First against fifteenth, with a draw-column caveat
Bayern München enter Friday's Round 10 fixture at the Allianz Arena top of the Bundesliga: nine games, six wins, three draws, no defeats, 21 points, 29 goals scored and nine conceded. 1. FC Köln arrive 15th with nine points from nine games, two wins, three draws and four defeats, 10 goals for and 14 against. On the plain table, this is first against fifteenth, and Bayern's home split makes the gap look wider still: three wins and a draw in four home league games, with 20 goals scored and four conceded.
The tension is not whether Bayern are favored; the data makes them overwhelming. It is that Bayern's unbeaten league record contains three draws, and their recent sample also includes home draws with Borussia Dortmund and Arsenal, both 2-2. Those results do not prove a template for Köln, but they do show that Bayern's season has not been a procession of wins. A team that has dropped points three times in nine league games can be held again, even if the model thinks it unlikely.
Köln's away split is the weak link
Köln's overall record is modest, but their away split is the number that should worry them most. In four away league games they have no wins, one draw and three defeats, scoring three and conceding eight. That is the mirror image of Bayern's home split: 20 goals for and four against in four games at the Allianz Arena. The matchup of those two samples is the core of the case for a home win.
Recent results point the same way, with a caveat. Bayern's recent run reads 2-1 at FSV Mainz 05, 2-0 at Atletico Madrid, 2-2 at home to Borussia Dortmund, 3-1 at SC Freiburg and 2-2 at home to Arsenal: three wins and two draws. Köln's reads 1-2 at home to Bayer Leverkusen, 1-1 at Union Berlin, 0-0 at home to FC Schalke 04, 0-1 at Eintracht Frankfurt and 2-1 at home to Borussia Mönchengladbach: one win, two draws and two defeats. Those are recent samples, not complete season records; they are useful for shape, not for declaring a definitive trend.
The AI read and the case for it
Nutmegly's AI read makes Bayern a heavy home favorite: 87.7% for a home win, 7.8% for a draw and 4.5% for a Köln win. The same model puts over 2.5 goals at 76.7%. Those are forecasts, not results, and a leading outcome is not a guarantee. But the direction is consistent with the table: Bayern lead the league with 21 points and a plus-20 goal difference, while Köln have the away record of a team that has struggled to score on the road.
The case for the model is straightforward. Bayern have scored 29 goals in nine league games and 20 in four at home; Köln have scored three in four away and conceded eight. Köln's recent 0-0 and 1-1 show they can be involved in low-scoring games, but their away sample has produced only one point from four. If the game follows the broad shape of those records, Bayern should control the scoreboard.
The counterargument and the watchpoint
The strongest counterevidence is not Köln's attack; it is Bayern's draw habit and the uncertainty around selection. Bayern have drawn three of nine league games, and their recent sample includes two home draws. Köln have drawn three of nine overall and one of four away, with recent 0-0 and 1-1 results. If Köln can keep the score low, the model's 7.8% draw and 4.5% away-win shares are the part of the distribution that matters. That is a conditional route, not a prediction that it will happen.
Selection is the other qualifier. Starting lineups are not confirmed, and the model's own counterpoint notes that late tactical and personnel choices could change the result picture. No injury absences were supplied, but that is not proof of a fully healthy squad; it simply means no absences were reported in the available data. With no confirmed lineups, any preview has to leave room for a late shift.
The watchpoint is the scoreline's elasticity. Bayern's home sample says goals; Köln's away sample says a struggle to score. If Köln concede early, the model's 76.7% over-2.5 probability becomes the more relevant number, and the game could turn into a damage-limitation exercise. If Köln are still within one goal deep into the second half, the draw and away-win shares become live, not because the model expects it, but because the data's low-scoring route for Köln would be materializing.
Either way, the match is a useful test of how much weight to put on a table gap versus the draw column. Bayern's unbeaten start is real; so are the three draws. Köln's away record is poor; so is their recent scoring output. The model has made its call. The game now has to answer whether Köln can turn a mismatch on paper into a long, awkward evening.