Last week the model went 7–2–1, pushing the 2026 record to 45–18–6 — 71.43% on the year. One of the strongest weeks of the season, and the kind of run that makes the process worth trusting.

This week, the UFC heads to Winnipeg for a lighter card than we’ve seen recently. Only two fights making it into the breakdown this week, which actually creates a perfect opportunity to do something a little different. Before we get into the numbers on the card, we’re pulling back the curtain on how the “Fighting With Numbers” model actually works, the model nicknamed “El Modelo” by Dave Ross. How things like the xR%, the tier system, and how win probability translates into actionable betting decisions. If you’ve been following the picks all year and wondering what’s driving them, this is the week to find out.

No First Strike this week, but we’ll be back next week with Dave Ross for a brand new episode. In the meantime, the numbers don’t take a week off, so we won’t.

Let’s dig in.

Under The Model Hood

Every week this column spits out win probabilities, tier labels, and betting recommendations. But what’s actually happening behind the numbers? Let’s walk through it.

The model is built on UFC fight data. Not promotional records, not reputation, not narrative, just straight octagon data. What a fighter actually does inside the cage, measured consistently across their UFC career, is what feeds the machine. The goal is simple: identify which fighter is more likely to win a given matchup based on how they’ve actually performed, and then compare that probability against what the sportsbooks are implying with their odds.

The xR% — Expected Round Percentage

One of the first numbers you’ll see in every breakdown is the xR%, and it’s probably my favorite single metric in the model.

xR% measures how often a fighter wins the rounds they compete in, based on the statistical profile of each round rather than the judges’ scorecards. It’s an expected value — hence the “x” — meaning it’s calculated from the underlying performance data rather than the official result.

Think of it as a fighter’s baseline competitiveness. A fighter sitting at 70% xR% is winning roughly seven out of every ten rounds they fight. A fighter at 50% is essentially a coin flip every round. Drop below 50% and you’re looking at a fighter who is statistically losing more rounds than they win, a red flag that shows up in the breakdown every time.

The xR% becomes especially useful when evaluating fighters whose records don’t tell the full story. A fighter can be 5–1 in the UFC and still carry a 48% xR% if their wins came via early finish while they were getting outworked in the rounds that did happen. For example, there’s a reason a fighter like Jiri Prochazka carries a 53% xR%. Conversely, a fighter can absorb a few losses and still carry a strong xR% if they were competitive in those fights. The record tells you outcomes. The xR% tells you process.

Win Probability and What It Actually Means

Once the model runs both fighters’ profiles against each other, accounting for striking output, differentials, grappling control, takedown metrics, finishing ability, and the xR%…it produces a win probability for each fighter.

That number is not a prediction. It’s a frequency estimate. A 70% win probability means the model believes that fighter wins this matchup seven times out of ten if you ran it a hundred times. It does not mean they win on Saturday. It means the edge is real and consistent.

The win probability only becomes actionable when you compare it to the implied probability baked into the betting odds. That’s the gap we’re hunting every week.

If the model gives a fighter 65% and the sportsbooks are pricing them at 50% implied, boom, that’s an edge. If the model gives a fighter 65% and the sportsbooks are already at 75% implied, well now we stay away, regardless of who the model thinks wins. We’re not just picking winners. We’re finding where the market is wrong.

The Tier System

Not every model output is created equal, and the tier system is how we communicate confidence level alongside the win probability.

A Full Confidence score means the model sees a clear, well-supported edge. The data is clean, the sample size is sufficient, and the gap between the model’s probability and the implied odds is meaningful. These are the plays worth putting money behind. I will find a way to have a Full Confidence fighter on my betting slip. Traditionally labeled as Tier 1 Elite Fav.

A Live Dog label means the model likes an underdog with real winning potential at the current price. The edge exists, but there’s enough variance in the matchup that we’re not treating it like a lock. These are calculated shots, worth taking at the right price, worth passing on if the line moves. Tier 1 Live Dog.

A Caution grade means the model has a lean, but something in the data gives it pause. Maybe the sample size is thin. Maybe the opponent brings a stylistic threat the numbers don’t fully capture. Maybe the confidence level just isn’t high enough to justify the price. Caution spots can still be played, but with reduced sizing and eyes open. Tier 2 Strong Fav and Tier 3 Volatile Fav for favorites. Tier 2 Puncher’s Chance for underdogs.

A Stay Away means exactly that. The model may have a lean, but the odds don’t support the play. Either the favorite is too expensive, the edge is too small, or the variance is too high to justify a bet. The model having a lean doesn’t mean we have to act on it. Tier 4 Fragile Fav or Tier 5 Trap Fav. You’ll also see the occasional Tier 3 Dead Dog for those plus money fighters.

Why This Matters for a Light Card

A low data fight card like UFC Winnipeg is actually where the tier system earns its keep. There’s no pressure to manufacture action across 13 fights. The model either finds an edge or it doesn’t, and this week it finds one clearly and dismisses the other. That’s the discipline that keeps a 71.43% record moving in the right direction.

Now let’s get into some fight numbers.

Embed from Getty Images

Kyler Phillips vs Charles Jourdain

A fun stylistic puzzle in this one.

Kyler Phillips enters on a two-fight losing streak, most recently dropping a decision to Vinicius Oliveira back in July of 2025. Since his UFC debut in 2020 he’s gone 6–3 in the promotion. Across from him is Charles “Air” Jourdain, back-to-back submission victories now on his record, most recently finishing Davey Grant in November. He sits at 8–7–1 across his UFC career dating back to 2019.

