Season 1 Championship: AI Betting Picks & ResultsBeta

The OddsGPT AI Championship publicly tracks AI betting strategies, closing line value (CLV), ROI and verified betting results.

Each strategy represents a different betting hypothesis and must build a statistically meaningful track record before entering the official rankings.

Jun 17, 2026 · May 31 – Jul 30, 2026 · Completed

AI Beyond Picks.

Value Beyond Odds.

Every AI strategy starts with a clean slate and must build a track record before entering the official rankings.

Premium betting site stake: New users can use the promo code FHn2uDc1 to receive $100 cash.

Season Review

Season 1 Championship Review

Anti-Steam AI won the Season 1 Championship with a season ROI of +118.2% across 60 settled picks.

Balanced AI finished second at +69.6% ROI.

Conservative AI recorded the highest win rate of the season at 100.0%.

Across five official competitors, the season produced 301 settled picks from May 31 – Jul 30, 2026.

🏆 Champion

😈 Anti-Steam AI

ROI +118.2% · Win Rate 68.3% · 41-19

AI Profile →

Final Standings

# AI Sample Confidence Positive CLV ROI Record Win Rate Units Won
1 ⚖️ Balanced AI 75 / 50 High Theory Verified 44% +47.4% 9-66 12.0% +35.6u
2 😈 Anti-Steam AI 62 / 50 High Theory Verified 61% +33.0% 20-42 32.3% +16.5u
3 🚀 Momentum AI 135 / 50 High Theory Verified 45% -12.5% 21-114 15.6% -32.2u

Provisional Standings

Building evidence toward official qualification.

Day Picks

⚖️ Balanced AI

1 Uzbekistan vs Colombia World Cup 2026-06-17 BET Uzbekistan Win 9.00 9.60 -0.7pp 78 🔥
Entry 104d ago
9.00
→
Expected Closing
9.60
Expected CLV -0.7pp
Why

Uzbekistan has a significant positive closing edge (28.91) and strong steam signals, indicating market undervaluation. The model probability (40.02%) is much higher than the market probability (11.11%). Main risk: Uzbekistan's lower ELO rating (-27.4) compared to Colombia.

Replaced with updated pick
Decision Lifecycle
  1. ✓ Pick Generated
  2. ✓ Market Tracked
  3. ✓ Closed
  4. ○ ROI Recorded
Bet Tracking Timeline Candidate pick @ 8.50
  1. 2026-06-11 Lens filter Lens reject: requires edge≥5 or steam≥75. (Current: steam=66, breadth=69%, signal=market_steam, kickoff_in=169.0h)
  2. 2026-06-12 Lens filter Lens reject: requires edge≥5 or steam≥75. (Current: steam=68, breadth=77%, signal=market_steam, kickoff_in=145.0h)
  3. 2026-06-13 Candidate pick @ 8.50
  4. 2026-06-13 Withdrawn before validation @ 8.50
  5. 2026-06-14 Pass Ranked 13/117 in candidate pool, exceeding daily cap of 3 picks. (Pre-LLM score 83.03). Main risk: N/A.
  6. 2026-06-15 Candidate pick @ 8.50
  7. 2026-06-15 Replaced with updated pick @ 8.50
  8. 2026-06-16 Candidate pick @ 9.00 (Candidate pick 8.50)
  9. 2026-06-17 Update @ 9.00 (Candidate pick 8.50)
  10. 2026-06-17 Replaced with updated pick @ 9.00 (Candidate pick 8.50)
  11. 2026-06-18 Candidate pick @ 10.00 (Candidate pick 8.50)
  12. 2026-06-18 Replaced with updated pick @ 10.00 (Candidate pick 8.50)
  13. — Closing @ 9.60
2 Tunisia vs Japan World Cup 2026-06-17 BET Tunisia Win 6.50 8.00 -2.9pp 74 🔥
Entry 104d ago
6.50
→
Expected Closing
8.00
Expected CLV -2.9pp
Why

Tunisia's odds have drifted significantly (from 4.75 to 6.5), creating a strong closing edge (23.92). The model probability (39.3%) is much higher than the market probability (15.38%). Main risk: Tunisia's large ELO deficit (-230.4) against Japan.

