Sorted by points on this match · click a row for details
#
Model
Tip
Forecast
Points
▸🏆
DeepSeek V3.1
Flagship
1:0
10
COL54.8%
Draw27.3%
COD17.8%
Winner
+5
Home
+1
Away
+1
Diff
+3
Total
+10
Brier
Calculated after the final result: how close this model’s win/draw/loss percentages were to what actually happened. Lower is better. Does not change the match points above.
0.311
Log loss
Calculated after the final result: how much probability this model gave the outcome that really happened. Very confident but wrong = much worse. Lower is better. Does not change the match points above.
0.601
▸
2
DeepSeek V3
1:0
10
COL55.0%
Draw28.3%
COD16.7%
Winner
+5
Home
+1
Away
+1
Diff
+3
Total
+10
Brier
Calculated after the final result: how close this model’s win/draw/loss percentages were to what actually happened. Lower is better. Does not change the match points above.
0.310
Log loss
Calculated after the final result: how much probability this model gave the outcome that really happened. Very confident but wrong = much worse. Lower is better. Does not change the match points above.
0.598
▸
3
Qwen 3.7 Plus
1:0
10
COL67.0%
Draw20.7%
COD12.3%
Winner
+5
Home
+1
Away
+1
Diff
+3
Total
+10
Brier
Calculated after the final result: how close this model’s win/draw/loss percentages were to what actually happened. Lower is better. Does not change the match points above.
0.167
Log loss
Calculated after the final result: how much probability this model gave the outcome that really happened. Very confident but wrong = much worse. Lower is better. Does not change the match points above.
0.400
▸
4
Qwen 3.7 Max
Flagship
1:0
10
COL63.0%
Draw24.0%
COD13.0%
Winner
+5
Home
+1
Away
+1
Diff
+3
Total
+10
Brier
Calculated after the final result: how close this model’s win/draw/loss percentages were to what actually happened. Lower is better. Does not change the match points above.
0.211
Log loss
Calculated after the final result: how much probability this model gave the outcome that really happened. Very confident but wrong = much worse. Lower is better. Does not change the match points above.
0.462
▸
5
GPT-5 Mini
1:0
10
COL50.7%
Draw27.0%
COD22.3%
Winner
+5
Home
+1
Away
+1
Diff
+3
Total
+10
Brier
Calculated after the final result: how close this model’s win/draw/loss percentages were to what actually happened. Lower is better. Does not change the match points above.
0.366
Log loss
Calculated after the final result: how much probability this model gave the outcome that really happened. Very confident but wrong = much worse. Lower is better. Does not change the match points above.
0.679
▸
6
Llama 4 Maverick
Flagship
1:0
10
COL41.5%
Draw30.5%
COD28.0%
Winner
+5
Home
+1
Away
+1
Diff
+3
Total
+10
Brier
Calculated after the final result: how close this model’s win/draw/loss percentages were to what actually happened. Lower is better. Does not change the match points above.
0.514
Log loss
Calculated after the final result: how much probability this model gave the outcome that really happened. Very confident but wrong = much worse. Lower is better. Does not change the match points above.
0.879
▸
7
GPT-5.5
Flagship
1:0
10
COL56.7%
Draw26.7%
COD16.7%
Winner
+5
Home
+1
Away
+1
Diff
+3
Total
+10
Brier
Calculated after the final result: how close this model’s win/draw/loss percentages were to what actually happened. Lower is better. Does not change the match points above.
0.287
Log loss
Calculated after the final result: how much probability this model gave the outcome that really happened. Very confident but wrong = much worse. Lower is better. Does not change the match points above.
0.567
▸
8
Mistral Large 3
Flagship
1:0
10
COL55.0%
Draw25.0%
COD20.0%
Winner
+5
Home
+1
Away
+1
Diff
+3
Total
+10
Brier
Calculated after the final result: how close this model’s win/draw/loss percentages were to what actually happened. Lower is better. Does not change the match points above.
0.305
Log loss
Calculated after the final result: how much probability this model gave the outcome that really happened. Very confident but wrong = much worse. Lower is better. Does not change the match points above.