Phillips holds the physical edge — three inches of reach at 72 inches — and the xR% tells a clear story. Phillips checks in at 68%, a strong mark that reflects consistent round-winning. Jourdain sits at 52%, dangerously close to the red flag zone. The fact that he’s on a win streak despite that number tells you those submission finishes are doing a lot of the heavy lifting. He’s getting bailed out before the scorecards matter.

The finishing profiles are worth noting. Jourdain has six UFC wins by knockout or submission, scored seven knockdowns, but has also been dropped four times and carries losses by both KO and submission. Phillips, by contrast, has never been finished inside the UFC octagon.

The striking numbers are close. Phillips lands 4.95 significant strikes per minute with a +1.12 differential. Jourdain edges him in volume at 5.48 per minute with a +1.23 differential, though he absorbs slightly more as a result at 4.25 per minute compared to Phillips’ 3.85.

The grappling is where the gap widens. Phillips holds a 61% control rate, has scored 19 takedowns at 45% accuracy, and defends takedowns at 75%. Jourdain checks in with a 43% control rate, only four takedowns at 21% accuracy, and a concerning 47% takedown defense. We know he’s not afraid of the ground, his eight submission attempts and four submission victories prove that, but he could be on the wrong end of the grappling exchanges more often than not against Phillips.

The model runs it through and delivers a 72.50% win probability for Kyler Phillips, earning him a Live Dog designation. His odds have moved since I locked in my play, but +120 currently still represents real value against the model’s number. We got in at +145 earlier in the week, which is an even better number, but the edge remains at the current price.

Kyler Phillips is on the slip.

Gilbert Burns vs Mike Malott

Once upon a time, Gilbert Burns was one of the more dangerous welterweights on the roster. He challenged for the title in 2021, pushed Kamaru Usman harder than most and built a reputation as one of the most well-rounded fighters in the division. That version of Burns was genuinely special.

This is 2026, and the numbers tell a different story.

Burns enters on a four-fight losing streak, most recently a TKO loss to Michael Morales nearly a year ago last May. Since challenging for the belt he’s gone 3–5, with losses coming against the best of the division — no shame in that. But the data underneath those results has been trending in one direction for a while now. Burns is 39 years old, and at some point the mileage shows up in the model whether you want it to or not.

Across from him is Mike Malott, 34 years old and riding a three-fight win streak after a decision over Kevin Holland last October. Malott is ascending while Burns is declining, and the numbers reflect that gap clearly.

The xR% says it all. Malott walks in at an excellent 78%. A dominant round-winning profile. Burns has slipped to 53%, nearly at the coin-flip line, which for a fighter of his caliber is a significant red flag.

The striking numbers confirm the decline. Burns lands 3.15 significant strikes per minute with a –0.49 differential. So he’s absorbing more than he’s landing on the feet. Malott lands 3.82 per minute with a solid +1.07 differential, absorbing fewer than three significant strikes per minute in return. Clean, efficient, and effective.

The knockdown history is the most alarming data point for Burns. He has been dropped 10 times in his UFC career. Malott has not been knocked down once across seven UFC fights. Burns has also lost two of his last four by knockout, the kind of trend that accelerates as a fighter ages.

On the mat, Burns is still the more active grappler. He holds a 66% control rate and has scored 42 takedowns in his UFC career, though only at 38% accuracy. Malott counters with an 82% control rate and 57% takedown accuracy across his eight scored takedowns. Plus Burns’ submission game remains a threat — it always will be — but his 38% accuracy on takedown attempts and Malott’s overall physical profile suggests he may not be able to get this fight to the mat consistently enough to matter.

The model lands on Malott at 65.81% win probability. The problem is the market already has him at –350, implying 77%. The model doesn’t support that price, and this is a clean Stay Away designation. Tier 5 Trap Fav.

Burns’ toughness and championship-level experience could make this scrappier than the odds suggest, and that’s exactly why we’re not laying –350 on it. Malott is the right side. He’s the official prediction. But the bankroll stays on the sideline for this one.

One play on the card this week in Kyler Phillips, with a clear stay-away on the main event despite the model’s lean toward Malott. Sometimes the smartest bet is the one you don’t make.

The model sits at 45–18–6 on the 2026 season — 71.43% on the year. Light cards like UFC Winnipeg are a good reminder that protecting the bankroll is just as important as building it. Five fights modeled, one real edge, a few stay-away. That’s the discipline.

No First Strike this week, but Dave Ross and I will be back next week with a full card breakdown. Make sure you’re subscribed wherever you listen to podcasts and on YouTube so you don’t miss it.

Follow along on X for live reactions, odds movement, and any last-minute updates heading into fight night. Find me at @TheRobbeo and Dave at @drosssports.

Good luck Saturday. Enjoy the violence.

Model Prediction | Win Prob.% | Tier

  • Allan Nascimento | 68.07% | Tier 4 Fragile Favorite
  • Jamey-Lyn Horth | 63.77% | Tier 5 Trap Favorite
  • Jasmine Jasudavicius | 69.16% | Tier 3 Volatile Favorite
  • Kyler Phillips | 72.50% | Tier 1 Live Dog
  • Mike Malott | 65.81% | Tier 5 Trap Favorite

The post First Strike Podcast Model Predictions for UFC Winnipeg: Burns vs Malott appeared first on VSiN.