Bayern v 
Köln
Fri, Nov 20 · 20:30 CET
Bayern by 2.5 goals on average, with home advantage.
- Likeliest scores
- 2–0 13%3–0 13%4–0 10%
- Goals expected
- Bayern 3.0 Köln 0.5
6+54321Draw123456+ ← Bayern win by Köln win by →
Bundesliga 2026-27 · Matchday 10 · Fri, Nov 20 – Sun, Nov 22
Bundesliga predictions, matchday 10
Every Bundesliga match of the matchday, forecast from the goals each club has scored and let in this season, rated against the opponents it has played, and from where it finished last season. Win, draw or loss, as chances. No betting information goes in. Tested on 5,330 league matches of 2021-22 to 2025-26, each forecast from the matches before it, the favourite won 70.4% of the matches that had a winner.
Built on the results our source has through 8 Nov; it is updated by volunteers, so the latest matchdays can arrive late and the ratings catch up when they do.
Updated 10 Nov, 02:56 UTC. Each forecast is frozen at kickoff and scored when the match is played. How it has done
| Match | Kickoff | Home | Draw | Away |
|---|---|---|---|---|
| Bayern v Köln | Fri, Nov 20 | 86% | 11% | 4% |
| Leverkusen v Paderborn | Sat, Nov 21 | 67% | 21% | 12% |
| Elversberg v Freiburg | Sat, Nov 21 | 33% | 26% | 41% |
| Gladbach v Dortmund | Sat, Nov 21 | 15% | 23% | 62% |
| Frankfurt v Hoffenheim | Sat, Nov 21 | 44% | 24% | 32% |
| Union Berlin v Leipzig | Sat, Nov 21 | 20% | 24% | 56% |
| Schalke v Stuttgart | Sat, Nov 21 | 22% | 26% | 52% |
| Augsburg v Mainz | Sun, Nov 22 | 44% | 26% | 30% |
| Werder v HSV | Sun, Nov 22 | 52% | 26% | 22% |
Leverkusen v 
Paderborn
Sat, Nov 21 · 15:30 CET
Leverkusen by 1.3 goals on average, with home advantage.
- Likeliest scores
- 2–0 12%1–0 12%1–1 10%
- Goals expected
- Leverkusen 2.1 Paderborn 0.8
6+54321Draw123456+ ← Leverkusen win by Paderborn win by →
Elversberg v 
Freiburg
Sat, Nov 21 · 16:00 CET
Freiburg by 0.2 goals on average, even with Elversberg at home.
- Likeliest scores
- 1–1 12%1–2 9%0–1 8%
- Goals expected
- Elversberg 1.4 Freiburg 1.6
6+54321Draw123456+ ← Elversberg win by Freiburg win by →
Gladbach v 
Dortmund
Sat, Nov 21 · 16:00 CET
Dortmund by 1.1 goals on average, even with Gladbach at home.
- Likeliest scores
- 0–1 12%0–2 11%1–1 11%
- Goals expected
- Gladbach 0.8 Dortmund 1.9
6+54321Draw123456+ ← Gladbach win by Dortmund win by →
Frankfurt v 
Hoffenheim
Sat, Nov 21 · 16:00 CET
Frankfurt by 0.3 goals on average, with home advantage.
- Likeliest scores
- 1–1 10%2–1 9%2–2 7%
- Goals expected
- Frankfurt 1.8 Hoffenheim 1.5
6+54321Draw123456+ ← Frankfurt win by Hoffenheim win by →
Union Berlin v 
Leipzig
Sat, Nov 21 · 16:00 CET
Leipzig by 0.8 goals on average, even with Union Berlin at home.
- Likeliest scores
- 1–1 11%0–1 10%1–2 10%
- Goals expected
- Union Berlin 1.0 Leipzig 1.9
6+54321Draw123456+ ← Union Berlin win by Leipzig win by →
Schalke v 
Stuttgart
Sat, Nov 21 · 18:30 CET
Stuttgart by 0.7 goals on average, even with Schalke at home.
- Likeliest scores
- 1–1 12%0–1 11%0–2 10%
- Goals expected
- Schalke 1.0 Stuttgart 1.7
6+54321Draw123456+ ← Schalke win by Stuttgart win by →
Augsburg v 
Mainz
Sun, Nov 22 · 15:30 CET
Augsburg by 0.3 goals on average, with home advantage.
- Likeliest scores
- 1–1 12%2–1 9%1–0 8%
- Goals expected
- Augsburg 1.7 Mainz 1.3
6+54321Draw123456+ ← Augsburg win by Mainz win by →
Werder v 
HSV
Sun, Nov 22 · 17:30 CET
Werder by 0.7 goals on average, with home advantage.
- Likeliest scores
- 1–1 12%1–0 11%2–1 9%
- Goals expected
- Werder 1.7 HSV 1.0
6+54321Draw123456+ ← Werder win by HSV win by →
How good is it?
This season so far: 81 matches forecast, the favourite won 68.8% of those with a winner (ranked probability score 0.204, against 0.227 for the league’s usual shares of home wins, draws and away wins). The 54 matches played before the forecasts went live were forecast afterwards from what was known before kickoff.
Before it went live, the method was run through 10,660 league matches of ten seasons (the Premier League, La Liga and the Bundesliga, 2016-17 to 2025-26), each forecast only from the matches played before its matchday. Its two settings were picked on the first five seasons, and the table is the 5,330 matches of the five after, which it was never fitted to (2021-22 to 2025-26):
| Forecast | Winner picked right | Ranked probability score | Log loss |
|---|---|---|---|
| Ours: goals | 70.4% | 0.200 | 0.983 |
| The table (points) only | 70.6% | 0.201 | 0.984 |
| Only the league’s usual shares of home wins, draws and away wins | 59.6% | 0.229 | 1.068 |
The ranked probability score and log loss grade the chances, not just the pick: lower is better. A draw is the hardest result to call; about one match in four ends level whatever the clubs, so “winner picked right” counts only matches that had a winner.
What goes in, and what doesn’t
- This season. Each club’s goal difference (a match counts at most 3 goals) and how many goals its matches have, rated against the opponents it has played.
- Last season. Each club starts from where it finished, with 90% of its distance from average kept (a club new to the league starts like last season’s weakest). The matches played pull it away from there. That share and the strength of the rating were picked on the first five seasons of the test.
- Home advantage. Worth about a third of a goal.
Not in it: line-ups, injuries and suspensions, transfers, shots and chances, and the betting line.