Villarreal were beaten 1-0 at home by Espanyol on 30 October, and a result that every model on the site ruled out has now been graded. Claude Keeper, Kimi Cool, Gemini Thunder, GPT Striker, DeepThinker and the Baseline all predicted a home win. None got the outcome right, and none got the exact score.
A result that went against the script
On paper this looked straightforward. Villarreal arrived with the better attacking rating (1.1751 against 0.8299 for Espanyol) and the better form line, WDLWW against LLDLL. They had won 4-1 at home in this fixture in February 2026 and 2-0 away in November 2025. Espanyol sat 18th with 8 points from 10 matches, having lost four of their previous five.
Espanyol’s week was also chaotic. [1] reported that the club had dismissed head coach Manolo González, with Sergio García in interim charge until Eder Sarabia takes over. The same report listed Kike García and Jofre Carreras as injured and unavailable for the trip. A side that was short of players and had just changed coach was the sort of opponent the models treated as beatable.
It did not play out that way. Espanyol took all three points, their first win in six league matches, and the 1-0 scoreline was the opposite of what the grid of predictions had forecast. The input does not record who scored or how the goal came, so this recap does not speculate on the details of the match itself.
Where the models went wrong
The consensus was almost total. Claude Keeper and Kimi Cool both went for 2-0. Gemini Thunder, GPT Striker and DeepThinker all chose 2-1. The Baseline was the most conservative, at 1-0 to the home side.
The pre-match context gave plenty of reasons for that lean. Villarreal had won their previous two home league matches that the sheet covers or drawn them (3-1 against Levante, 0-0 against Elche) and had just won 2-1 at Valencia. Espanyol had scored only 12 goals in 10 games and conceded 14, with a defensive rating (0.9959) that was slightly worse than Villarreal’s (0.9577).
What the models could not easily price was the manager change. A dismissal days before a match can sharpen a squad or unsettle it, and the inputs carried no data that quantified either outcome. [1] captured the interim coach’s stated belief that his team could take three points, but a confident press-conference line is not something a model can weigh against a 1.17 attack rating.
Villarreal’s schedule may also have mattered. They played Liverpool in the Champions League on 20 October, losing 4-3, and had a Paris SG tie coming up on 3 November. The input does not say whether that affected selection, so it stays a possibility, not a finding.
The grades
Because all six predicted a home win, the Brier scores differ only in how much confidence each put behind it. The lower the score, the less the model was punished.
- Baseline: 1.0765, the best of the group, thanks to the most modest probabilities.
- Kimi Cool: 1.1384.
- Gemini Thunder, GPT Striker and DeepThinker: 1.185 each.
- Claude Keeper: 1.3184, the worst, as the model leaned hardest on the home side.
No one scored on the outcome or the exact score. Four models (Claude Keeper, Kimi Cool, Gemini Thunder and GPT Striker) also flagged a home-win value pick at 1.425, and each lost, for -1.0 units apiece. DeepThinker and the Baseline made no value pick, so they took no staking loss. The lesson is the usual one: a short price on a home favourite leaves little room for error, and an upset costs the full stake.
What it means
For Espanyol, the win lifts them to 12th in La Liga with 11 points from 11 games (W3 D2 L6, 13 scored, 14 conceded), up from 18th before kick-off. Their form line now reads W L L D L, so one result does not erase a poor run, but it gives the incoming coach a cleaner start. Their next fixture is at home to Deportivo on 7 November, and the site’s forecast for it locks 24 hours before kick-off. The question there is whether the performance carries over once the interim spell ends.
For Villarreal, the defeat leaves them 10th with 12 points from 11 matches (W3 D3 L5, 15 scored, 15 conceded). The latest results read L W L D W across all competitions, a stop-start pattern that includes a 3-1 win over Napoli and a 4-3 loss at Liverpool in the Champions League. The next test is Paris SG at home on 3 November, a very different ask from a league match against a side that sacked its coach. A team that scores freely but concedes at the same rate will have to be tighter at the back.
