AI crowd · UFC Fight Night: Buckley vs. Malott
UFC Edmonton AI crowd: Claude models back Malott, GPT models back Buckley
Our 100 AI forecasters lean Mike Malott in the main event at 52%, but the split runs almost exactly along company lines. The crowd also sides with Timmy Cuamba, whom the odds have as the underdog.
The setup
100 AI forecasters, played by five AI models, each playing 20. Each one reads the same fact sheet on every fighter (record, size, career stats, recent 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 Sonnet 5.5Anthropic · 20 forecasters
- Claude Opus 5.5Anthropic · 20 forecasters
- Claude Fable 5.1Anthropic · 20 forecasters
- AstraOpenAI · 20 forecasters
- GPT-6.1 SolOpenAI · 20 forecasters
The angles
Each forecaster is told to weigh the fights from one of these ten; "no angle" gets only the fact sheet and the question. Half see each fight with the fighters listed the other way round, so the order can't tilt the result.
- No angle
- The stand-up
- The grappling
- The numbers
- Durability and cardio
- Age and time away
- Recent form
- The matchup
- The sceptic's case
- The judging
The AI crowd takes Mike Malott over Joaquin Buckley in Saturday's main event, but only just. The average chance came out at 52%, and 64 of 100 forecasters picked him.
The crowd is 100 forecasters built on three Claude models from Anthropic and two GPT models from OpenAI, all reading the same fact sheet without seeing the odds.
Last week, at UFC Fight Night: Allen vs. Duncan, the crowd called 5 of 11 winners. Our model and the betting favourites both called 6 of 11. The forecasters given no angle managed only 4 of 11. The crowd's worst miss was Felipe Franco at 71%, who was submitted by Brendson Ribeiro at 1:49 of round 1.
The main event splits by company
The Claude forecasters went 58 of 60 for Malott. The GPT forecasters went 34 of 40 for Buckley. Every Claude Opus 5.5 and Claude Sonnet 5.5 forecaster picked Malott, while 19 of 20 Astra forecasters picked Buckley.
Each camp put its man at 57%. So the 52% headline is two confident groups cancelling out, not 100 forecasters shrugging.
The Claude case leaned on recent results. "Malott has won four straight, two by KO. Buckley landed just 5 of 31 strikes against Brady and is on two straight losses," said a Claude Opus 5.5 forecaster going by the numbers. The GPT case was about style: "Buckley's reach and power threaten Malott, who has not shown the sustained wrestling that neutralized Buckley in his last two fights," said an Astra forecaster weighing the matchup.
The betting market has Malott at 54% and our model has him at 53%, so the Claude side sits closer to both.
Where the crowd goes against the odds
The biggest gap on the card is Louis Jourdain vs Timmy Cuamba. The betting market and our model both make Jourdain a 64% favourite. The crowd went the other way almost to a forecaster: 98 of 100 picked Cuamba, at 57% on average.
The reasons were about experience. "Cuamba has 5 UFC fights of experience against a debutant who has been out for over a year, with a thin regional record," said a Claude Sonnet 5.5 forecaster given no angle.
The crowd also picks SuYoung You over Cody Chovancek, 68 of 100 at 54%, while the odds and our model favour Chovancek at 56%.
Opus and Sonnet go their own way
That You pick is really a GPT pick. All 40 GPT forecasters took You, but Claude Opus 5.5 went 17 of 20 for Chovancek and Claude Sonnet 5.5 went 15 of 20. "Chovancek is unbeaten at 9-0 with real finishing power and a home crowd," said a Claude Sonnet 5.5 forecaster weighing the stand-up.
The same pattern shows up at heavyweight. The GPT forecasters went 38 of 40 for Jhonata Diniz, while Opus (17 of 20) and Sonnet (16 of 20) picked Tanner Boser. The betting market has Diniz at 68% and our model at 71%, so the crowd's 53% for him is far less sure.
Sure, but never very sure
The crowd's firmest calls are Melissa Croden, picked by all 100 forecasters, and Erin Blanchfield, picked by 97 of 100. Both sit at 68%, the highest number the crowd gave anyone. Kyle Daukaus was picked by 99 of 100 but only at 61%.