Withdrawn before validation
Decision Lifecycle
  1. ✓ Pick Generated
  2. ✓ Market Tracked
  3. ✓ Closed
  4. ○ ROI Recorded
Bet Tracking Timeline Candidate pick @ 7.50
  1. 2026-06-14 Pass Ranked 68/117 in candidate pool, exceeding daily cap of 20 picks. (Pre-LLM score 53.02). Main risk: N/A.
  2. 2026-06-15 Pass Ranked 19/66 in candidate pool, exceeding daily cap of 3 picks. (Pre-LLM score 53.02). Main risk: N/A.
  3. 2026-06-16 Candidate pick @ 7.50
  4. 2026-06-16 Replaced with updated pick @ 7.50
  5. 2026-06-17 Candidate pick @ 6.50 (Candidate pick 7.50)
  6. 2026-06-17 Withdrawn before validation @ 6.50 (Candidate pick 7.50)
  7. 2026-06-18 Pass LLM did not select this match (rank 3/94, pre-LLM score 96.52). Main risk: N/A.
  8. 2026-06-19 Pass Ranked 9/118 in candidate pool, exceeding daily cap of 3 picks. (Pre-LLM score 92.2). Main risk: N/A.
  9. 2026-06-20 Pass Ranked 15/138 in candidate pool, exceeding daily cap of 3 picks. (Pre-LLM score 92.2). Main risk: N/A.
  10. 2026-06-21 Pass Ranked 8/113 in candidate pool, exceeding daily cap of 3 picks. (Pre-LLM score 94.67). Main risk: N/A.
  11. — Closing @ 8.00
3 FC Slutsk vs Dinamo Minsk 0-2 ✗ Incorrect Coppa 2026-06-17 BET FC Slutsk Win 13.50 12.00 +0.9pp 70 🔥
Entry 104d ago
13.50
→
Expected Closing
12.00
Expected CLV +0.9pp
Why

FC Slutsk shows a strong closing edge (24.98) and significant model-market disparity (32.39% vs. 7.41%). The odds movement (-30.8% move_pct) suggests market overreaction. Main risk: FC Slutsk's negative xG differential (-0.12) and underdog status.

🔥 Sharp Money Signal
Decision Lifecycle
  1. ✓ Pick Generated
  2. ✓ Market Tracked
  3. ✓ Closed
  4. ✓ ROI Recorded
Bet Tracking Timeline Candidate pick @ 13.50
  1. 2026-06-15 Pass Ranked 20/66 in candidate pool, exceeding daily cap of 3 picks. (Pre-LLM score 51.2). Main risk: N/A.
  2. 2026-06-16 Pass Ranked 28/60 in candidate pool, exceeding daily cap of 20 picks. (Pre-LLM score 53.65). Main risk: N/A.
  3. 2026-06-17 Candidate pick @ 13.50
  4. — Closing @ 12.00

🚀 Momentum AI

1 Uzbekistan vs Colombia World Cup 2026-06-17 BET Uzbekistan Win 9.00 9.60 -0.7pp 85 🔥
Entry 104d ago
9.00
→
Expected Closing
9.60
Expected CLV -0.7pp
Why

Uzbekistan has seen sharp odds movement, with a strong steam signal and significant closing edge (28.91%). Market probability is undervalued compared to model probability (11.11% vs. 40.02%). Main risk: Elo rating favors Colombia, indicating potential class difference.

Replaced with updated pick
Decision Lifecycle
  1. ✓ Pick Generated
  2. ✓ Market Tracked
  3. ✓ Closed
  4. ○ ROI Recorded
Bet Tracking Timeline Candidate pick @ 8.50
  1. 2026-06-11 Pass Uzbekistan's odds have drifted higher, but the steam score and breadth percentage are moderate compared to the other candidate. Additionally, there is no model probability or closing edge data to support a bet. Main risk: Lack of supporting data and limited historical pattern make this fixture less compelling.
  2. 2026-06-12 Pass Uzbekistan's steam indicators are moderate, but the significant elo disadvantage and low market probability reduce confidence in backing the home side. Main risk: Uzbekistan's weak comparative metrics against Colombia.
  3. 2026-06-13 Pass Ranked 50/60 in candidate pool, exceeding daily cap of 20 picks. (Pre-LLM score 77.11). Main risk: N/A.
  4. 2026-06-14 Pass Ranked 42/46 in candidate pool, exceeding daily cap of 20 picks. (Pre-LLM score 77.11). Main risk: N/A.
  5. 2026-06-15 Candidate pick @ 8.50
  6. 2026-06-15 Replaced with updated pick @ 8.50
  7. 2026-06-16 Candidate pick @ 9.00 (Candidate pick 8.50)
  8. 2026-06-17 Update @ 9.00 (Candidate pick 8.50)
  9. 2026-06-17 Replaced with updated pick @ 9.00 (Candidate pick 8.50)
  10. 2026-06-18 Candidate pick @ 10.00 (Candidate pick 8.50)
  11. 2026-06-18 Replaced with updated pick @ 10.00 (Candidate pick 8.50)
  12. — Closing @ 9.60
2 Tunisia vs Japan World Cup 2026-06-17 BET Tunisia Win 6.50 8.00 -2.9pp 80 🔥
Entry 104d ago
6.50
→
Expected Closing
8.00
Expected CLV -2.9pp
Why

Tunisia's odds have drifted significantly, with strong steam and a closing edge of 23.92%. Model probability (39.3%) far exceeds market probability (15.38%). Main risk: Elo difference heavily favors Japan, suggesting Tunisia may struggle against a stronger opponent.

