La Liga · Spain
full timeFriday, 18 September 2026
Brought forward to Friday 18 September at 20:00. Originally planned for Sunday 20 September at 16:00.
Change noted on 03/09.
What Janus-12 said
Espanyol or draw
76% success over 58 matches in this league
✗ Fell into the 23 %
A prediction at 77 % fails 23 % of the time: it happened.
Bet offered: Espanyol or draw
Promo code: janus-12
Partner link: Janus-12 is paid for accounts opened with this link or code.
We don't show odds: check them at the bookmaker. Availability checked on 18/09, 15:31.
Other predictions, most reliable first
| # | Team | P | W | D | L | Diff | Pts | Form |
|---|---|---|---|---|---|---|---|---|
| 1 | FC Barcelona | 7 | 7 | 0 | 0 | +24 | 21 | |
| 2 | Atlético de Madrid | 7 | 5 | 1 | 1 | +9 | 16 | |
| 3 | Real Betis | 7 | 5 | 1 | 1 | +2 | 16 | |
| 4 | Real Madrid | 7 | 5 | 0 | 2 | +10 | 15 | |
| 5 | Sevilla | 7 | 4 | 1 | 2 | +1 | 13 | |
| 6 | Deportivo Alavés | 7 | 3 | 2 | 2 | +5 | 11 | |
| 7 | Deportivo La Coruña | 7 | 2 | 4 | 1 | +2 | 10 | |
| 8 | Real Sociedad | 7 | 3 | 1 | 3 | -4 | 10 | |
| 9 | Villarreal | 7 | 2 | 2 | 3 | +1 | 8 | |
| 10 | Athletic Club | 6 | 2 | 2 | 2 | +1 | 8 | |
| 11 | Getafe | 7 | 2 | 2 | 3 | -3 | 8 | |
| 12 | Rayo Vallecano | 7 | 2 | 2 | 3 | -5 | 8 | |
| 13 | Osasuna | 7 | 2 | 2 | 3 | -7 | 8 | |
| 14 | Celta de Vigo | 7 | 1 | 4 | 2 | +2 | 7 | |
| 15 | Espanyol | 7 | 2 | 1 | 4 | 0 | 7 | |
| 16 | Santander | 7 | 2 | 1 | 4 | -10 | 7 | |
| 17 | Levante | 6 | 1 | 2 | 3 | -4 | 5 | |
| 18 | Elche | 7 | 1 | 2 | 4 | -6 | 5 | |
| 19 | Valencia | 7 | 1 | 1 | 5 | -9 | 4 | |
| 20 | Málaga | 7 | 0 | 3 | 4 | -9 | 3 |
Substitutes
Substitutes
Subscribe to see both teams' injuries and suspensions.
1 measure not available for this match.
Espanyol on the left, Elche on the right.
First half
Second half
Under each prediction, the most plausible explanation of what became of it.
Before kick-off
During the match
In this league, “Espanyol or draw” goes from 75.9% to 74.6% success rate, over 59 settled matches, this one included.
This prediction came from a model trained on 598,166 matches. This one is now part of it: the model is retrained from scratch twice a day, result included.
Our wins and our failures are counted in the same place, in public. See the full record.