Guides
The mathematics we run our own engine on, written out in full — including the bugs we shipped, the numbers we do not yet publish, and the reasons why.
Start here
- Are AI betting predictions accurate? — AI prediction sites advertise 78–87% accuracy.
- How to evaluate a tipster or prediction service — Nine checks that separate a verifiable record from a marketing claim: sample size, average odds, drawdown, statistic resets, verification method and who controls the numbers.
- De-vigging odds: how to remove bookmaker margin — Bookmaker odds contain margin, so implied probabilities sum to more than 100%.
All guides
| Guide | What it answers |
|---|---|
| Are AI betting predictions accurate? | AI prediction sites advertise 78–87% accuracy. |
| How to evaluate a tipster or prediction service | Nine checks that separate a verifiable record from a marketing claim: sample size, average odds, drawdown, statistic resets, verification method and who controls the numbers. |
| De-vigging odds: how to remove bookmaker margin | Bookmaker odds contain margin, so implied probabilities sum to more than 100%. |
| Closing line value: the metric that survives small samples | CLV compares the price you took against the price at kickoff. |
| The Poisson model for football predictions | A Poisson model turns two expected-goals numbers into a full matrix of scorelines, and that matrix prices every market at once — 1X2, over/under, both teams to score, correct score. |
| Dixon–Coles: the correction that fixes low-score draws | Independent Poisson under-predicts 0–0 and 1–1 and over-predicts 1–0 and 0–1. |
| Expected value, done properly | The usual EV formula assumes a bet either wins or loses. |
| How to measure whether a prediction model is any good | Hit rate cannot distinguish a good model from a cautious one. |
| How to build a football betting model | A complete, honest path: data collection, fitting attack and defence strengths, building the score matrix, de-vigging the market for comparison, and backtesting without the four biases that make every naive backtest look profitable. |
| Maximum drawdown: the number almost nobody publishes | Drawdown is the peak-to-trough fall in your bankroll. |
| Kelly criterion and risk of ruin | Kelly gives the mathematically optimal stake for long-run growth — and full Kelly is far too aggressive for anyone whose edge estimate might be wrong. |
| Why most events get a PASS | CONSENSUS runs a multi-layer analytical process over the same event and publishes only where the independent checks agree. |
| How to read betting odds | Three formats say the same thing in different words. |
| Line shopping: the cheapest edge available to anyone | The same outcome carries different prices at different bookmakers. |
| Asian handicap: what the quarter lines actually do | Asian handicap removes the draw and splits stakes across two lines at quarter increments. |
| Double chance and draw no bet | Two ways to shorten the odds by covering more outcomes, and one arithmetic trap that produces edges of 40 percentage points where none exist. |
| Parlays and accumulators: the margin compounds too | Combining selections multiplies the odds — and multiplies the bookmaker margin with them. |
| When not to bet | Most published analysis is about which bet to make. |
| Thin markets: when a price is one firm's opinion | A quote available at a single bookmaker carries none of the information that makes market prices useful. |
| Bankroll management, without the folklore | A separated bankroll, a flat unit, and a stake small enough to survive the worst run your strategy can produce. |
| Tracking your bets: the eleven fields that matter | A betting record is only useful if it can answer questions you did not think to ask when you made the bet. |
| Why your ROI is probably wrong | Six ways a return-on-investment figure overstates performance, all of them accidental and all of them common: turnover versus stake, excluded voids, forgotten bonus money, survivorship in the record, unequal stakes counted equally, and the sample being far too small to mean anything. |
| How many bets until you know whether you are any good | The uncomfortable answer is more than most people ever place. |
| When betting stops being a choice | The recognised warning signs, the tools that work regardless of willpower, and where to get help in the UK, US, Canada, Australia and across Europe. |
| Over/under totals: the market that prices the game, not the winner | Totals ask how much scoring there will be, not who wins. |
| Both teams to score: a bet on two independent failures | BTTS asks whether each side avoids a blank. |
| Player props: where the market is thinnest | Player markets carry the widest margins and the least scrutiny of anything on the board. |
| Live betting: what you pay for immediacy | In-play markets carry higher margins, shorter acceptance windows and a structural information disadvantage against the bookmaker. |
| How bookmakers actually set odds | A price is a model output, a margin, and a running adjustment for the money arriving. |
| Why odds move — and what a move actually tells you | Four distinct causes produce identical-looking price movements. |
| The favourite–longshot bias | Across a century of data and most betting markets, longshots are systematically overbet and favourites systematically underbet. |
| Value betting: the only definition that means anything | A value bet is one where your probability estimate exceeds the de-vigged market probability by more than your own uncertainty. |
| Arbitrage betting: the arithmetic works, the accounts do not | Backing every outcome across different bookmakers can lock a small profit regardless of result. |
| Matched betting: extracting bonuses, and the ceiling on it | Using a bookmaker bonus and covering the position at an exchange converts most of the bonus into cash. |
| Why accounts get restricted — and what actually triggers it | Winning is not the only trigger, and often not the first one. |
| Is sports betting profitable — and for whom? | For the large majority it is not, and the reason is structural rather than a matter of skill. |
Glossary
Short definitions with formulas and the mistake each term is usually involved in — 18 terms, one page each.
Check any of this against our record
Every signal CONSENSUS publishes carries the bookmaker odds fixed before the event starts and the settled result afterwards — including the drawdowns and the losing runs. The running total is on the front page and every entry is in the log.