🎯 The Prediction Engine
🌐 The Ecosystem
This system is built across five connected tools. Here is how they talk to each other:
1. The Data Centre
/fouls – The main engine. Enter player data, calculate probabilities, view tactical heatmaps, and generate Smart Bets.
2. The Betting Slips
/slips – The profile manager. Create user profiles and view the specific bets saved from the Data Centre.
3. Match Lists
/foul-viewer – The library. View, copy, and manage every match you’ve ever processed and saved.
4. The Audit Log
/checker – The feedback loop. Input actual match results to grade the engine’s accuracy and recalibrate algorithm bias.
5. Change Logs
/change-log – The history. A complete, version-controlled record of every mathematical and UI update made to the system.
🛠️ Step-by-Step: User Process Flow
Step 1: Match Setup & Data Entry
Navigate to The Data Centre. Enter the Home Team, Away Team, and Kick-Off details. Input the Starting XI for both teams. You will need:
- Position: Crucial for the tactical matchup multipliers.
- FC p90: Fouls Committed per 90 minutes.
- FW p90: Fouls Won per 90 minutes.
- Expected Mins: Adjust this if a player is likely to be subbed early. The engine recalculates their foul probability based on the time they are actually on the pitch.
Step 2: Generate & Save
Once your data is in, check “Lineup Confirmed” to view the top-down tactical pitch. Then, click Save Match. This pushes the data to the server and generates a unique 8-character code so you can reload it later via Match Lists.
Step 3: Analysis & Building Slips
Click Calculate All Predictions. The engine will instantly generate the dashboard:
- Smart Picks: Categorized into Bankers, Mids, and Long Shots based on risk-to-reward ratio.
- Forecast Matrix: A 0.0 to 9.9 Likelihood Rating for every player on the pitch.
- Smart Bet Builder: Scroll to the bottom to automatically generate value-hunting Accumulators that strictly avoid related contingencies.
Found a bet you like? Use the dropdown menus to Add to Slip. This sends the bet directly to that user’s profile on The Betting Slips page.
Step 4: The Feedback Loop
After the match finishes, go to The Audit Log. Enter the Actual fouls committed and won by each player. The system calculates a granular Accuracy Score. If the global average drops below 75.0, the system will prompt you to recalibrate the mathematical bias.
🧮 The Mathematics (Under the Hood)
The predictor does not just look at averages; it uses a calibrated model tailored to the chaos of real-world football.
📈 Calibrated Poisson Distribution
To determine the probability of a player hitting exactly 1+, 2+, or 3+ fouls, the engine uses a Poisson distribution formula based on their expected minutes. We then apply Realism Caps to account for the unpredictable nature of sport:
- 1+ Fouls: Hard capped at 88.0%
- 2+ Fouls: Hard capped at 54.0%
- 3+ Fouls: Hard capped at 32.0%
⚔️ Positional Friction Multipliers
Not all matchups are created equal. A winger running at a fullback generates more fouls than a center-back passing to another center-back. The algorithm compares the Aggressor’s FC/90 against the Victim’s FW/90 and applies a multiplier:
| Matchup Type | Examples | Multiplier |
|---|---|---|
| High Conflict | LW vs RB, ST vs CB | 1.6x |
| Midfield Battle | CDM vs CAM | 1.5x |
| Engine Room | CM vs CM | 1.4x |
| Low Friction | Non-clashing positions | 0.75x |
🟨 The Yellow Card Derivative & Fair Value
Statistically, a player committing multiple fouls is at a severe risk of a booking. The algorithm derives a “Yellow Card” probability by calculating 35% of the player’s 2+ Foul probability.
Furthermore, every percentage generated by the engine is instantly converted into Fair Value Odds (Decimal format) so you know exactly what price a bookmaker should be offering.