Four tracks, one discipline
Start with prices, move to evidence, then to risk, and finish by reproducing research. Lessons build on each other, but every one stands alone.
Price Intelligence
Settlement rules, implied probability, bookmaker margin, expected value and the gap between a quoted and an executable price.
Market formats and exact settlement rules
A price only means something next to the rule that settles it. Learn how the seven Value 7 market families settle in general football markets, and where Value 7's simplified win-or-lose rules change the maths.
Lesson 2Implied probability and bookmaker margin
Convert decimal odds into implied probabilities, measure the bookmaker's margin with the book sum, and see why prices that add up to more than 100% are a built-in cost to anyone who bets.
Lesson 3Removing the margin: four de-vig methods
The overround tells you how much margin to remove but not where it sits. Compare proportional, additive, power and Shin adjustments, see where each one breaks, and learn what published comparisons found.
Lesson 4Expected value under uncertainty
Expected value is one line of arithmetic, but it is only as good as the probability you feed it. Learn to compute EV, break-even and fair odds, then test whether an apparent edge survives a realistic range of estimates.
Lesson 5Variance, correlation and sample size
Even a real edge is buried in noise. Measure the variance of a single bet, the standard error of a track record and the number of bets needed to see an edge, and learn why picks from the same match are not independent evidence.
Lesson 6Quoted prices, executable prices and settled returns
A price on a screen is not a price you can necessarily get. Separate forecasts, quotes, executable prices and settled returns, and measure how slippage, partial fills, stale quotes and limits erode an edge on paper.
Finding and Validating an Edge
Consensus prices, market equivalence, closing line value, football score models, pricing biases and honest backtesting.
Building a consensus fair price
Pool several bookmakers into one fair-price benchmark: remove each book's margin, average probabilities rather than odds, and avoid counting the same opinion twice.
Lesson 8Line shopping and exact market equivalence
A better price only counts if it is for exactly the same contract. Learn to check period, settlement rule, line and voids before comparing prices — and why Value 7's no-push rule changes the answer.
Lesson 9Closing line value: a diagnostic, not a verdict
Closing line value compares the price you took with a stated de-vigged closing price. It is a practitioner diagnostic, not realised ROI; the sources reviewed here do not establish long-run profit from it.
Lesson 10Football score models: Poisson and Dixon–Coles
Build a score matrix from team goal rates, correct it for low scores with the Dixon–Coles adjustment, and turn it into consistent fair prices for match result, BTTS and totals.
Lesson 11The favourite–longshot bias: test it, don't assume it
Long shots have often returned less per unit staked than favourites, but whether that holds depends on market structure, the test you use and the period. The lesson is to test it in your exact market — not to always back favourites.
Lesson 12Backtesting without fooling yourself
Most impressive betting backtests fail for mundane reasons: data that were not available at the time, duplicated rows, too many tries on the same history. Learn point-in-time discipline, multiple-testing corrections and walk-forward validation.
Risk and Bankroll Science
Kelly and fractional Kelly, uncertainty-aware decisions, drawdowns, correlated exposure, execution costs and model calibration.
Kelly and fractional Kelly
Derive the Kelly stake from expected log growth, see why staking about twice Kelly erases long-run growth even on a positive-EV bet, and weigh what fractional Kelly gives up and gains.
Lesson 14Probability uncertainty and robust decisions
Treat a probability as an interval rather than a point: test whether the edge survives the downside, size stakes from the lower bound, and see why estimation error calls for shrinking Kelly.
Lesson 15Drawdowns and risk of ruin
Measure drawdowns properly, read their distribution from simulation, and set stop rules and affordability limits — then see how a five-point overestimate turns Kelly staking into ruin.
Lesson 16Portfolio thinking: correlation and exposure
Picks that share a match, team, league, model or kickoff are not independent. Measure their correlation from a score model, avoid multiplying probabilities for same-game combinations, and cap exposure by what picks have in common.
Lesson 17Liquidity, execution, limits and costs
Turn a theoretical edge into a realised one by accounting for partial fills, slippage, latency, commission and stake limits — and see why hypothetical backtest profits overstate what real markets deliver.
Lesson 18Calibration, scoring rules and model drift
Separate calibration, discrimination and accuracy; score forecasts with Brier and log loss; read reliability diagrams and calibration by odds bucket; and monitor models for drift against a benchmark.
Professional Research Lab
Reproduce published studies, catch false positives, weigh conflicting evidence and write a reproducible research report.
Reproduce a Dixon–Coles case study
Rebuild the Dixon–Coles workflow step by step — team strengths, the low-score τ correction, time-decay weighting and a chronological backtest — then compare its probabilities with a de-vigged hypothetical market.
Lesson 20Replicate a cross-bookmaker discrepancy study
Rebuild a consensus-versus-outlier betting rule in the style of Kaunitz, Zhong & Kreiner (preprint), then add quote timestamps, stake caps, account restrictions, fill rates and costs to see how a nominal edge erodes.
Lesson 21Anatomy of a false-positive backtest
Take a synthetic backtest showing +161.9% ROI apart flaw by flaw — an erroneous outlier price, look-ahead quotes, a duplicate row and an unexecutable stake — and see why a published correction found the same pattern in real research.
Lesson 22Persistent bias or single-season noise?
Weigh claims of persistent pricing bias against multi-season evidence: why significant single seasons are expected even in efficient markets, how theory and competition evidence fit together, and how to pre-register a cross-season test.
Lesson 23Audit a machine-learning sports forecast
Audit a forecasting model the way a reviewer would: base rates, calibration by bucket, Brier and log loss against a market baseline, leakage checks, multiple model comparisons and drift — then read Walsh & Joshi (2024) with its corrigendum.
Lesson 24Write a reproducible research report
Turn an analysis into an auditable memo: question, null hypothesis, data lineage, point-in-time joins, pre-registered rules, confidence bounds, execution assumptions, failure cases, conflicting evidence and an explicit decision — including "no edge demonstrated".