This football analytics glossary covers 63 terms, updated 12 October 2026: 58 of 63 are explained in depth across 11 guides, and 5 have an explainer of their own. Each entry links straight to the section that defines it, with worked examples and the TACTIX angle on how the model uses it.
Start with how accurate TACTIX is to see calibration applied to settled predictions, the methodology for how the model is built, and the betting calculator to turn odds into implied probability.
What Is in This Football Analytics Glossary?
The terms fall into two families. The first is analytics: the numbers used to describe how teams and players perform, and how a prediction model is built and judged. The second is betting: the language of prices, markets and staking. Each guide below takes one topic from start to finish, so related ideas sit together instead of on separate thin pages.
Prediction model metrics: Brier score, log loss, calibration
How to judge a forecast: scoring rules that reward honest probabilities, and the traps that make a model look better than it is.
Expected goals metrics explained: npxG, xA, xGOT and more
The xG family beyond the headline number: chance quality without penalties, the pass before the shot, and shot placement.
- Non-penalty xG (npxG)
- Expected assists (xA)
- Expected goals on target (xGOT)
- Goals prevented (GK)
- Set-piece xG
- Shot-creating actions (SCA)
Pressing and territory stats: PPDA, field tilt and xT
Measures of how high a team presses, where the ball lives and which passes move it forward.
- PPDA (passes per defensive action)
- Field tilt
- Progressive passes
- Packing
- Expected threat (xT)
- Possession-adjusted stats
Reading football stats fairly: per-90, sample size and form
The checks to run before trusting a stat: minutes played, sample size, schedule and the pull back to average.
- Per-90 statistics
- Sample size
- Regression to the mean
- Form (recent form)
- Game state
- Strength of schedule
- Home advantage
How football prediction models work: Poisson to simulation
The building blocks of a match model, from goal distributions and ratings to simulating a season.
- Poisson distribution
- Dixon-Coles model
- Elo rating
- Monte Carlo simulation
- Expected points (xPts)
- Pythagorean expectation
- In-play win probability
Betting odds explained: implied probability, margin and value
What a price says as a probability, how the bookmaker margin is built in, and what value means.
- Implied probability & overround
- Vigorish (juice)
- Odds-on and odds-against
- Expected value (EV)
- Closing line value (CLV)
- Steam move
- Sharp and square
Goals markets explained: over/under, clean sheets, correct score
Markets settled on the number and timing of goals rather than the result.
Handicap betting explained: Asian, European and draw no bet
Markets that give one side a head start, and the ones that refund your stake if the match is drawn.
Goalscorer markets and bet builders explained
Player scoring markets and combining several outcomes from one match.
Bankroll management and staking: Kelly, plans and variance
How much to stake and why the size of each stake matters as much as the selection.
Cash out and hedging explained, plus each way and arbitrage
Settling early, laying off a position, each-way terms and why arbitrage is rarer than it looks.
Stand-alone explainers
Five ideas come up so often that they have a page of their own rather than a section in a guide. Expected goals and calibration sit underneath almost every number on the site; value, both teams to score and accumulators are the betting terms readers ask about most. They are listed here so the football analytics glossary stays complete in one place.
- Expected goals (xG) — Chance quality — the probability each shot becomes a goal.
- Model calibration — Whether a model’s probabilities are honest: when it says 60%, does it happen 60% of the time?
- Value / edge — The gap between the model’s probability and the price’s implied probability.
- Both teams to score (BTTS) — Wins if both teams score at least once.
- Accumulators — Several selections combined into one bet where every leg must win.
How Should You Use a Football Analytics Glossary?
Use the football analytics glossary as a reference, not a reading list. When a match preview mentions a Brier score, PPDA or closing line value, the link takes you to the part of the guide that explains it, and the guide shows the neighbouring ideas that usually matter at the same time. A probability is only as useful as its calibration, and a price only matters next to the probability behind it, so the guides read best in that order: the model, how it is measured, then the markets. The definitions are written in plain English first and the maths second, and where a term changes how a TACTIX prediction should be read, the guide says so next to it. New terms join the guide they belong to rather than a page of their own.
See the live track record and how TACTIX’s model works.
18+. Information and analysis only — not betting advice. Please gamble responsibly — BeGambleAware.org.
This post is generated from TACTIX’s match data by our software and checked automatically. How it works →
See today’s free slate
Free, every day: the model’s fixed slate of picks, each with its confidence tier and that tier’s settled record — plus yesterday’s slate, graded. Pro adds the full board, the value edge and the written analysis. No payment card.
Create free account