Eredivisie Week 25 AI Model Performance Audit
Gemma 3 12B led Eredivisie predictions with 1.78 avg points/match, followed by Llama 4 Scout and MiniMax M2.1 at 1.44. Models achieved 19.30% correct tendency overall, with the 2-3 upset in NEC Nijmegen vs Fortuna Sittard being the biggest surprise.
Gemma 3 12B led Eredivisie predictions with 1.78 avg points/match, followed by Llama 4 Scout and MiniMax M2.1 at 1.44. Models achieved 19.30% correct tendency overall, with the 2-3 upset in NEC Nijmegen vs Fortuna Sittard being the biggest surprise.
Eredivisie Regular Season - 25 featured 9 matches, including fixtures like Twente vs Feyenoord and Heracles vs PSV Eindhoven. AI prediction accuracy is critical for assessing model reliability in competitive league rounds. This audit examines statistical performance based on provided data.
Top 10 Models
| # | Model | Matches | Total Points | Avg Pts/Match | Tendency % | Exact % |
|---|---|---|---|---|---|---|
| 1 | Gemma 3 12B (OpenRouter) | 9 | 16 | 1.78 | 22.2% | 22.2% |
| 2 | Llama 4 Scout (OpenRouter) | 9 | 13 | 1.44 | 22.2% | 11.1% |
| 3 | MiniMax M2.1 (OpenRouter) | 9 | 13 | 1.44 | 22.2% | 11.1% |
| 4 | Phi-4 (OpenRouter) | 9 | 13 | 1.44 | 22.2% | 11.1% |
| 5 | Kimi K2.5 (OpenRouter) | 9 | 12 | 1.33 | 33.3% | 0.0% |
| 6 | GLM-4.7 (OpenRouter) | 9 | 12 | 1.33 | 33.3% | 0.0% |
| 7 | GLM-5 (OpenRouter) | 9 | 10 | 1.11 | 22.2% | 11.1% |
| 8 | DeepSeek R1-0528 (OpenRouter) | 9 | 10 | 1.11 | 22.2% | 11.1% |
| 9 | Devstral 2 (OpenRouter) | 9 | 10 | 1.11 | 22.2% | 11.1% |
| 10 | Mistral Small 3.2 24B (OpenRouter) | 9 | 10 | 1.11 | 22.2% | 11.1% |
Match-by-Match Audit
- Excelsior vs GO Ahead Eagles: Result 0-1, tendency 10.5%, exact 0.0%, consensus D (73.7%) incorrect.
- Utrecht vs AZ Alkmaar: Result 2-0, tendency 0.0%, exact 0.0%, consensus D (57.9%) incorrect.
- Twente vs Feyenoord: Result 2-0, tendency 0.0%, exact 0.0%, consensus A (52.6%) incorrect.
- FC Volendam vs Groningen: Result 3-2, tendency 10.5%, exact 0.0%, consensus D (68.4%) incorrect.
- PEC Zwolle vs Ajax: Result 0-0, tendency 42.1%, exact 0.0%, consensus A (52.6%) incorrect.
- NEC Nijmegen vs Fortuna Sittard: Result 2-3, tendency 5.3%, exact 0.0%, consensus H (84.2%) incorrect.
- Heracles vs PSV Eindhoven: Result 1-3, tendency 94.7%, exact 57.9%, consensus A (94.7%) correct.
- Heerenveen vs Sparta Rotterdam: Result 2-1, tendency 10.5%, exact 5.3%, consensus A (52.6%) incorrect.
- Telstar vs NAC Breda: Result 3-0, tendency 0.0%, exact 0.0%, consensus D (84.2%) incorrect.
Biggest Consensus Misses
- NEC Nijmegen vs Fortuna Sittard (2-3) | Consensus: H (84.2%) | Counts H/D/A: 16/2/1
- Telstar vs NAC Breda (3-0) | Consensus: D (84.2%) | Counts H/D/A: 0/16/3
- Excelsior vs GO Ahead Eagles (0-1) | Consensus: D (73.7%) | Counts H/D/A: 3/14/2
- FC Volendam vs Groningen (3-2) | Consensus: D (68.4%) | Counts H/D/A: 2/13/4
- Utrecht vs AZ Alkmaar (2-0) | Consensus: D (57.9%) | Counts H/D/A: 0/11/8
Methodology
kroam.xyz uses a quota-based scoring system that rewards both accuracy and boldness:
Tendency Points (2-6 points): Models earn points for correctly predicting the match outcome (home win, draw, or away win). The points awarded depend on prediction rarity—if most models predicted a home win but the away team won, models who correctly predicted the away win earn more points (up to 6). Common predictions earn fewer points (minimum 2).
Goal Difference Bonus (+1 point): If the model predicts the correct goal difference (e.g., predicted 2-1 and result was 3-2, both +1 difference), they earn a bonus point.
Exact Score Bonus (+3 points): Predicting the exact final score earns 3 additional points.
Maximum: 10 points per prediction (6 tendency + 1 goal diff + 3 exact).
This system ensures that models taking calculated risks on unlikely outcomes are rewarded when correct, while also recognizing precision in exact score predictions. Learn more about our methodology.
Frequently Asked Questions
Q: Which AI model performed best in Eredivisie Regular Season - 25? A: Gemma 3 12B (OpenRouter) performed best with 1.78 average points per match.
Q: How accurate were AI predictions for Eredivisie this round? A: Models achieved 19.30% correct tendency and 7.02% exact score hit rate on average.
Q: What was the biggest upset in Eredivisie Regular Season - 25? A: The biggest consensus miss was NEC Nijmegen vs Fortuna Sittard (2-3), with 84.2% of models incorrectly predicting a home win.
Q: How does kroam.xyz score AI football predictions? A: kroam.xyz uses a quota-based system awarding up to 10 points per match for tendency, goal difference, and exact score accuracy.
Generation cost: $0.0021
Tokens: 4,947 input + 1,810 output
Frequently Asked Questions
What is this article about?
Which AI model performed best in Eredivisie Regular Season - 25?**?
Q: Which AI model performed best in Eredivisie Regular Season - 25?
Q: How accurate were AI predictions for Eredivisie this round?
You might also like
Eredivisie Week 24 AI Model Audit: Llama 4 Scout Leads
Llama 4 Scout (OpenRouter) performed best with 4.00 avg points/match, followed by Trinity Large Preview (3.22) and Llama 3.3 70B Instruct (3.00). Overall accuracy was 52.05% correct tendency. The biggest upset was Twente's 2-1 win over Groningen, which fooled 84.2% of models predicting a draw.
Feb 23, 2026
UEFA Conference League Round of 32 AI Prediction Audit
GPT-OSS 20B led UEFA Conference League predictions with 2.88 points per match, followed by Trinity Large Preview (2.63) and GLM-5 (2.25). Models achieved 38.16% correct tendency overall, with Fiorentina vs Jagiellonia (2-4) as the biggest upset.
Mar 2, 2026
UEFA Europa League Round of 32 AI Model Performance Audit
Mistral Small 3.2 24B led predictions with 3.38 avg points/match, followed by Phi-4 (2.88) and Llama 4 Scout (2.75). Models achieved 38.82% correct tendency. VfB Stuttgart's 0-1 loss to Celtic was the biggest consensus miss.