If the UFC had a chess-like rating...

I knew Jones would be top but Iam curious what Fedor's peak ELO was during his 28 fight win streak.
 
Владислав IV said:
The Elo rating system can be used as an objective approach to determining a ranking based on the skill level of UFC fighters, as opposed to the silly and arbitrary pound-for-pound rankings.
With a start rating at 100, this is how the current top 15 would look like:
  1. Jon Jones (363.863)
  2. Georges St-Pierre (346.526)
  3. Khabib Nurmagomedov (292.587)
  4. Dustin Poirier (287.057)
  5. Islam Makhachev (286.728)
  6. Kamaru Usman (285.422)
  7. Leon Edwards (280.988)
  8. Charles Oliveira (277.861)
  9. Alexander Volkanovski (274.928)
  10. Beneil Dariush (271.992)
  11. Aljamain Sterling (269.404)
  12. Demetrious Johnson (268.642)
  13. Amanda Nunes (266.690)
  14. Francis Ngannou (263.779)
  15. Robert Whittaker (263.539)
And out of the 2373 fighters that have fought in the UFC, these are the 5 highest recorded ratings:
  1. Jon Jones (363.863)
  2. Georges St-Pierre (346.526)
  3. Kamaru Usman (322.851)
  4. Anderson Silva (315.715)
  5. Tony Ferguson (297.683)
Fun fact:
The biggest mismatch (Elo discrepancy) in UFC history was Demetrious Johnson (260.357) vs Tim Elliott (71.886)

If anyone's interested in the full list check out this table https://nkshv.github.io/
Click to expand...
A couple of things jump out at me. ELO doesn't work so well for MMA since there are weight classes in chess and men don't play women. The great thing about chess is kids play adults, men play women etc...., so it's just about wins and losses and skill level. There are so many factors like injuries and huge dips in performance that don't work for MMA (Renan Barao for example). You don't have those type of huge drop offs so fast in chess, so it's much harder to get an accurate ELO for MMA. It's too subjective. The other glaring things would be the ELO of Amanda Nunes. Her record would imply that she could beat Ngannou and Whittaker. I know it's for skill level but then that's totally subjective. This is fun but probably meaningless.
 
Eternalwarriorking said:
Do the goofy chess analogies from commentary piss you off? I'm a pretty avid chess player and they always grind my gills. lol
Click to expand...

Hahah yes, I know exactly what you’re talking about. My gf actually uses “blunder” all the time now, and it’s so cringe.
 
Liverkick-king56 said:
I'm being trolled right? Joe-B, Cejudo, Dodson, Gooch, Mccall, Moraes, Torres...... None of these guys aren't top level talent.
Click to expand...

Your original post said “Khabib, Aljo, Islam, Dariush and Poirier”, which to all I would say I agree with the Elo rankings. I don’t know why you think that’s a troll.

Now I would disagree with some of the next batch of names, except Cejudo. I have never put MM anywhere near my Mt Rushmore, he just had so little competition for all those years. Nobody could make that weight cut, division almost got scrapped. Respect the hell out of the guy though, and his longevity has been impressive.
 
I like the idea and always hoped for the UFC to do something like this, a point based ranking system.

And it's not flawed if you just see it as a "how many points fighter X achieved" instead of "how good fighter X is."

So for instance, something like a Cerrone would be above someone like Islam. Is he better? No. But fought more and better guys so far.
(just an example, don't know their numbers)
 
3g64.gif


Great thread. Please provide updates if you ever tweak the formula.
 
I'm going to make one of these lists give really exacting criteria and just put whatever number I want next to fighters in an order that grinds people's gears. Like Anderson silva 393 Jon jones 236 Fedor 235 gsp 511.

Then i will be very adamant about the scientific nature of the approach. Aww if only I had the time...
 
I don't think such a rating system is useful for predicting future outcomes nor can it be used to compare fighters who are not in the same division since rating depends upon how large the difference is in rating between opponents. The shallower the division the harder it becomes to compare in any useful fashion.
Ramon Antonio said:
The sample size in MMA is not enough for a proper ELO system to work well.
Click to expand...
I second that.
 
Tokoloshe said:
Glorious idea. Fun to see. There is a video showing chess champions and their ratings evolutions over time. It would be awesome to see the same for the UFC.

Click to expand...

This is how the top 1 would look like over time:

Royce Gracie [1993 - 2005]
148.0 -> 249.051

Chuck Liddell [2005 - 2010]
253.637 -> 279.363

Georges St-Pierre [2010 - 2012]
281.885 -> 303.336

Anderson Silva [2012 - 2013]
306.942 -> 315.715

Georges St-Pierre [2013 - 2020]
321.804 -> 346.526

Jon Jones [2020 - ]
253.890 -> 363.862 (current)
 
Владислав IV said:
Another problem is that some fighters benefit by winning against aging fighters, whose Elo doesn't reflect their current skills. Donald Cerrone as the 6th best fighter of all time is a glaring error and a much better example as to why numbers don't tell the whole story.
Click to expand...
.

Is there a way to acccount for fighting over-the-hill fighters by incorporating the fight record/win% of the opponent at the time a win or loss was tallied against them?
 
Dustin Poirier's makes no fucking sense. Elo higher than Oliveira even after being strangled by him, when his highest elo win is Conor who is at #61 probably inflately as well. Then the fellow sucky gatekeepers Gaethje and Chandler at #101 and fucking #724. Lmaoooo.
 
You must log in or register to reply here.

Latest posts

Forum statistics

Threads
1,280,189
Messages
58,264,687
Members
175,986
Latest member
Dakota DeSousa

Share this page

Back
Top