The number that undercuts the lean
Sevilla host Osasuna at the Ramón Sánchez-Pizjuán on Monday night with the home side sixth in La Liga on 14 points and the visitors eleventh on 11, and Nutmegly's model has duly come down on the home side. The trouble is how softly it has done so. The AI read splits the result three ways at 42.1% Sevilla, 39.5% draw and 18.4% Osasuna. A home lean of roughly two and a half percentage points over the draw is not a conviction, it is a shrug with a favourite attached, and any preview that treats this as a straightforward home assignment is reading the table and ignoring the model.
There is a second number in that forecast that is harder to explain away. The model puts over 2.5 goals at just 15.76%, which is a firm expectation of a tight, low-scoring night. Now look at what these two teams have actually been serving up. Sevilla's three most recent recorded results read 1-4, 3-3 and 1-3. Osasuna's recent record includes a 2-5 and a 0-4. Neither side has been living in the kind of match the goal forecast describes. That gap between the forecast and the recent evidence is the most interesting thing about this fixture, and it is where the argument should start.
What the table actually says
The case for Sevilla rests less on brilliance than on arithmetic. Nine games, four wins, two draws, three defeats, 14 scored and 16 conceded — a mildly negative goal difference attached to a top-half position. Osasuna's nine games read three wins, two draws, four defeats, 10 scored and 18 conceded. The eight-point gap in goals conceded is the real separator between sixth and eleventh.
The away split sharpens it further. Osasuna have played four times on the road in the current league table and lost three of them, scoring six and conceding 15. Their two most recent recorded away trips were a 2-5 defeat at Real Betis and a 0-4 defeat at Atlético Madrid. If you are looking for the single strongest piece of evidence behind a home lean, that is it: an away defence shipping goals at a rate that would make any host feel entitled to the points.
The counterweight is timing. Osasuna's most recent recorded result is a 2-0 home win over Racing Santander on 18 October, their only clean-sheet-looking scoreline in the recent sample. A side that has just won without conceding is not the same proposition as the one that shipped nine goals across two away days in September, and the model's read does not get to lean on that win because of when it was produced.
Sevilla's home record is not a fortress
Here is where the home lean gets shakier. Sevilla's home split in the current table is four games, two wins, no draws and two defeats, with five goals scored and seven conceded. That is a negative home goal difference. Their recent home results in the stored record are a 1-3 defeat to Barcelona and a 1-0 win over Valencia. The Sánchez-Pizjuán has not been a place where Sevilla have dominated, and the model's 42.1% is being carried more by Osasuna's away numbers than by anything Sevilla have done in front of their own crowd.
That matters for how the match might unfold. If Sevilla's home defence — seven conceded in four — is the version that turns up, Osasuna's away attack, six goals in four away games, has enough about it to make the evening uncomfortable. If instead the Sevilla side that beat Valencia 1-0 and won 1-0 at Deportivo La Coruña shows up, the low-goal forecast starts to look reasonable. The data supports both readings, which is precisely why the model landed near a coin flip.
The forecast is older than the evidence
One caveat deserves to be stated plainly rather than buried. The model's read was generated on 7 October, before both clubs played their two most recent recorded matches. It was also produced without any odds summary, without an external forecast for cross-checking, without confirmed lineups and without injury reports. The result probabilities have not been calibrated against a market, and the model's own counterpoint notes that starting selections are not settled and could still change the shape of the game. Treat the percentages as an uncalibrated estimate, not a verdict.
The stored results also have gaps — Sevilla's record jumps from 19 September to 12 October, Osasuna's from 19 September to 11 October — so the recent sample should be read as a set of available results rather than a continuous run of form. Four-game home and away splits are small samples, and small samples of goals are noisy by nature.
The strongest case against the lean
The best argument against the home read is not Osasuna's quality, it is the draw. At 39.5% the draw sits within touching distance of the home win, and the model is effectively saying the two most likely outcomes are almost equally likely. That is a fair reflection of a fixture where the host has a negative home goal difference and the visitor has just won 2-0.
But there is a counterargument to the counterargument, and it is a good one. Neither split has produced a single draw: Sevilla's four home games are two wins and two defeats, Osasuna's four away games are one win and three defeats. Across eight matches in these specific conditions, the draw has not appeared once. That does not disprove a 39.5% estimate — eight games is a tiny sample and draws are the outcome most sensitive to a single goal — but it does mean the model's second-favourite result is one that neither team's home or away record has yet delivered.
The question to watch
Watch the first goal, and specifically who gets it. If Sevilla score first, the 42.1% starts to look like the sensible side of the ledger and Osasuna's away record — 15 conceded in four — becomes the story of the night. If Osasuna score first, the model's own numbers say the draw is the live outcome, and the 15.76% over-2.5 forecast gets tested immediately by two teams whose recent matches have been anything but quiet. The lean says Sevilla; the evidence says the draw is close enough that one goal could decide which number was right.