Championship Round 3: the model found four tidy wins, but the scorers had other ideas

The round was mostly right about who would win, and mostly wrong about how they would do it.

Four exact scorelines landed, which is enough to keep a model honest without letting it get smug. The problem was that the misses were not random little slips. They clustered around games where Fipster liked the home side, liked a narrow margin, and then watched the match run away from that shape entirely. Southampton did not just beat Millwall; they turned it into a rout. Norwich did not just edge Burnley; they opened them up. Middlesbrough and Bristol City were closer to the script, but even there the goals arrived from places the pre-match calls had not properly mapped out. That is usually where a round starts telling you something useful rather than merely noisy things in a louder voice than necessary.

The cleanest hit was Middlesbrough’s win over West Brom.

This was one of the few places where the model’s basic read held together for more than just the result line. Fipster backed a home win and got that part right, but more importantly it had identified Middlesbrough as the side more likely to control proceedings. The actual game followed that logic once William Lankshear got going after half-time. His two goals changed a tight contest into something more comfortable, and Alfie Jones added the late cushion. The key point is not that Fipster named every scorer — it did not — but that its underlying read on territory and outcome survived contact with the afternoon. That matters more than pretending every correct scoreline is equally instructive.

The miss that stung most was Southampton against Millwall.

This was where the model over-weighted caution and under-read ceiling. It had Southampton down for a routine home win by a single goal; instead they scored five, and Millwall’s early lead only made things look briefly respectable before everything tilted hard in one direction. The real damage came from players Fipster had not called at all: Cyle Larin scored twice, Finn Azaz ran the game with a goal and an assist, Léo Scienza added another goal and an assist, and Tom Fellows finished it off late. Mark Sykes had already put Millwall ahead, which only sharpened how badly Southampton then took over. The model saw Southampton as winners. It did not see them as this sort of winner.

The round belonged to players Fipster never properly named up front.

That is where this week gets its edge. The obvious headline names were not always the decisive ones; often they were simply absent from Fipster’s pre-match list altogether. William Lankshear scored twice for Middlesbrough without being called. Mohamed Touré and Andre Chance Brooks each scored twice for Norwich without being called. Rodrigo Gomes did the same for Wolves in a match Fipster had tipped Stoke to nick away from home. Phil Neumann’s brace for Birmingham also sat outside the model’s player calls, as did Karlan Ahearne-Grant’s winner for Charlton and Benjamin House’s strike for Lincoln. Even when Fipster got the result right, as with Southampton or Norwich or Middlesbrough, it still left room for someone else to walk through the door and take over.

Callum Doyle at Wrexham was another reminder that one good individual can break a neat forecast without needing much help from anyone else. Birmingham’s 2-1 win there came through Phil Neumann’s two goals, while Callum Doyle at least gave Wrexham something back after half-time. The broader lesson is plain enough: Fipster has started this Championship season with some sound directional reads, but it still needs sharper instincts on which attackers are actually about to matter most. That is calibration work rather than crisis work, which is fortunate because crisis work tends to be tedious and expensive.