Frequently asked questions about football data analysis

What is 11Stat?

11Stat is a football data-analytics platform that breaks matches down into probability and statistics — it is not a betting service. For each fixture it produces a measurable readout rather than a single tip: 1X2 win probability, expected goals (xG), a form index and model agreement.

• Multi-engine model — a Poisson/Dixon-Coles goal engine, an Elo/form secondary engine and an advanced bivariate engine run together.
• Monte Carlo simulation — each match is simulated 10,000 times to derive the score distribution and market probabilities.
• Calibration — probabilities are adjusted per league and market so that "70%" has historically meant roughly 70%.

Outputs are probabilities, not advice, and no result or profit is guaranteed. You can see how the numbers are produced on our model calibration page.

Are today's match predictions reliable?

A one-line "prediction" is not reliable; what is reliable is a calibrated probability readout. 11Stat never says "this team wins" — it gives a percentage split across home, draw and away, and transparently measures how well those percentages have held up.

• Reliability = calibration — roughly 60% of the matches we call "60%" should land that way, checked against paper backtests.
• Uncertainty flags — thin data, missing squads or high-variance fixtures are flagged explicitly.
• Model agreement (N/M) — how many independent models point the same way; low agreement is a caution signal.

High probability is not certainty, and no readout guarantees an outcome. Accuracy figures shown are simulated past performance and never imply a future result.

Who wins this match?

No model can know a match winner for certain — the only honest answer is a probability distribution. A sample match card might read "Home 49% · Draw 27% · Away 24%": the home side is favoured, but the outcome is open.

• 1X2 probability split — the full percentage table of three outcomes, not a single choice.
• xG edge — a measure of real superiority independent of the scoreline.
• Last 5 / last 10 form — the direction and stability of momentum.

You can inspect a match's full probability breakdown on the match analysis page. Being the favourite does not mean winning; a probability is not a promise.

Is there a lock or guaranteed match?

No — there is no such thing as a guaranteed or "lock" match, and 11Stat makes no such claim. Football is uncertain by nature; even the highest-probability readouts can go the other way.

• An 80% probability fails roughly one time in five over the long run — that is the nature of probability, not an error.
• Markets are efficient — most public information is already priced in, so a guaranteed win is mathematically impossible.
• Transparency — we measure uncertainty and show it rather than hiding it.

11Stat is not a betting-advice service and guarantees no outcome. For anyone searching for a "lock", the most honest value is a probability showing how open or one-sided a match really is.

How is a match prediction actually made with data?

Sound analysis doesn't guess; it breaks each match into measurable inputs. 11Stat follows these steps:

• Expected goals (xG) — the chance quality each team creates and concedes is estimated.
• Poisson / Dixon-Coles — a score-probability matrix is built from xG; Dixon-Coles corrects low-score dependence.
• Monte Carlo 10,000 — the match is simulated tens of thousands of times to derive 1X2, both-teams-to-score and over/under probabilities.
• Elo/form and model agreement — secondary engines corroborate the result, reported as N/M agreement.

Method details are on our methodology page. This is not betting advice; the goal is to ground your decision in transparent numbers.

Can AI predict football matches?

"AI" is not a magic oracle; match analysis is statistical models trained on historical data producing probabilities. Instead of a single "AI" box, 11Stat uses several statistical engines that cross-check each other.

• Poisson/Dixon-Coles models goal production; Elo/form tracks strength and momentum; an advanced engine captures the dependence between the two sides' scoring.
• The models are combined and reported as N/M model agreement — sturdier than trusting one algorithm blindly.
• Outputs are calibrated and presented with their uncertainty.

Even the most advanced model cannot know the future; the output is a probability, not a certainty. AI or statistics, no result is guaranteed.

Which is the most reliable match-analysis site?

A source's reliability is measured by transparency and calibration, not by loud "hit-rate" claims. When you evaluate one, look for:

• Does it show probability? — a percentage 1X2 split and xG, not a bare pick.
• Is calibration open? — has "70% means ~70%" actually been measured on history?
• Does it state uncertainty? — are thin data and high variance flagged?
• Does it use guarantee language? — a source promising "lock/guaranteed" is exactly the one not to trust.

11Stat is built around these criteria: probability not advice, calibration not claims. Even so, no platform can guarantee an outcome.

How is both-teams-to-score (BTTS) analysed?

"Both teams to score" is not a certainty but a computable probability. 11Stat derives it from the score matrix:

• A full score-probability matrix is built from each team's xG via Poisson/Dixon-Coles.
• The sum of the cells where both teams score ≥1 gives the BTTS probability.
• Monte Carlo simulation corroborates it, calibrated with attacking/defensive form and the league base rate.

For example, a "BTTS 58%" readout also means roughly one match in two where only one team scores — high probability is not certainty.

See the full probability breakdown of today's matches in the live match analysis and start your free 3-day trial. 11Stat is a football data-analytics platform, not betting advice.