5.8M+ predictions logged. The full paper-trading record is below. The day a model earns verified live status, its picks publish free to Telegram — subscribe to be first in the room.
We run 13 independent models in parallel across every sport — each one a separate experiment in finding a real edge against the market. The table below shows the top-performing combinations: every (sport, market, model) combination clearing +5% ROI on a minimum of 10 settled paper bets, out of the full set we evaluate. Most combinations won't hold an edge — that's expected. Searching this wide means most experiments fail, and a handful prove out. The table is the current leaders; the graduation gates exist to separate a genuine edge from a lucky run before any combination is ever published as a live track record.
| Sport | Market | Model | Record (W-L) | Strike rate | ROI |
|---|---|---|---|---|---|
| AFL | H2H | Model B | 39-26 | 60.0% | +11.2% |
| AFL | H2H | Model D | 41-29 | 58.6% | +10.2% |
| AFL | Spreads | Model C | 50-35 | 58.8% | +14.2% |
| AFL | Spreads | Model G | 36-27 | 57.1% | +13.0% |
| AFL | Spreads | Model B | 51-37 | 58.0% | +12.3% |
| AFL | Spreads | Model I | 52-38 | 57.8% | +12.2% |
| AFL | Spreads | Model F | 51-40 | 56.0% | +8.3% |
| AFL | Spreads | Model D | 50-41 | 54.9% | +6.5% |
| A-League | H2H | Model D | 8-8 | 50.0% | +95.8% |
| A-League | H2H | Model F | 13-16 | 44.8% | +84.5% |
| A-League | H2H | Model A | 11-15 | 42.3% | +43.7% |
| A-League | H2H | Model I | 6-15 | 28.6% | +12.9% |
| A-League | Totals | Model D | 8-3-1 | 72.7% | +46.7% |
| A-League | Totals | Model G | 9-4 | 69.2% | +45.8% |
| A-League | Totals | Model I | 7-4-1 | 63.6% | +30.7% |
| A-League | Totals | Model A | 11-12-1 | 47.8% | +10.4% |
| Bundesliga | H2H | Model H | 23-31 | 42.6% | +38.3% |
| Bundesliga | H2H | Model E | 20-23 | 46.5% | +35.9% |
| Bundesliga | H2H | Model C | 17-34 | 33.3% | +24.7% |
| Bundesliga | H2H | Model G | 24-31 | 43.6% | +24.5% |
| Bundesliga | Spreads | Model E | 23-15-4 | 60.5% | +22.5% |
| Bundesliga | Spreads | Model F | 21-12-6 | 63.6% | +13.5% |
| Bundesliga | Spreads | Model H | 23-20-7 | 53.5% | +8.7% |
| Bundesliga | Totals | Model M | 31-19-4 | 62.0% | +24.1% |
| Bundesliga | Totals | Model A | 28-21-3 | 57.1% | +16.6% |
| Bundesliga | Totals | Model E | 24-19 | 55.8% | +12.0% |
| Bundesliga | Totals | Model D | 27-24-3 | 52.9% | +10.7% |
| EPL | H2H | Model D | 8-13 | 38.1% | +45.9% |
| EPL | H2H | Model F | 15-38 | 28.3% | +15.3% |
| EPL | H2H | Model A | 20-38 | 34.5% | +14.7% |
| EPL | H2H | Model K | 20-26 | 43.5% | +13.3% |
| EPL | H2H | Model J | 9-16 | 36.0% | +13.3% |
| EPL | Spreads | Model A | 36-15-10 | 70.6% | +20.7% |
| EPL | Spreads | Model M | 32-18-12 | 64.0% | +9.0% |
| EPL | Totals | Model A | 38-22-6 | 63.3% | +29.9% |
| EPL | Totals | Model B | 39-24-6 | 61.9% | +26.0% |
| EPL | Totals | Model M | 35-24-5 | 59.3% | +22.4% |
| EPL | Totals | Model F | 13-10-4 | 56.5% | +14.3% |
| La Liga | H2H | Model D | 13-22 | 37.1% | +34.8% |
| La Liga | H2H | Model I | 19-39 | 32.8% | +17.4% |
| La Liga | H2H | Model A | 26-44 | 37.1% | +15.5% |
| La Liga | H2H | Model F | 14-31 | 31.1% | +14.2% |
| La Liga | H2H | Model H | 22-42 | 34.4% | +13.6% |
| La Liga | H2H | Model E | 25-34 | 42.4% | +11.9% |
| La Liga | H2H | Model G | 31-43 | 41.9% | +6.8% |
| La Liga | Totals | Model E | 29-21-1 | 58.0% | +20.1% |
| La Liga | Totals | Model G | 33-30-2 | 52.4% | +13.7% |
| La Liga | Totals | Model J | 37-33-2 | 52.9% | +10.2% |
| MLB | H2H | Model C | 99-126-1 | 44.0% | +5.8% |
| NBA | H2H | Model D | 23-41 | 35.9% | +36.7% |
| NBA | H2H | Model G | 42-52 | 44.7% | +35.4% |
| NBA | H2H | Model J | 19-45 | 29.7% | +25.8% |
| NBA | H2H | Model E | 39-55 | 41.5% | +17.4% |
| NBA | H2H | Model K | 15-21 | 41.7% | +12.2% |
| NBA | H2H | Model C | 36-55 | 39.6% | +9.4% |
| NBA | H2H | Model I | 36-61 | 37.1% | +6.0% |
| NBA | H2H | Model A | 47-54 | 46.5% | +5.4% |
| NBA | Spreads | Model F | 38-37 | 50.7% | +7.6% |
| NBA | Totals | Model G | 37-29 | 56.1% | +21.5% |
| NBA | Totals | Model E | 40-25 | 61.5% | +18.9% |
| NBA | Totals | Model B | 45-34 | 57.0% | +16.2% |
| NBA | Totals | Model J | 40-33 | 54.8% | +11.7% |
| NBA | Totals | Model M | 45-40 | 52.9% | +6.6% |
| NCAAB | H2H | Model E | 7-4 | 63.6% | +28.2% |
| NCAAB | H2H | Model J | 10-5 | 66.7% | +16.7% |
| NCAAB | H2H | Model H | 10-6 | 62.5% | +13.4% |
| NCAAB | H2H | Model I | 7-5 | 58.3% | +12.2% |
