UFC This Weekend — AI Predictions for Saturday’s Main Card

UFC Fight Night heads to Winnipeg this Saturday, April 18, with Gilbert Burns vs Mike Malott headlining at the Canada Life Centre. It's a Fight Night card — no title fights — but the main card features several matchups where the model sees clear edges. Our UFC prediction accuracy sits at 68.0% (17 correct out of 25 graded picks) this season, making combat sports one of our strongest categories. Here's every pick for Saturday's card with confidence scores and method-of-victory projections.

Full predictions with live updates are on the UFC AI predictions page.

Main Event: Gilbert Burns vs Mike Malott

The headliner is a welterweight bout between Gilbert Burns, the veteran former title challenger, and Mike Malott, the Canadian fighting on home soil.

The model is split on this one. Across our prediction entries, confidence ranges from 65-68% — some favoring Burns by unanimous decision, others leaning Malott. The honest read: Burns at 65% confidence by decision, but this is a genuine coin-flip-plus fight where the hometown factor for Malott adds uncertainty the model can't fully quantify.

Burns's path to victory runs through his grappling — he's one of the best Brazilian jiu-jitsu practitioners in the UFC and can neutralize Malott's aggression on the mat. Malott's path is through pressure and volume, using the Winnipeg crowd to fuel a pace that tires Burns in the championship rounds.

At 65% confidence, this is a fight where the value bet finder is most useful — check whether the sportsbook odds imply a wider gap than the model sees. If the market has Burns at 55% implied probability and the model sees 65%, the moneyline carries value. If the market already agrees, there's no edge on the winner — look at method of victory or round props instead.

Main Card Predictions

Kyler Phillips vs Charles Jourdain — Model: Jourdain at 68% confidence. Projected method: submission in round 2. Jourdain's submission game has been sharp, and Phillips has shown vulnerability on the ground in recent outings. This is the model's strongest lean on the main card outside of the headliner.

Jasmine Jasudavicius vs Karine Silva — Model: Jasudavicius at 70% confidence. Projected method: decision. Jasudavicius's wrestling-heavy approach should control the pace against Silva. The confidence is moderate — 70% in a women's flyweight fight reflects the model's read that Jasudavicius has the tools to win but the margin isn't large.

Thiago Moises vs Gauge Young — Model: Moises at 65% confidence. Projected method: submission in round 2 or 3. Moises is the more experienced fighter and his submission game is the clear path to victory. Young is dangerous on the feet, which keeps the confidence from climbing higher.

Dennis Buzukja vs Marcio Barbosa — Model: Barbosa at 78% confidence. One of the higher-confidence picks on the card. The model sees Barbosa's skill set as a significant mismatch in this matchup.

All main card predictions with live updates are on the UFC predictions page.

Prelim Highlights

Mitch Raposo vs Allan Nascimento — Model: Nascimento at 85% confidence. The highest confidence pick on the entire card. The model sees a clear skill gap here, and Nascimento's grappling should dominate the matchup. At 85%, this is the closest thing to a lock on the Winnipeg card.

Jamie Siraj vs John Yannis — Model: Siraj at 82-85% confidence. Another high-confidence prelim pick. Siraj's striking volume and pressure style should overwhelm Yannis across three rounds.

JJ Aldrich vs Jamey-Lyn Horth — Model: Horth at 75-80% confidence. The model likes Horth in this women's strawweight bout. Aldrich is the more experienced fighter, but Horth's recent improvement and physical advantages give her the edge.

Tanner Boser vs Gokhan Saricam. The model is split — some entries favor Boser at 75% (fighting at home in Canada), others favor Saricam at 65% by KO/TKO in round 2. Heavyweight fights are inherently volatile because one punch can end it. At these confidence levels, the model is essentially saying “lean Boser, but anything can happen.”

Value Angles for Saturday

Here's where the model sees the best betting value on the Winnipeg card.

Nascimento at 85% is the highest-confidence pick and likely the best straight bet on the card. If the market has Nascimento at -200 or shorter (implying 67%), the model sees significantly more edge than the line suggests.

Jourdain by submission as a method-of-victory prop could carry value. If the sportsbook prices Jourdain by submission at +300 or longer but the model's primary projection is exactly that outcome, there's a meaningful probability gap.

Parlay consideration: Nascimento + Siraj + Horth as a three-leg parlay of the model's highest-confidence prelim picks. All three are at 75%+ confidence. A three-leg parlay at those confidence levels has an implied win probability of roughly 48-52% — not a lock, but better than most three-leg parlays. Build it through the parlay builder for correlation analysis.

How Our UFC AI Works

Our UFC model evaluates fighter matchups using recent form, striking and grappling differentials, reach and physical advantages, stylistic matchup dynamics (striker vs grappler, pressure vs counter-fighter), and real-time data via search grounding. The model generates both a winner prediction with confidence score and a projected method of victory when the data supports it. For the full methodology breakdown, read our UFC AI Predictions accuracy report — where we went 17/24 correct at UFC 327.

Track the Results

After Saturday's card, every prediction will be graded automatically — hits and misses, all public. You can track the results on the predictions page and see how this card affects our overall UFC accuracy on the UFC predictions hub. For context on what the confidence scores mean, read How to Read AI Confidence Scores. For how we grade every prediction, see How We Grade Predictions.

Winnipeg. Saturday night. Let's see who the AI has right.