Withdrawn before validation
Decision Lifecycle
  1. ✓ Pick Generated
  2. ✓ Market Tracked
  3. ✓ Closed
  4. ○ ROI Recorded
Bet Tracking Timeline Candidate pick @ 7.50
  1. 2026-06-14 Pass Ranked 21/46 in candidate pool, exceeding daily cap of 20 picks. (Pre-LLM score 93.8). Main risk: N/A.
  2. 2026-06-15 Pass Ranked 7/17 in candidate pool, exceeding daily cap of 5 picks. (Pre-LLM score 93.8). Main risk: N/A.
  3. 2026-06-16 Candidate pick @ 7.50
  4. 2026-06-16 Replaced with updated pick @ 7.50
  5. 2026-06-17 Candidate pick @ 6.50 (Candidate pick 7.50)
  6. 2026-06-17 Withdrawn before validation @ 6.50 (Candidate pick 7.50)
  7. 2026-06-18 Pass Ranked 8/39 in candidate pool, exceeding daily cap of 5 picks. (Pre-LLM score 103.95). Main risk: N/A.
  8. 2026-06-19 Pass Ranked 21/59 in candidate pool, exceeding daily cap of 20 picks. (Pre-LLM score 99.89). Main risk: N/A.
  9. 2026-06-20 Pass Ranked 24/78 in candidate pool, exceeding daily cap of 20 picks. (Pre-LLM score 99.89). Main risk: N/A.
  10. 2026-06-21 Pass Ranked 17/54 in candidate pool, exceeding daily cap of 5 picks. (Pre-LLM score 102.36). Main risk: N/A.
  11. — Closing @ 8.00
3 Landvetter IS vs Jonsered 1-0 ✗ Incorrect Division 2 - Västra Götaland 2026-06-17 BET Jonsered Win 5.25 5.26 -0.0pp 78 🔥
Entry 104d ago
5.25
→
Expected Closing
5.26
Expected CLV -0.0pp
Why

Jonsered has strong steam and a sharp odds movement (+34.6%). Model probability (35.17%) is significantly higher than market probability (19.05%). Main risk: Elo difference favors Landvetter IS, indicating Jonsered may be outmatched.

🔥 Very Strong Money Signal
Decision Lifecycle
  1. ✓ Pick Generated
  2. ✓ Market Tracked
  3. ✓ Closed
  4. ✓ ROI Recorded
Bet Tracking Timeline Candidate pick @ 5.25
  1. 2026-06-16 Lens filter Lens reject: steam score must be ≥65; requires strong money signal. (Current: steam=36, breadth=0%, signal=model_edge, kickoff_in=40.0h)
  2. 2026-06-17 Candidate pick @ 5.25
  3. — Closing @ 5.26
4 Molodechno-DYuSSh 4 vs Bate Borisov 1-4 ✗ Incorrect Coppa 2026-06-17 BET Molodechno-DYuSSh 4 Win 6.84 9.00 -3.5pp 77 🔥
Entry 104d ago
6.84
→
Expected Closing
9.00
Expected CLV -3.5pp
Why

Molodechno-DYuSSh 4 has strong steam and significant odds movement (+42.2%). Model probability (32.85%) is much higher than market probability (14.62%). Main risk: Elo and xG differences favor Bate Borisov, suggesting a class gap.

🔥 Sharp Money Signal
Decision Lifecycle
  1. ✓ Pick Generated
  2. ✓ Market Tracked
  3. ✓ Closed
  4. ✓ ROI Recorded
Bet Tracking Timeline Candidate pick @ 6.84
  1. 2026-06-15 Lens filter Lens reject: steam score must be ≥65; requires strong money signal. (Current: steam=32, breadth=0%, signal=model_edge, kickoff_in=60.0h)
  2. 2026-06-16 Lens filter Lens reject: steam score must be ≥65; requires strong money signal. (Current: steam=39, breadth=0%, signal=model_edge, kickoff_in=36.0h)
  3. 2026-06-17 Candidate pick @ 6.84
  4. — Closing @ 9.00
5 Ecuador vs Curaçao World Cup 2026-06-17 FADE Draw 11.00 8.00 +3.4pp 65 🔥
Entry 104d ago
11.00
→
Expected Closing
8.00
Expected CLV +3.4pp
Why

The draw option for Ecuador vs Curaçao has seen extreme odds movement (+83.3%), but historical patterns are limited, and volatility is high. Elo and xG metrics slightly favor Ecuador. Main risk: The sharp odds movement may reflect speculative betting rather than genuine value.

