AI crowd · UFC Fight Night: Adesanya vs. Magomedov
AI crowd narrowly picks Magomedov at UFC Fight Night: Adesanya vs. Magomedov
The AI crowd gives Shara Magomedov a 55% chance against Israel Adesanya, while the betting market has Adesanya favoured. Its most confident pick is Mario Pinto.
The setup
10 AI forecasters, played by three AI models. Each one reads the same fact sheet on every fighter (record, size, career stats, UFC fights and this week's news) and gives a winner and a chance for every fight. None of them sees the betting odds, our model or each other's picks, and none can search the web. The crowd's number is the average of their chances.
The models
- Claude Opus 5.5Anthropic ยท 3 forecasters
- Claude Sonnet 5.5Anthropic ยท 4 forecasters
- Claude Fable 5.2Anthropic ยท 3 forecasters
The AI crowd has Shara Magomedov beating Israel Adesanya, but only just. It puts his chance at 55%, with 6 of 10 forecasters picking him. This week's crowd is 10 AI forecasters, all Claude models from Anthropic, each given the same fact sheet and no odds or web search.
On the last card, UFC Fight Night: Prochazka vs. Stirling, the crowd got 6 of 11 fights right. The betting favourites went 7 of 11, and our model was right on 2 of 6. The crowd's biggest miss was Jake Matthews, its pick at 65%, who lost a unanimous decision to Tahir Abdullayev.
| Fight | AI crowd | Our model | Betting market |
|---|---|---|---|
| Adesanya vs Magomedov | Magomedov 55% | ๐ | Adesanya 59% |
| Garry vs Amosov | Garry 66% | ๐ | Garry 62% |
| Spivac vs Pinto | Pinto 72% | ๐ | โ |
| Chokheli vs Dulatov | Dulatov 56% | ๐ | โ |
| Santos vs Abdullaev | Santos 54% | ๐ | โ |
| Tuchalov vs Cutelaba | Tuchalov 60% | ๐ | โ |
| Silva vs Filho | Filho 57% | ๐ | โ |
| Satybaldiev vs Gnizditskiy | Even | ๐ | โ |
| Tulio vs Park | Park 52% | ๐ | โ |
| Andrusca vs Osmanli | Andrusca 66% | ๐ | โ |
| Rodrigues vs Duclos | Duclos 54% | ๐ | โ |
| Hussein vs Sosa | Hussein 51% | ๐ | โ |
Our model and the betting market as of Nov 21, 10:05 PM ET.
Every forecaster, fight by fight
Each dot is one AI forecaster, placed by the chance it gave its pick. Left of the middle line it picked the red-corner fighter, right of it the blue corner. Dots pile up in 5-point steps, Claude models above the line and GPT models below it.
Model by model
Model by model
Each model's own call, from its forecasters given no angle, next to the whole crowd. The number is the chance it gave its pick; red means it picked the red-corner fighter, blue the blue corner, and the deeper the colour, the surer it was.
| Fight | AI crowd | Opus 5.5 | Sonnet 5.5 | Fable 5.2 | GPT-6.2 Astra | GPT-6.2 Sol |
|---|---|---|---|---|---|---|
| Adesanya v Magomedov | 55 for Magomedov | 50 for neither | 64 for Magomedov | 52 for Adesanya | 58 for Adesanya | 56 for Adesanya |
| Garry v Amosov | 66 for Garry | 63 for Garry | 69 for Garry | 66 for Garry | 61 for Garry | 51 for Garry |
| Spivac v Pinto | 72 for Pinto | 76 for Pinto | 70 for Pinto | 71 for Pinto | 66 for Pinto | 69 for Pinto |
| Chokheli v Dulatov | 56 for Dulatov | 63 for Dulatov | 53 for Chokheli | 63 for Dulatov | 55 for Dulatov | 58 for Dulatov |
| Santos v Abdullaev | 54 for Santos | 62 for Santos | 50 for neither | 52 for Santos | 61 for Abdullaev | 58 for Abdullaev |
| Tuchalov v Cutelaba | 60 for Tuchalov | 64 for Tuchalov | 56 for Tuchalov | 61 for Tuchalov | 54 for Cutelaba | 50 for neither |
| Silva v Filho | 57 for Filho | 59 for Filho | 62 for Filho | 51 for Filho | 58 for Filho | 60 for Filho |
| Satybaldiev v Gnizditskiy | 50 for neither | 53 for Gnizditskiy | 52 for Satybaldiev | 50 for neither | 58 for Gnizditskiy | 57 for Gnizditskiy |
| Tulio v Park | 52 for Park | 51 for Tulio | 56 for Park | 51 for Tulio | 58 for Park | 54 for Park |
| Andrusca v Osmanli | 66 for Andrusca | 73 for Andrusca | 64 for Andrusca | 63 for Andrusca | 72 for Andrusca | 70 for Andrusca |
| Rodrigues v Duclos | 54 for Duclos | 57 for Duclos | 58 for Duclos | 55 for Rodrigues | 55 for Duclos | 56 for Duclos |
| Hussein v Sosa | 51 for Hussein | 51 for Sosa | 52 for Hussein | 51 for Hussein | 50 for neither | 56 for Hussein |
The main event comes down to one model
All 4 Claude Sonnet 5.5 forecasters picked Magomedov, at an average of 64%. The other two models weren't convinced. Claude Opus 5.5 came out even at 50%, though 2 of its 3 forecasters took Adesanya. Claude Fable 5.2 had Adesanya at 52%, also with 2 of 3.