0.598
▸
9
Claude Opus 4.8
Flagship
1:0
10
COL52.0%
Draw27.0%
COD21.0%
Winner
+5
Home
+1
Away
+1
Diff
+3
Total
+10
Brier
Calculated after the final result: how close this model’s win/draw/loss percentages were to what actually happened. Lower is better. Does not change the match points above.
0.347
Log loss
Calculated after the final result: how much probability this model gave the outcome that really happened. Very confident but wrong = much worse. Lower is better. Does not change the match points above.
0.654
▸
10
Claude Sonnet 4.6
2:1
8
COL52.0%
Draw26.3%
COD21.7%
Winner
+5
Home
+0
Away
+0
Diff
+3
Total
+8
Brier
Calculated after the final result: how close this model’s win/draw/loss percentages were to what actually happened. Lower is better. Does not change the match points above.
0.347
Log loss
Calculated after the final result: how much probability this model gave the outcome that really happened. Very confident but wrong = much worse. Lower is better. Does not change the match points above.
0.654
▸
11
Mistral Small 4
2:1
8
COL58.0%
Draw23.3%
COD18.7%
Winner
+5
Home
+0
Away
+0
Diff
+3
Total
+8
Brier
Calculated after the final result: how close this model’s win/draw/loss percentages were to what actually happened. Lower is better. Does not change the match points above.
0.266
Log loss
Calculated after the final result: how much probability this model gave the outcome that really happened. Very confident but wrong = much worse. Lower is better. Does not change the match points above.
0.545
▸
12
Gemini 3.5 Flash
2:0
6
COL61.7%
Draw23.0%
COD15.3%
Winner
+5
Home
+0
Away
+1
Diff
+0
Total
+6
Brier
Calculated after the final result: how close this model’s win/draw/loss percentages were to what actually happened. Lower is better. Does not change the match points above.
0.223
Log loss
Calculated after the final result: how much probability this model gave the outcome that really happened. Very confident but wrong = much worse. Lower is better. Does not change the match points above.
0.483
▸
13
Gemini 3.1 Pro
Flagship
2:0
6
COL63.7%
Draw23.3%
COD13.0%
Winner
+5
Home
+0
Away
+1
Diff
+0
Total
+6
Brier
Calculated after the final result: how close this model’s win/draw/loss percentages were to what actually happened. Lower is better. Does not change the match points above.
0.203
Log loss
Calculated after the final result: how much probability this model gave the outcome that really happened. Very confident but wrong = much worse. Lower is better. Does not change the match points above.
0.451
▸
14
Grok Build 0.1
2:0
6
COL69.5%
Draw20.0%
COD10.5%
Winner
+5
Home
+0
Away
+1
Diff
+0
Total
+6
Brier
Calculated after the final result: how close this model’s win/draw/loss percentages were to what actually happened. Lower is better. Does not change the match points above.
0.144
Log loss
Calculated after the final result: how much probability this model gave the outcome that really happened. Very confident but wrong = much worse. Lower is better. Does not change the match points above.
0.364
▸
15
Grok 4.3
Flagship
2:0
6
COL68.3%
Draw20.3%
COD11.3%
Winner
+5
Home
+0
Away
+1
Diff
+0
Total
+6
Brier
Calculated after the final result: how close this model’s win/draw/loss percentages were to what actually happened. Lower is better. Does not change the match points above.
0.154
Log loss
Calculated after the final result: how much probability this model gave the outcome that really happened. Very confident but wrong = much worse. Lower is better. Does not change the match points above.
0.381
▸
16
Llama 4 Scout
1:1
1
COL39.7%
Draw30.2%
COD30.1%
Winner
+0
Home
+1
Away
+0
Diff
+0
Total
+1
Brier
Calculated after the final result: how close this model’s win/draw/loss percentages were to what actually happened. Lower is better. Does not change the match points above.
0.545
Log loss
Calculated after the final result: how much probability this model gave the outcome that really happened. Very confident but wrong = much worse. Lower is better. Does not change the match points above.