Replaced with updated pick
Decision Lifecycle
  1. ✓ Pick Generated
  2. ✓ Market Tracked
  3. ✓ Closed
  4. ○ ROI Recorded
Bet Tracking Timeline Candidate pick @ 11.00
  1. 2026-06-14 Pass Ranked 22/46 in candidate pool, exceeding daily cap of 20 picks. (Pre-LLM score 93.22). Main risk: N/A.
  2. 2026-06-15 Pass LLM did not select this match (rank 2/17, pre-LLM score 103.95). Main risk: N/A.
  3. 2026-06-16 Candidate pick @ 10.00
  4. 2026-06-16 Replaced with updated pick @ 10.00
  5. 2026-06-17 Candidate pick @ 11.00
  6. 2026-06-17 Replaced with updated pick @ 11.00
  7. 2026-06-18 Candidate pick @ 11.00
  8. 2026-06-18 Withdrawn before validation @ 11.00
  9. 2026-06-19 Pass Ranked 10/59 in candidate pool, exceeding daily cap of 5 picks. (Pre-LLM score 102.36). Main risk: N/A.
  10. 2026-06-20 Pass Ranked 15/78 in candidate pool, exceeding daily cap of 5 picks. (Pre-LLM score 102.36). Main risk: N/A.
  11. — Closing @ 8.00
Hypothesis: Late market moves contain information that improves betting decisions.
Theory Building Confidence: Medium
Building Qualification 20 / 50 Settled Bets
30 more settled picks required for official ranking
ROI
-100.0%
Positive CLV
Insufficient Data
Record 0-2 · Win Rate 0.0% · Profit -4.0u · Total Staked 4.0u · Avg Stake 2.00u
Recent Results L L
Latest Picks
  • Uzbekistan Win
  • Tunisia Win
  • Jonsered Win
Hypothesis: Combining multiple signals is more reliable than relying on a single factor.
Theory Building Confidence: Medium
Building Qualification 12 / 50 Settled Bets
38 more settled picks required for official ranking
ROI
-100.0%
Positive CLV
Insufficient Data
Record 0-1 · Win Rate 0.0% · Profit -1.0u · Total Staked 1.0u · Avg Stake 1.00u
Recent Results L
Latest Picks
  • Uzbekistan Win
  • Tunisia Win
  • FC Slutsk Win
Hypothesis: Recurring market patterns appear often enough to support systematic betting.
No picks tracked yet
Theory in Testing Confidence: Low
Building Qualification 0 / 50 Settled Bets
50 more settled picks required for official ranking
Hypothesis: Selective betting beats betting volume.
No picks tracked yet
Theory in Testing Confidence: Low
Building Qualification 1 / 50 Settled Bets
49 more settled picks required for official ranking
Hypothesis: The market sometimes overreacts, creating value on the opposite side.
No picks tracked yet
Theory in Testing Confidence: Low
Building Qualification 9 / 50 Settled Bets
41 more settled picks required for official ranking

Confidence reflects sample size only. Each AI needs 50 settled picks to enter the official ranking. CLV rate appears after 10 settled picks.

Community Benchmark

Hypothesis: Multiple independent opinions outperform individual opinions.
No picks tracked yet
Theory in Testing Confidence: Low
Building Qualification 4 / 50 Settled Bets
46 more settled picks required for official ranking

Why This Matters

Every bettor has a theory.

  • Follow the market.
  • Fade the market.
  • Trust only high-confidence picks.
  • Look for recurring patterns.

We track each approach publicly and measure the results over time.

The goal is simple:

Which ideas actually beat the market?

Example Betting Hypotheses

  • Following major market moves reveals useful information.
  • Betting against popular market sentiment creates value.
  • Fewer high-conviction bets outperform high-volume betting.
  • Combining multiple signals is more reliable than relying on a single factor.
  • Recurring market patterns can be identified and exploited over time.
  • Consensus opinions outperform individual opinions.

The Championship does not assume any of these ideas are true.

The purpose is to test them publicly.

← Current Championship