| NHL | H2H | Model E | 70-60 | 53.8% | +11.0% |
| NHL | H2H | Model I | 52-56 | 48.1% | +6.0% |
| NHL | Spreads | Model J | 35-45-4 | 43.8% | +6.6% |
| NHL | Spreads | Model E | 39-27-2 | 59.1% | +5.6% |
| NHL | Totals | Model F | 37-31-3 | 54.4% | +9.7% |
| NHL | Totals | Model D | 51-47-5 | 52.0% | +6.3% |
| NRL | H2H | Model H | 46-35 | 56.8% | +20.0% |
| NRL | H2H | Model C | 25-37 | 40.3% | +15.8% |
| NRL | H2H | Model J | 35-24 | 59.3% | +12.4% |
| NRL | H2H | Model K | 35-27 | 56.5% | +11.0% |
| NRL | Spreads | Model G | 16-13 | 55.2% | +11.9% |
| NRL | Spreads | Model J | 46-37 | 55.4% | +9.3% |
| NRL | Spreads | Model D | 46-39 | 54.1% | +6.7% |
| NRL | Spreads | Model B | 45-38 | 54.2% | +6.4% |
| NRL | Spreads | Model F | 42-36 | 53.8% | +5.5% |
| NRL | Totals | Model D | 57-30 | 65.5% | +26.4% |
| NRL | Totals | Model C | 49-31 | 61.3% | +18.2% |
| NRL | Totals | Model A | 51-36 | 58.6% | +13.0% |
| NRL | Totals | Model J | 41-32 | 56.2% | +8.3% |
| Serie A | H2H | Model B | 23-24 | 48.9% | +38.6% |
| Serie A | H2H | Model H | 19-24 | 44.2% | +32.3% |
| Serie A | H2H | Model F | 18-17 | 51.4% | +22.2% |
| Serie A | H2H | Model K | 18-26 | 40.9% | +8.9% |
| Serie A | H2H | Model G | 22-29 | 43.1% | +8.2% |
| Serie A | Spreads | Model I | 13-16-7 | 44.8% | +7.0% |
| Serie A | Totals | Model E | 28-15 | 65.1% | +34.6% |
| Serie A | Totals | Model G | 22-16-2 | 57.9% | +25.2% |
| Serie A | Totals | Model M | 27-22-1 | 55.1% | +14.6% |
| Serie A | Totals | Model J | 27-25-1 | 51.9% | +10.5% |
| Serie A | Totals | Model B | 26-25-1 | 51.0% | +8.6% |
Data current as of July 2026. Showing the top-performing combinations — ROI above +5% on at least 10 settled paper bets. The graduation gates exist to prove that edge before anything is ever published live.
All figures are based on simulated paper trading — not real-money wagering. Each prediction is recorded with a flat notional stake of AUD $100 at the best available global bookmaker odds at prediction time. ROI is calculated as total simulated profit or loss divided by total notional amount staked. These figures do not account for bookmaker account limitations, odds movement between prediction and placement, market liquidity, or other real-world execution factors. Actual results could differ materially from the reported figures. Paper trading since February 2026.
We build statistical models that analyse sports data. Here is exactly what they do, what they cannot do, and why we tell you both.
Our 13 independent models ingest historical match data, team performance statistics, and real-time bookmaker odds across 16 sports. Each model uses a different statistical methodology to estimate outcome probabilities. When a model's probability diverges significantly from the market's implied probability, we flag it and publish the analysis.
Every model runs independently -- no shared predictions, no committee decisions. We track each model's accuracy by sport and market type over time. The model with the strongest verified track record in a given sport will earn the right to generate subscriber alerts. Models that underperform lose their reporting slot. The data decides, not us.
Models are retrained as new data accumulates. New statistical approaches are tested in shadow mode -- logging predictions without generating alerts -- until they demonstrate sustained accuracy against real market odds.
Each model uses a different statistical approach — running independently, with no shared predictions, using real-time odds from multiple bookmakers.
Every market is scanned for statistical divergence. An alert is generated only when a model's calculated probability exceeds the bookmaker's implied probability by at least 5%. Data, not opinion.
Each sport and market combination is tracked across all 13 models. A model only generates subscriber alerts after it has earned its position through sustained, verified accuracy. Sports where no model has yet qualified do not generate alerts.
When a qualified model identifies a divergence, the alert is published — with the probability, the odds, and the data behind it.
Follow the public channel to see real model analysis in real time before subscribing. No credit card. No commitment.
We're building something different — rigorous, transparent AI sports analytics. Register your interest to be among the first to access gpunt when we launch Foundation Subscriptions.