A narrow lean at Balaídos
Celta Vigo host Alaves on Sunday at Municipal de Balaídos in La Liga's ninth round, and the table offers the first tension. Alaves arrive seventh on 11 points from eight games; Celta sit ninth on eight. Nutmegly's AI read, generated on 20 September, leans away: Alaves 38.6%, Celta 33.8%, draw 27.6%.
That is a lean, not a verdict. The gap between the leading outcome and the home win is under five percentage points, and the draw band is nearly as large as the home win. The model also puts over 2.5 goals at 53.99%, a slight nudge toward goals rather than a forecast of a shootout.
What the recorded samples show
Celta's stored results, newest first, read: 0-0 at Elche on 11 October, a 5-0 home win over Racing Santander on 19 September, 1-1 at home to Malaga, 1-1 at Getafe, and 0-0 at Real Sociedad. Unbeaten across those five, but four draws.
Alaves' sample runs: 1-2 at home to Atletico Madrid on 10 October, 0-0 at Athletic Club, 0-1 at home to Valencia, 1-2 at Racing Santander, and a 5-2 home win over Osasuna. Three defeats in five. Across eight recorded matches each, Celta average 1.25 points per game and Alaves 1.38; within the La Liga sample, that becomes 1.20 against 1.40. The latest entries are from 10-11 October, and the stored record does not establish current-season form beyond what the table shows.
Why the model looks away
The away lean has a real basis. Alaves are higher in the table, have a better points-per-game return in the stored sample, and a better same-competition efficiency. Celta's home split is not a fortress: one win, one draw, two defeats in four league home games, with seven scored and five conceded. In the broader stored same-venue sample, Alaves concede 0.80 goals per away game, while Celta score 1.60 per home game — a matchup that could keep the game tight rather than open. If Alaves reproduce the defensive control suggested by that away sample, the 38.6% away win is the most plausible single outcome.
The counterargument: Alaves haven't won away
The strongest supplied counterevidence is Alaves' current league away record: three games, no wins, two draws, one defeat, two goals scored and three conceded. That is a winless away profile, and it sits awkwardly against a model that makes them favourites. Celta, meanwhile, are unbeaten in their five most recent recorded matches, even if four were draws.
The model's own counterpoints flag that the three-way probabilities are not clearly separated and that the draw interval must be retained. With no odds, no external forecast, no confirmed lineups and no injury reports in the data, the read is uncalibrated and lineup-sensitive. The head-to-head sample is only two matches, too thin to settle anything. And the prediction was generated on 20 September, so it did not see the 10-11 October results.
The watchpoint
The match turns on which sample proves more relevant. If Celta's home scoring — seven goals in four league home games — carries into this one, the away lean weakens and the draw band becomes the more honest read. If Alaves' away defence holds and they take the lead, the model's 38.6% looks justified. Watch the first goal and the game state that follows: Celta's home record includes two defeats, so an early Alaves goal would test their response; Alaves' away record includes no wins, so an early Celta goal would test whether they can chase. With over 2.5 at 53.99%, goals are slightly more likely than not, but not by much. The clearest conditional: if the game reaches half-time level, the 27.6% draw is the outcome the evidence best supports.