Dream fight, simulated

Prime Khabib vs Prime GSP: Who Actually Wins?

It is the greatest fight MMA never made. So we made it, 400,000 times, and the engine landed on the one answer nobody wanted to hear: a dead heat, 50-50. But the two men win completely different fights, and that is where it gets interesting.

Written by Harry Ali

Published 16 Jul 2026 · 7 min read · 400,000 simulated fights, verified against the live 30-0 engine

Two mirrored columns of gold light rising inside an octagon and meeting in a burst of sparks at the centre, a dead heat

The greatest fight MMA never made

Two men retired with a claim on the same throne. Georges St-Pierre walked away 26-2, with a record nine welterweight title defences and a second belt won at middleweight against Michael Bisping. Khabib Nurmagomedov walked away at 29-0, having never really been in trouble, let alone beaten. Put their peaks in the same cage and you are not settling a fight. You are settling the greatest-of-all-time argument itself.

It never happened, and the reason is unglamorous: weight. Khabib wanted it at lightweight and would not move; GSP, older and years removed from a serious cut, wanted a 165-pound catchweight. The promotion never pushed. So the fight lived where it still lives, in group chats and comment sections, undefeated on both sides forever. We decided to stop arguing and run it.

The case for Khabib

Nobody solved him. Not the wrestlers, not the strikers, not the one man (Conor McGregor) with the aura to rattle him. Khabib's game was a slow catastrophe: a takedown you could see coming and still could not stop, then a chest on your chest and a forearm on your jaw until the round, and eventually your will, ran out. In our ratings his grappling is a 99 and his submission defence a 97, the highest marks in the game, and his cardio (96) means the maul never slows down.

Against GSP, the plan does not change, because it never changed. Get it to the mat, make it ugly, drown the clever man in deep water. If Khabib gets top position and keeps it, no amount of fight IQ scores you points from the bottom.

The case for GSP

Here is the problem for that plan: GSP is the one welterweight who might win the wrestling. He out-wrestled decorated wrestlers for a decade, and he did it with a jab in your face and a level change you never saw start. Our GSP card is the more complete striker by a distance, a 90 to Khabib's 80, with the fight IQ to turn 25 minutes into a maths problem the other man cannot solve.

GSP does not need to stop the takedown every time. He needs to stop it enough, and to make Khabib pay in transition and on the feet for every attempt that fails. The blueprint that beat everyone at welterweight is the same one that could frustrate a lightweight climbing to meet him: control the distance, win the exchanges, and never, ever panic.

GSP vs Khabib: where the fight is won

Strip away the names and the fight is a single question: can Khabib get GSP down and keep him there, or does GSP's takedown defence and his own wrestling turn it into a kickboxing match he is better equipped to win? Every other thread hangs off that one. Khabib's edge is a razor: his grappling is a 99 to GSP's 97, his submission defence a 97 to GSP's 95. Real, but thin. GSP's edge, the ten-point striking gap, is the wider one, but it only matters in the seconds the fight is standing.

So we ran it, hundreds of thousands of times, and let those margins fight it out.

400,000 fights between the prime cards, run in both corners so neither gets a home advantage. It splits down the middle.

Fifty-fifty. Not a cop-out, a finding: with two 97s this evenly matched, the model cannot separate them, which is exactly what a decade of the argument has been telling us. But a coin flip is where most people stop, and it is the least interesting part. Look at how each man wins:

The same 50-50, told two ways. Khabib's wins come on the mat. GSP's come on the scorecards. Neither man beats the other the way he beats everyone else.

That is the real story the number hides. A Khabib win is 25 minutes of top control ending in a tap or a shutout. A GSP win is a clinic: jabs, timed takedowns of his own, and a decision so lopsided the judges barely have to think. They arrive at the same 50-50 from opposite directions, which is the most fitting outcome this fight could possibly have.

The one thing the engine cannot see

Here is where I stop trusting my own model. The engine rates skill, and on skill these two are a coin flip. It does not rate size, and size is the exact factor most likely to break the tie. GSP is the bigger man, and it is not close. He fought his entire career at welterweight and won a belt as high as middleweight. Khabib never competed an ounce above lightweight. At any weight they could have agreed on, GSP is the one rehydrating into the cage as the larger, stronger human being, and in a fight this even, being the bigger man is not a tiebreaker you wave away.

So if you want my actual pick, past the numbers: at a fair catchweight, I lean GSP, narrowly, on size and the wider striking gap. At a true 155 that Khabib would never surrender, I lean Khabib. Which is to say the weight class was never a technicality. It was the fight.

Here is the uncomfortable version nobody says out loud: the fight never happened because, deep down, both camps knew what the simulation just found. It is a coin flip. And you do not risk a 29-0, or the cleanest GOAT resume in the sport, on a coin flip. The 0 was worth more to both men than the answer.

Harry "Haroon" Ali

MMA enthusiast · dabbled in boxing, kickboxing, muay thai, jiu jitsu & judo

So who wins?

On the numbers, nobody. It is the rarest thing a dream fight can be: genuinely even, a real 50-50, decided by which cage they agree to walk into and which day it is. The engine settles the skill and refuses to settle the fight, because the fight was never settleable. That is not a failure of the model. It is the correct answer, and it is the reason two of the greatest to ever do it will stay undefeated against each other for good.

The ratings behind all of this are ours, built from public records and results only; the method is in how we rate fighters. And if you think the cards are wrong, there is a cage where you can prove it.

Draft them. Settle it yourself.

Both cards are in the game. Build the camp, run the fights, and find out whose blueprint survives 30 of them. Free, no account, about two minutes.

Play 30-0 free

Frequently asked questions

Who would win, Khabib or GSP?

In our simulation of 400,000 fights between the prime versions of each, it is a dead heat: 50% to 50%. But the two men win different fights. Khabib wins mostly by submission or grinding decision; GSP wins mostly on points, outboxing and out-positioning over 25 minutes.

Why did Khabib and GSP never fight?

A weight standoff. Khabib wanted the fight at lightweight (155 lb) and refused to move up; GSP, older and coming off a long layoff, did not want to cut back down to 155 and preferred a 165 lb catchweight. The UFC never seriously pursued the bout, so it stayed a what-if.

Is Khabib or GSP the MMA GOAT?

Both are in the conversation. GSP retired 26-2 with a record nine welterweight title defences and a second title at middleweight. Khabib retired a perfect 29-0. In our ratings both hold the joint-top overall of 97, one of only five cards rated that high across the whole game.

Was GSP bigger than Khabib?

GSP is the naturally bigger man: he fought his whole career at welterweight (170) and won a title as high as middleweight (185), while Khabib never competed above lightweight (155). Off-season the gap narrows, but on fight night at any shared weight GSP would likely be the larger, stronger fighter. Our engine rates skill, not size, which is the one edge it cannot see.

Harry "Haroon" Ali

Creator of 30-0 and writer of The Corner. A lifelong MMA fan who has dabbled in boxing, kickboxing, muay thai, jiu jitsu and judo, and who built the engine these fights were run on. More about Harry.