How the model works
Every FootyMind pick goes through the same five steps, applied the same way to every fixture – a Champions League night or a second-division Tuesday.
1. Team ratings
Each team carries two numbers: an attack rating (how many goals it tends to score against an average side) and a defence rating (how many it tends to concede). They are rebuilt after every result from roughly a season of matches, with recent games weighted more heavily, and adjusted for the strength of the opposition.
2. Expected goals
For each fixture the two attack and defence ratings, plus a home-advantage factor for the league, give each side's expected goals – for example 1.8 for the home team and 0.9 for the visitors.
3. The score grid
Those two numbers become a grid of every plausible scoreline – 0–0, 1–0, 1–1, 2–1 and so on – each with its own probability. Low scores are adjusted slightly, because real football produces more 0–0 and 1–1 results than a basic model expects.
4. Every market from one grid
Add up the right cells and the grid answers every market at once: home, draw or away; double chance; over/under 1.5, 2.5 and 3.5; both teams to score; Asian handicap margins; the likeliest correct score; and, by splitting the projection between halves, the half-time markets. Because they share one grid, a match's picks never contradict each other.
5. The Mind Score and the price
The Mind Score is simply the model's probability for a pick, written out of 100. For each match the model looks for the market with the best balance of a high Mind Score and a fair bookmaker price – that becomes the Mind pick. A pick is only published when its price is not shorter than the model can justify; when the price is clearly longer than fair, it also appears among the value picks.
From picks to slips
The banker is the single strongest pick at a fair price. The free slips combine high-scoring legs into the target price with the best joint probability, using only senior fixtures from 11:00 WAT onwards. Very short prices (around 1.12 and below) are kept out of featured picks unless they are a straight home or away win.
Settling and the record
Picks lock at kick-off and settle on the score after 90 minutes plus stoppage time – extra time and penalties do not count, and postponed games are void. Every result goes onto the track record, which also checks calibration: do MS 70 picks really land about 70% of the time?
What the model cannot see
Ratings do not know about a late injury, a rotated line-up or a manager under pressure until those things show up in results. Our data has no corner or card counts, so those two pages list games by profile and say so. Treat every Mind Score as a well-informed probability – never a certainty.