A Claude Sonnet 5.5 forecaster made the case for Magomedov at 66%. "Adesanya, 37, has lost three straight by stoppage; Magomedov is younger, more active and accurate despite no five-round experience and being blind in one eye."
A Claude Fable 5.2 forecaster went the other way, with Adesanya at 56%. "Seven-inch reach edge and five-round experience against a non-wrestling striker Michael Page outpointed from range; Adesanya's recent stoppage losses keep it close."
Where the crowd and the odds part ways
The betting market has Adesanya at 59%, so it and the crowd favour different men in the main event.
They agree on Ian Machado Garry against Yaroslav Amosov. The crowd has Garry at 66% and the market has him at 62%. All 10 forecasters picked him, and a Claude Fable 5.2 forecaster at 68% pointed to his takedown defence. "Stopped all seven Belal Muhammad takedowns and survived Makhachev and Rakhmonov; much sharper striker than Amosov at distance."
Surest on Pinto, stuck on Satybaldiev and Gnizditskiy
Mario Pinto over Serghei Spivac is the crowd's most confident pick, at 72% and 10 of 10. A Claude Opus 5.5 forecaster had him at 79%. "Takes opponents down and finishes on the ground; Spivac has 53% takedown defence and just lost every round to Petrino." Pavel Andrusca, at 66% over Mahammadali Osmanli, and Magomed Tuchalov, at 60% over Ion Cutelaba, were also picked by all 10.
Uran Satybaldiev against Ivan Gnizditskiy has no crowd pick. Satybaldiev was picked by 6 of 10 forecasters, but the 4 on Gnizditskiy were surer, and the average came out even. Abdul Hussein at 51% over Hecher Sosa and JunYong Park at 52% over Marco Tulio were nearly as tight.
Sonnet and Fable each go it alone
Sonnet also stood apart on Levan Chokheli against Islam Dulatov. All 4 of its forecasters took Chokheli, at an average of 53%. All 3 from Opus and all 3 from Fable took Dulatov, and both models averaged 63%. That leaves Dulatov as the crowd's pick at 56%.
Fable had its own disagreement on Modestino Rodrigues against Matthieu Duclos. Its 3 forecasters all picked Rodrigues, at 55%. Every Opus and Sonnet forecaster took Duclos, at 57% and 58%, and the crowd settled on him at 54%.
How the AI crowd works
The crowd is 10 AI forecasters, played by Claude Opus 5.5, Claude Sonnet 5.5 and Claude Fable 5.2 from Anthropic. None of them was told how to weigh the fights: each got only the fact sheet and the question.
All of them read the same fact sheet: each fighter's record, size, career stats and UFC fights from our database, plus this week's news (the reports, and what the fighters and their coaches said), gathered once for each fight. None of them saw the betting odds, our model or each other's picks, and none could search the web. Half of them saw each fight with the fighters in the opposite order, so the order they're listed in can't tilt the result.
Each forecaster picked a winner in every fight and gave a chance. The crowd's number is the average of those chances. It's an experiment in how AI reads fights, not betting advice.
The AI crowd: every card so far, and how it works →
UFC Fight Night: Adesanya vs. Magomedov: the full card, odds and results →