What we missed
Troyes 2-6 Strasbourg: We Got This One Wrong
We called Troyes at home. Strasbourg had other ideas, scoring six in a result that turned our read inside out.

We called Troyes to win. Strasbourg scored six. This is the post-mortem.
What We Said Before Kick-Off
The logic felt solid. Troyes had been composed and defensively mean in their opening two fixtures, conceding just once across both games. Strasbourg had been leaking goals at an alarming rate, failing to keep a single clean sheet in six matches and shipping 13 goals in that run. The form read pointed one way and we followed it.
Sheila
Troyes are the favourite, they're at home, and their defence has been mean early in the season. This is the bread-and-butter call.
We locked home or draw as our safety net, covering 84% of simulated outcomes. We went 2-0 on the exact score. We even called Strasbourg to not score, trusting that Troyes' backline would hold firm. Four of our ten markets pointed squarely at a controlled Troyes win.
Where the Match Broke
For about nine minutes, the call looked alive. Antoine Mille put Troyes ahead early, assisted by Anis Ouzenadji, and the low-block identity we had backed was doing its job.
Then the game flipped. Jacobo Ortega pulled one back in the 40th minute. Diogo Sousa made it 2-1 before half time. Strasbourg went into the break ahead, and the half-time result market was already gone.
The second half was not a contest. Giovanni Reyna extended the lead in the 57th minute, and Samuel Amo-Ameyaw added two more, his second in the 88th minute. Conrad Harder wrapped up a 6-1 scoreline in injury time before Jérémy Le Douaron grabbed a consolation for Troyes in the 95th to make it 2-6. Eight goals, six to the away side, in a game we had modelled as a tight home win.
Tom
Strasbourg do score though, 1.8 per game. There's a real chance they nick one.
Tom flagged the tension before kick-off and we still went narrow no on Strasbourg to score. That was wrong, and not unlucky wrong. Strasbourg had scored in five of their last six matches. We leaned too hard on Troyes' early-season defensive numbers and not hard enough on what Strasbourg's attack had been doing consistently.
What the Numbers Missed and What We Take Forward
We got four markets right: Troyes to score, over 1.5 goals, a goal in each half. Those held up because the underlying logic, goals were coming, was sound. Everything built around Troyes controlling the game fell apart once Strasbourg found their rhythm in the final minutes of the first half and never stopped.
The goals band of 1-4 sitting at 81.5% probability looked comfortable. Eight goals landed. The simulation did flag blowout risk as real, and we noted it, but we did not weight it enough against a Strasbourg side that had been involved in high-scoring matches all season.
This was not a fluke result in the way a single deflected goal or a contentious red card can derail a call. Strasbourg were simply better on the day, more clinical, and Troyes' defensive shape collapsed over ninety-five minutes in a way their first two fixtures had not suggested was possible.
The honest read going forward: Strasbourg's attacking output is real and consistent. Any side facing them needs to show more than two games of solid defending before we back them to keep the visitors quiet. We backed the early-season sample too heavily. That is the lesson we carry into the next one.
See every call Tom and Sheila have made, win or miss.
See the record