lists in Matt Serra. Overall, if I had to make a call on who the greatest UFC fighter of all time is, it would have to be Anderson Silva. But I will acknowledge that Silva’s highest-rated opponents were not quite as tough as those of St-Pierre and Jones. For Silva, it’s his longevity that makes him great more than anything else. Give any MMA fighter 16 tough opponents in a row, and he’s bound to lose at some point. But Silva hasn’t.
I know what you’re thinking – what about Fedor Emelianenko? It’s very tough to include Fedor in such a comparison. For one, PRIDE had a different mentality towards title fights than the UFC does. Title fights in PRIDE were relatively rare, occurring about once a year per division. But if I included non-title fights, it would almost certainly hurt Fedor, as he took quite a few fights against over-matched opponents like Yuji Nagata. In addition, Fedor’s prime took place earlier than the primes of the fighters listed above, and ratings inflation would not be kind to him.
In the end, performing analysis of this kind is very tricky, and almost certainly requires the construction of arbitrary criteria. In tomorrow’s post, I’ll look at each fighter’s ten toughest opponents overall, and I’ll expand it to include Fedor. But the only conclusion that can be drawn is that judging greatness in MMA is an exercise of opinion, regardless of how many numbers and stats I throw your way.
Here are my rapid-fire, knee-jerk thoughts about the just concluded UFC 153:
-Physical size as a factor in a fight is very overrated. I don’t know how many people bought into the idea that Stephan Bonnar was a serious threat against Anderson Silva because of his size, but Silva showed why it’s almost a non-factor most of the time. If Silva fought a heavyweight next, how many heavyweights would seriously have a decent chance of winning against him? I’ll give you the elites – Dos Santos, Velasquez, Cormier, and Overeem – but after that, I’d strongly favor Silva. Technique is what matters.
–Dave Herman is just plain weird. Forget all his talk about jiu-jitsu. What was he doing in the second round, when he left his right arm fully extended, and not attacking? By keeping his hands away from his chin, and lifting his chin in the air every time Antonio Rodrigo Nogueira attacked him, Herman was begging to be hurt. Nogueira did just that on multiple occasions, and followed up by shattering Herman’s trash talk with an armbar. Dana White has spoken repeatedly about how he likes “serious” guys. I didn’t think Herman was in danger of being cut from the UFC because he took this fight on short notice, but now, I’m not so sure.
-Two UFC fights, and two annihilations by Glover Teixeira. This time, he badly hurt Fabio Maldonado early, and just kept pouring it on. I personally had the fight scored 20-15 through two rounds, and I don’t think I’ve ever done that before. Maldonado is amazingly tough, but when I think of this fight, Maldonado reminds me of Danny Downes – a fighter who got clobbered non-stop, and somehow kept fighting through it. When the best thing one can say about a fighter is that he can take a lot of punishment, it’s not a good thing. Teixeira is the real deal, and hopefully we’ll get to see him against a top 10 opponent next time.
-You’re probably expecting me to say “I told you so” about Erick Silva. I’ll be honest – it’s hard not to. But instead, what I’ll do is give Silva some praise. Jon Fitch had a very hard time keeping Silva down in the first two rounds, got hit hard a couple times, and was in serious danger of being submitted in the second round. Silva’s cardio is a problem, and it will likely always be a problem, but he’s going to be a very dangerous opponent for any UFC welterweight in the first 8-10 minutes of a fight. The man is very talented. As for Fitch… all I can say is that I hope he finally gets some respect from the fans after putting on a great fight.
-As expected, Phil Davis was far too much for Wagner Prado. Davis beat Prado the same way he’s beaten almost all his opponents – by dominating them with wrestling and grappling. There were way too many people siding with Prado based on the very short fight he had against Davis two months ago. It’s one thing to get behind a fighter after a spectacular 90-second victory; it’s another to get behind a fighter after 90 seconds of just landing a few strikes.
–Demian Maia looked fantastic against Rick Story, winning by submission for the first time in a long time. When Maia is focused on imposing his jiu-jitsu on his opponents, he’s an elite fighter and extremely hard to beat. When he decides to engage in a boxing match, he struggles. There are not many welterweights who will be able to stop Maia as long as he fights like he did tonight.
-The UFC needs to stop using significant strike data on their broadcasts. For a long time, I tried to use significant strike data for meaningful analysis, and I’ve pretty much given up on it. There are two fundamental problems with it. One is that strikes are divided into “significant” and “not significant” only, and there’s no way to separate those without making it a judgment call. The other is that not all significant strikes are created equal – Stephan Bonnar may have landed some significant strikes on Anderson Silva, but none were nearly as significant as Silva’s knee that ended the fight. Like it or not, determining the performance of a fighter needs to be based on subjectivity, and can’t be done with statistics. And that’s coming from a statistics nerd.
-Maybe I should confine my analysis and predictions to main card fights only. I seem to do a much better job with those than with preliminary fights…
UPDATED ELO RATINGS