Top picks
18+The FootyGrounds Top Picks: up to five higher-confidence statistical match predictions each weekend and midweek, with probabilities, confidence scores, combined risk and full reasoning. Not betting advice.
These are the model's statistically strongest predictions. They are not safe bets — even individually strong predictions fail regularly, and combining five outcomes dramatically increases the overall risk.
This midweek's higher-confidence picks
Model generated from 2 modelled fixtures in the 3 Nov 2026 – 5 Nov 2026 window; 2 met the confidence and data-quality criteria. Model version fg-match-predictor-1.0.0, last calculated 31/10/2026, 16:20:14.
Only 2 fixtures met our confidence threshold this edition. We publish fewer predictions rather than lowering the standard to manufacture five.
1. Queens Park Rangers to win
Queens Park Rangers v Burnley · EFL Championship · Tue 3 Nov, 19:45
60%
Model probability
Data: excellentConfidence: ModerateFootyGrounds confidence score64 / 100 · ModerateWhy the model selected this match
- Queens Park Rangers strength rating 1588 v Burnley 1528
- Queens Park Rangers home record this season: 3W 2D 1L
- Burnley away form: 1W 1D 4L, 9 scored / 15 conceded
- Superior goal difference: Queens Park Rangers +7 v Burnley -8
- Opponent-adjusted model probability: 60% (Elo 57%, goals 62%, recent form 60%)
Model components
- Elo strength model57%
- Goals model62%
- Recent-form model60%
Main uncertainty: No verified injury or suspension data is available for this division.
Full match analysis2. Birmingham City to win
Birmingham City v Millwall · EFL Championship · Tue 3 Nov, 20:00
56%
Model probability
Data: excellentConfidence: Low–ModerateFootyGrounds confidence score56 / 100 · Low–ModerateWhy the model selected this match
- Birmingham City strength rating 1599 v Millwall 1552
- Birmingham City home record this season: 2W 4D 0L
- Millwall away form: 2W 1D 3L, 8 scored / 12 conceded
- Superior goal difference: Birmingham City +10 v Millwall +2
- Birmingham City have won 4 of their last 5
- Opponent-adjusted model probability: 56% (Elo 55%, goals 56%, recent form 61%)
Model components
- Elo strength model55%
- Goals model56%
- Recent-form model61%
Main uncertainty: No verified injury or suspension data is available for this division.
Full match analysis
Chance of all 2 winning
33.6%
Chance of at least one selection failing
66.4%
Individual probabilities: 60%, 56%. Even when individual selections have relatively high model probabilities, combining 2 outcomes makes an accumulator substantially less likely to succeed.
The combined figure multiplies the individual probabilities together, a simplified independence assumption. Real matches are not perfectly independent, so treat it as an approximation rather than an exact number.
