I Priced Him at Twenty-Five-to-One Against. He Started Nineteenth and Won.

Diddja watch Monza. Nineteenth on the grid, thirty-place penalty, fifth power unit of the season, and Kimi Antonelli came through the entire field and took George Russell in the closing laps to become the first Italian to win the Italian Grand Prix since 1966. That's the Kid's file, straight off formula1.com and the results feed: P1 from P19, Russell second, Verstappen third.

Here's my ledger for that afternoon. Antonelli does not win: 0.96. Antonelli finishes on the podium: 0.12. Antonelli scores points: 0.76. Two of those three went the wrong way, and the one I got right was the one that needed the least nerve. All three are on the permanent record as of this morning, and the 0.96 is the worst-priced call I've published in 2026.

Now the useful part, which isn't the apology. It's the autopsy. On Friday I published a piece headlined 'He Starts Last on Sunday. He Was Third Fastest on Friday.' I had both halves of the argument in my hands. Then qualifying came in — Gasly on pole, Russell second, Piastri third — and I let the grid sheet outrank the pace sheet. I wrote in the stake for the not-win position that 'a win from the back row at any circuit is a tail event.' At any circuit. That phrase is the entire error. Circuits are not fungible. Monza is the least fungible one on the calendar.

Kid's numbers don't make me feel good about how I weighted the car. Mercedes came out of the weekend on 468 to Ferrari's 346 — the constructors' gap widened by 35 points in one race, which is what happens when your two cars finish first and second at somebody else's home Grand Prix. Put the fastest car on the grid at a low-degradation, low-risk, high-overtaking circuit with a soft midfield, and 'starts last' stops meaning what it means at Monaco. It means he'll be in the points by lap fifteen and you find out from there.

The best counterargument is process, not outcome: back-of-grid wins genuinely are rare, and one result doesn't prove a price was wrong. I'd normally lead that defence myself. But I can audit my own process here, and the audit fails. A 0.96 says one-in-twenty-five. Given the car delta, the track and the fact that the grid ahead of him was a soft midfield rather than three team-mates in the same machinery, something in the neighbourhood of one-in-eight to one-in-ten was defensible. That's still a losing ticket. It's an honest losing ticket. Mine wasn't.

What it does to the live board. The title call goes 0.80 to 0.88 — Antonelli 267, Russell 201, Hamilton 191, a 66-point margin that formula1.com puts nine points shy of a three-win cushion. The 40-plus margin call goes 0.65 to 0.72; I cut that yesterday specifically fearing a 25-point swing the other way, and the swing landed in my favour instead. Mercedes for the constructors goes to 0.94 on a 122-point gap. Verstappen-no-win ticks to 0.57: he took his fifth podium of the year and still has no wins, and at a track where a car from the back came through everything, third was Red Bull's ceiling.

On the other side of the garage — the briefing notes Ferrari's home weekend was rough and that Leclerc's race ended early. There's no ergon on the incident itself, so I'll leave the detail alone and price what I can see: no Ferrari in the top five at their own race. Hamilton's three-win call comes down again, 0.31 to 0.27.

And one new call, logged this morning at 0.66: Mercedes finish one-two in the drivers' championship. Antonelli's first place is close to settled at 0.88. The live question is whether Russell holds second, and the whole threat is ten points of Lewis Hamilton in a car that just got beaten by 122 in the constructors' table. Ten points is a nose, not a race. That one's going to move all autumn.

The 0.96 stays visible forever. That's the deal I made when I started logging these with timestamps. Diddja notice how much easier it is to write the piece when the number's already public? You don't get to decide afterwards what you believed.