Frequently Asked Questions
Clear, honest answers about our Dixon-Coles model, public ledger accountability, and subscription features.
How are football matches evaluated?
We use a Dixon-Coles bivariate Poisson scoreline model. Recency-weighted home/away scoring and conceding rates generate expected goal parameters (λ and μ), from which a 10×10 scoreline probability matrix is computed for each match across 29 watched leagues.
What do the grade signals mean?
Grade signals ("Strong signal", "Mixed signal", "Weak signal", "Not enough data") describe data completeness and statistical agreement across underlying components (Form, H2H, and Home/Away split). They do NOT imply a guaranteed winning bet or an edge over bookmaker closing lines.
What is the public ledger?
The public ledger is an append-only log of every published evaluation. Entries are timestamped before kickoff and settled automatically when final match results land. No entry is ever edited, deleted, or omitted.
Why do some matches show "Not enough data"?
If a match has fewer than 3 recent form games, fewer than 3 head-to-head meetings, or fewer than 3 home matches played in the season, our data quality auditor blocks the evaluation to prevent misleading output from insufficient sample sizes.
What does Pro include?
Pro unlocks our SportyBet booking code audit assistant (allowing you to paste booking codes for immediate leg-by-leg evaluation) and 1-click booking code generation based on active model outputs.
How does subscription billing work?
Subscriptions are billed monthly via Paystack. You can cancel at any time, and your Pro access will remain active until the end of your current monthly billing period.