Europa League predicted table 2026/27
Europe · 2026/27
Not the current table: our estimate of the final league-phase standings, from 10,000 simulations. Updated UK time (GMT).
Predicted final table
| # | Team | Now | Pld | Points now | Exp. pts | Last 16 (top 8) | Knockout play-offs | Eliminated |
|---|---|---|---|---|---|---|---|---|
| 1 | Bayer Leverkusen | 2 ▲1 | 3 | 9 | 18 | 84% | 16% | <1% |
| 2 | Benfica Leader now | 1 ▼1 | 3 | 9 | 18 | 82% | 17% | <1% |
| 3 | Bournemouth | 3 | 3 | 9 | 17 | 73% | 27% | <1% |
| 4 | Olympiakos Piraeus | 5 ▲1 | 3 | 9 | 17 | 67% | 33% | <1% |
| 5 | Ferencvarosi TC | 4 ▼1 | 3 | 9 | 17 | 70% | 30% | <1% |
| 6 | Juventus | 7 ▲1 | 3 | 7 | 16 | 62% | 36% | 1% |
| 7 | Union St. Gilloise | 6 ▼1 | 3 | 7 | 14 | 45% | 51% | 3% |
| 8 | Lyon | 12 ▲4 | 3 | 6 | 14 | 40% | 55% | 5% |
| 9 | Beşiktaş | 8 ▼1 | 3 | 7 | 14 | 41% | 55% | 4% |
| 10 | Sunderland | 9 ▼1 | 3 | 7 | 14 | 34% | 60% | 6% |
| 11 | Crystal Palace | 15 ▲4 | 3 | 4 | 14 | 37% | 56% | 7% |
| 12 | AZ Alkmaar | 10 ▼2 | 3 | 6 | 13 | 30% | 62% | 8% |
| 13 | Rennes | 13 | 3 | 5 | 12 | 19% | 66% | 14% |
| 14 | Omonia Nicosia | 11 ▼3 | 3 | 6 | 12 | 13% | 68% | 19% |
| 15 | Real Sociedad | 14 ▼1 | 3 | 4 | 11 | 16% | 64% | 20% |
| 16 | AC Milan | 24 ▲8 | 3 | 3 | 11 | 10% | 65% | 24% |
| 17 | Celta Vigo | 25 ▲8 | 3 | 3 | 11 | 10% | 63% | 26% |
| 18 | NEC Nijmegen | 20 ▲2 | 3 | 4 | 11 | 8% | 61% | 31% |
| 19 | Levski Sofia | 17 ▼2 | 3 | 4 | 11 | 9% | 63% | 28% |
| 20 | Anderlecht | 22 ▲2 | 3 | 3 | 11 | 10% | 62% | 28% |
| 21 | Sturm Graz | 19 ▼2 | 3 | 4 | 10 | 5% | 56% | 38% |
| 22 | Dinamo Zagreb | 18 ▼4 | 3 | 4 | 10 | 7% | 56% | 38% |
| 23 | OFI | 26 ▲3 | 3 | 3 | 10 | 5% | 58% | 36% |
| 24 | Sparta Praha | 16 ▼8 | 3 | 4 | 10 | 6% | 57% | 37% |
| 25 | Torreense | 28 ▲3 | 3 | 3 | 9 | 4% | 52% | 44% |
| 26 | 1899 Hoffenheim | 23 ▼3 | 3 | 3 | 9 | 4% | 49% | 47% |
| 27 | Red Bull Salzburg | 21 ▼6 | 3 | 4 | 9 | 2% | 43% | 55% |
| 28 | Jagiellonia | 27 ▼1 | 3 | 3 | 8 | 1% | 36% | 62% |
| 29 | Plzen | 33 ▲4 | 3 | 0 | 7 | 1% | 31% | 68% |
| 30 | Celje | 29 ▼1 | 3 | 1 | 7 | 1% | 28% | 72% |
| 31 | Celtic | 35 ▲4 | 3 | 0 | 6 | 1% | 23% | 76% |
| 32 | Marseille | 34 ▲2 | 3 | 0 | 6 | <1% | 24% | 76% |
| 33 | Hapoel Beer Sheva | 30 ▼3 | 3 | 1 | 6 | <1% | 19% | 80% |
| 34 | Lech Poznan | 32 ▼2 | 3 | 1 | 6 | <1% | 18% | 81% |
| 35 | Ararat-Armenia | 36 ▲1 | 3 | 0 | 6 | <1% | 21% | 79% |
| 36 | Lillestrom | 31 ▼5 | 3 | 1 | 6 | <1% | 17% | 83% |
Sorted by expected points. # = predicted final place; Now = place today, with the expected move (▲ climb, ▼ drop). Zone columns: % of simulations. Zones: Last 16 (top 8) = 1–8; Knockout play-offs = 9–24; Eliminated = 25–36. Bold = more likely than not. 0% and 100% appear only when the outcome is already mathematically settled.
How sure is this?
It is our estimate of how the 36-team league phase ends, based on 10,000 simulations: we play out the remaining 90 matches 10,000 times using team ratings from European matches this season and the last two (each competition rated like a division, with the step between competitions learned from clubs that moved), and show how often each team finished in each zone: 1–8 straight to the last 16, 9–24 knockout play-offs, 25–36 out. Updated UK time (GMT).
After 3 matchdays, a team's chances can still swing a lot. Early in the season the ratings lean more on last season.
How this table is made
Each team gets an attack and a defence rating. The starting point is its rating from the last two seasons (last season counts twice as much as the one before), moved up or down a division with a factor learnt from promoted and relegated clubs; clubs new to our data start like the weaker clubs of the division. This season's results are then added on top: the history counts like 20 matches, so after about half a season this season's form outweighs it. Ratings are estimates, so in every simulation each team's strength is also drawn with some uncertainty, more early in the season, less later; that keeps the chances from becoming overconfident. We then play every remaining match 10,000 times with goals drawn from those ratings, and count how often each team ends up in each zone.
Table, played and remaining matches all come from one snapshot of the same match list (54 played + 90 remaining = 144); a run where these numbers do not add up is not published. Current points include any points deductions; ties are broken on goal difference, then goals scored.
Tested on the last two finished seasons from the 6th and 11th rounds: the two-season history made expected final points 5–10% closer to the real ones, and the rating uncertainty made the title, promotion and relegation chances more accurate (log loss 0.206 → 0.199). It uses no bookmaker prices, news or AI. Methodology