That restraint stands out next to the odds. The betting market has Nick Galanti at 82%; the crowd, with 97 of 100 picking him, says 61%.
| Fight | AI crowd | Our model | Betting market |
|---|---|---|---|
| Buckley vs Malott | Malott 52% | Malott 53% | Malott 54% |
| Blanchfield vs Jasudavicius | Blanchfield 68% | Blanchfield 81% | Blanchfield 73% |
| Nelson vs Perez | Perez 57% | Perez 78% | Perez 75% |
| Barriault vs Daukaus | Daukaus 61% | Daukaus 72% | Daukaus 64% |
| Jourdain vs Cuamba | Cuamba 57% | Jourdain 64% | Jourdain 64% |
| Nallo vs Landwehr | Nallo 59% | Nallo 73% | Nallo 70% |
| Boser vs Diniz | Diniz 53% | Diniz 71% | Diniz 68% |
| Leblanc vs Galanti | Galanti 61% | — | Galanti 82% |
| Mahjoub vs Faglier | Mahjoub 57% | Faglier 54% | Mahjoub 51% |
| Cerminara vs Stiller | Cerminara 57% | Cerminara 70% | Cerminara 70% |
| Anheliger vs Koslow | Koslow 53% | Koslow 60% | Koslow 56% |
| Croden vs Chandler | Croden 68% | Croden 76% | Croden 70% |
| Chovancek vs You | You 54% | Chovancek 56% | Chovancek 56% |
Our model works from fight stats only and never sees the odds. The betting market's chance has the bookmakers' margin removed. Both are as of Oct 13, 10:46 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, angle by angle
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 5.1 | Astra | GPT-6.1 Sol |
|---|---|---|---|---|---|---|
| Buckley v Malott | 52 for Malott | 58 for Malott | 59 for Malott | 52 for Malott | 61 for Buckley | 60 for Buckley |
| Blanchfield v Jasudavicius | 68 for Blanchfield | 66 for Blanchfield | 69 for Blanchfield | 72 for Blanchfield | 72 for Blanchfield | 72 for Blanchfield |
| Nelson v Perez | 57 for Perez | 59 for Perez | 58 for Nelson | 56 for Perez | 60 for Perez | 62 for Perez |
| Barriault v Daukaus | 61 for Daukaus | 58 for Daukaus | 62 for Daukaus | 62 for Daukaus | 63 for Daukaus | 64 for Daukaus |
| Jourdain v Cuamba | 57 for Cuamba | 56 for Cuamba | 64 for Cuamba | 50 for neither | 58 for Cuamba | 57 for Cuamba |
| Nallo v Landwehr | 59 for Nallo | 59 for Nallo | 52 for Nallo | 59 for Nallo | 60 for Nallo | 62 for Nallo |
| Boser v Diniz | 53 for Diniz | 54 for Boser | 50 for neither | 59 for Diniz | 60 for Diniz | 61 for Diniz |
| Leblanc v Galanti | 61 for Galanti | 57 for Galanti | 54 for Galanti | 63 for Galanti | 61 for Galanti | 61 for Galanti |
| Mahjoub v Faglier | 57 for Mahjoub | 56 for Mahjoub | 66 for Mahjoub | 58 for Mahjoub | 56 for Mahjoub | 56 for Mahjoub |
| Cerminara v Stiller | 57 for Cerminara | 54 for Cerminara | 54 for Cerminara | 60 for Cerminara | 65 for Cerminara | 64 for Cerminara |
| Anheliger v Koslow | 53 for Koslow | 55 for Anheliger | 58 for Anheliger | 55 for Koslow | 57 for Koslow | 58 for Koslow |
| Croden v Chandler | 68 for Croden | 69 for Croden | 67 for Croden | 65 for Croden | 69 for Croden | 72 for Croden |
| Chovancek v You | 54 for You | 55 for Chovancek | 56 for Chovancek | 58 for You | 62 for You | 62 for You |
Angle by angle
The same for each angle the forecasters were told to weigh, every model together. “No angle” is the forecasters given only the fact sheet and the question.
| Fight | No angle | Stand-up | Grappling | Numbers | Durability | Age, time off | Form | Matchup | Sceptic | Judging |
|---|---|---|---|---|---|---|---|---|---|---|
| Buckley v Malott | 50 for neither | 53 for Malott | 53 for Malott | 53 for Malott | 50 for neither | 52 for Malott | 55 for Malott | 50 for neither | 50 for neither | 52 for Malott |
| Blanchfield v Jasudavicius | 70 for Blanchfield | 69 for Blanchfield | 65 for Blanchfield | 67 for Blanchfield | 70 for Blanchfield | 65 for Blanchfield | 68 for Blanchfield | 68 for Blanchfield | 67 for Blanchfield | 67 for Blanchfield |
| Nelson v Perez | 56 for Perez | 58 for Perez | 55 for Perez | 57 for Perez | 58 for Perez | 56 for Perez | 57 for Perez | 59 for Perez | 55 for Perez | 55 for Perez |
| Barriault v Daukaus | 62 for Daukaus | 57 for Daukaus | 63 for Daukaus | 61 for Daukaus | 58 for Daukaus | 60 for Daukaus | 62 for Daukaus | 62 for Daukaus | 60 for Daukaus | 62 for Daukaus |
| Jourdain v Cuamba | 57 for Cuamba | 56 for Cuamba | 57 for Cuamba | 57 for Cuamba | 56 for Cuamba | 58 for Cuamba | 57 for Cuamba | 57 for Cuamba | 56 for Cuamba | 56 for Cuamba |
| Nallo v Landwehr | 58 for Nallo | 60 for Nallo | 58 for Nallo | 59 for Nallo | 60 for Nallo | 59 for Nallo | 58 for Nallo | 60 for Nallo | 57 for Nallo | 59 for Nallo |
| Boser v Diniz | 55 for Diniz | 54 for Diniz | 53 for Diniz | 52 for Diniz | 53 for Diniz | 51 for Diniz | 54 for Diniz | 55 for Diniz | 53 for Diniz | 53 for Diniz |
| Leblanc v Galanti | 59 for Galanti | 60 for Galanti | 63 for Galanti | 60 for Galanti | 61 for Galanti | 59 for Galanti | 61 for Galanti | 61 for Galanti | 60 for Galanti | 62 for Galanti |
| Mahjoub v Faglier | 58 for Mahjoub | 54 for Mahjoub | 60 for Mahjoub | 54 for Mahjoub | 55 for Mahjoub | 58 for Mahjoub | 54 for Mahjoub | 59 for Mahjoub | 58 for Mahjoub | 58 for Mahjoub |
| Cerminara v Stiller | 59 for Cerminara | 59 for Cerminara | 52 for Cerminara | 59 for Cerminara | 57 for Cerminara | 55 for Cerminara | 57 for Cerminara | 58 for Cerminara | 56 for Cerminara | 61 for Cerminara |
| Anheliger v Koslow | 52 for Koslow | 50 for neither | 56 for Koslow | 54 for Koslow | 51 for Koslow | 52 for Koslow | 51 for Koslow | 55 for Koslow | 53 for Koslow | 54 for Koslow |
| Croden v Chandler | 68 for Croden | 70 for Croden | 69 for Croden | 69 for Croden | 69 for Croden | 67 for Croden | 67 for Croden | 69 for Croden | 67 for Croden | 69 for Croden |
| Chovancek v You | 54 for You | 51 for You | 53 for You | 57 for You | 52 for You | 56 for You | 53 for You | 53 for You | 55 for You | 55 for You |
How the AI crowd works
The crowd is 100 AI forecasters, played by Claude Opus 5.5, Claude Sonnet 5.5 and Claude Fable 5.1 from Anthropic, and Astra and GPT-6.1 Sol from OpenAI. Most were told to weigh the fights from one of nine angles: the stand-up, the grappling, the numbers, durability and cardio, age and time away, recent form, the matchup, the sceptic's case and the judging. The rest got no angle at all, only the fact sheet and the question.
All of them read the same fact sheet: each fighter's record, size, career stats and recent UFC fights from our database, plus this week's news, 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: Buckley vs. Malott: the full card, odds and results →