How is Over 2.5 probability calculated?
We sum joint Poisson probabilities for all scores with i+j ≥ 3 (strictly greater than 2.5).
Over/Under Probability Calculator
Goal totals (Over 2.5, Under 3.5, team goals) from expected goals — Poisson extension for football betting.
Quick answer
Enter xG → see Over/Under probabilities for 0.5–4.5 goals plus team totals. Half-lines (.5) never push.
Over/Under goal markets explained
Totals betting asks whether combined goals go over or under a line (e.g. Over 2.5 means 3+ goals). Half-lines (.5) have no push; integer lines can push in some markets.
Enter home and away xG to see Poisson-derived probabilities and fair odds for popular lines — compare with bookmakers to spot value.
Educational probabilities only — not betting advice. Gamble responsibly where legal.
Match total goals — Over / Under
| Line | Over | Under |
|---|---|---|
| 0.5 | 93.28% | 6.72% |
| 1 | 93.28% | 6.72% |
| 1.5 | 75.13% | 24.87% |
| 2 | 75.13% | 24.87% |
| 2.5 | 50.64% | 49.36% |
| 3 | 50.64% | 49.36% |
| 3.5 | 28.59% | 71.41% |
| 4 | 28.59% | 71.41% |
| 4.5 | 13.71% | 86.29% |
Worked example (current xG inputs)
Values update when you change home or away expected goals above.
Home xG: 1.5 · Away xG: 1.2
| Market / line | Model probability | Fair decimal odds |
|---|---|---|
| Over 1.5 | 75.13% | 1.33 |
| Under 1.5 | 24.87% | 4.02 |
| Over 2.5 | 50.64% | 1.97 |
| Under 2.5 | 49.36% | 2.03 |
| Over 3.5 | 28.59% | 3.5 |
| Under 3.5 | 71.41% | 1.4 |
Home team goals
| Over 0.5 | 77.69% | 22.31% |
| Over 1 | 77.69% | 22.31% |
| Over 1.5 | 44.22% | 55.78% |
| Over 2 | 44.22% | 55.78% |
| Over 2.5 | 19.12% | 80.88% |
| Over 3 | 19.12% | 80.88% |
| Over 3.5 | 6.56% | 93.44% |
| Over 4 | 6.56% | 93.44% |
| Over 4.5 | 1.86% | 98.14% |
Away team goals
| Over 0.5 | 69.88% | 30.12% |
| Over 1 | 69.88% | 30.12% |
| Over 1.5 | 33.74% | 66.26% |
| Over 2 | 33.74% | 66.26% |
| Over 2.5 | 12.05% | 87.95% |
| Over 3 | 12.05% | 87.95% |
| Over 3.5 | 3.38% | 96.62% |
| Over 4 | 3.38% | 96.62% |
| Over 4.5 | 0.77% | 99.23% |
Poisson Model for Goal Totals
Total goals G = home goals + away goals. Under independent Poissons, we sum P(i,j) over the score grid where i+j meets the Over/Under rule (e.g. Over 2.5 → 3+ goals).
Example: Over 2.5 with λ 1.5 + 1.2
Combined λ ≈ 2.7 goals expected. Over 2.5 probability typically lands near mid-50% to 60% depending on split — use the table for the exact Poisson sum. Under 2.5 is the complement.
Over 2.5 vs BTTS vs Asian Totals
Over 2.5 is about total goals only. BTTS requires both teams to score. Asian goal lines may quarter-split like handicaps — see our Asian Handicap calculator for handicap-style markets.
How to Use the Over/Under Calculator
Enter xG, review match total lines, then home/away individual goal markets.
- Enter expected goals: λ_home + λ_away drives the total goals distribution.
- Pick your line: Half-lines (.5) have no push; integer lines can push on exact totals in some books.
- Switch probability or odds: Fair odds = 1 / model probability for quick comparison.
- Cross-check BTTS and 1X2: High total xG often lifts Over 2.5 but BTTS still needs both sides to score.
Over/Under Tips
- Over 1.5 first-half markets need HT xG — this tool uses full-match λ unless you adjust inputs.
- Weather and red cards shift live totals — pre-match xG is a baseline only.
- Shop lines: 2.5 vs 2.75 Asian totals differ — know your book rules.
- Compare multiple lines (2.5, 3.5) to see where the model disagrees most with the market.
What's Next?
See BTTS calculator, Poisson markets, and football predictions.
Glossary: totals betting
- Over 2.5
- Wins when total goals are 3 or more.
- Under 3.5
- Wins on 0–3 goals; loses on 4+.
- xG
- Expected goals per team; drives Poisson λ.
- Push
- Stake returned on exact integer totals at some books.
- Fair odds
- 1 ÷ model probability for the selected line.
- Closing line
- Final pre-kickoff total — benchmark for totals +EV on OddsGPT.
- Asian total (quarter line)
- Splits stake across two half-lines (e.g. 2.75 = half on 2.5, half on 3).
- Expected value (EV)
- Weighted profit if model % beats implied odds after vig removal.
Complete guide: Over/Under goal probability from xG
Totals markets (Over/Under 2.5, 3.5, team goals) price the distribution of combined scoring. Poisson goal models are industry-standard because they convert shot quality metrics into probabilities for each goal line.
This page sums the joint score matrix for each line — half-goals (.5) avoid pushes while integer lines may refund on exact totals depending on the book. Always confirm house rules for Asian vs European totals.
Workflow: (1) set xG, (2) read Over/Under %, (3) compare to closing totals, (4) strip vig, (5) calculate EV. Pair with BTTS when analyzing attack-vs-defence matchups.
Team total tables isolate home or away scoring — useful when books offer independent team goals rather than combined match totals.
Advanced users export the same λ into the Poisson calculator for correct-score context or Poisson Pro when draws cluster at 0-0 and 1-1.
Why use this Over/Under calculator?
Full match and team total lines from one xG pair — same Poisson grid as BTTS and correct score. Toggle fair odds, compare to closing totals, and pair with BTTS for attack/defence context.
BTTS vs Over/Under vs Asian Handicap vs 1X2
Same xG inputs, different settlement rules — use the right OddsGPT tool for each market.
| Market | Wins when… | OddsGPT tool |
|---|---|---|
| BTTS (Both Teams To Score) | Each team scores ≥1 goal (Yes) or at least one blank (No). | BTTS probability calculator |
| Over/Under (e.g. 2.5) | Total goals above/below the line (3-0 → Over 2.5). | Over/Under goal probability |
| Asian Handicap | Home team covers the goal handicap (quarter lines split). | Asian handicap cover % |
| 1X2 (Match result) | Home win, draw, or away win after 90 minutes. | xG win probability |
OddsGPT football & betting tools
- All betting calculators hub
- Poisson match calculator (full markets)
- Poisson Pro — Dixon–Coles & ZIP
- BTTS probability from xG
- Over/Under goal probability
- Asian Handicap cover calculator
- xG match simulator
- Expected Value (EV) calculator
- Implied probability converter
- No-vig fair odds tool
- Closing odds database
- Football match predictions
- Football xG statistics
- Betting help & glossary
Educational tool only. Gamble responsibly. BeGambleAware · NCPG
Frequently Asked Questions
What does Under 3.5 mean?
Under 3.5 wins on 0–3 total goals; four or more loses.
Why use xG for totals?
xG stabilizes shot quality into expected scoring rates, smoother than raw past goals for forecasting.
Are team total goals independent?
We model each team with its own Poisson λ for home/away team total tables.
Half-lines vs integer lines?
.5 lines avoid pushes; integer totals (e.g. Over 3) may push on exactly 3 goals at some books.
How do I find value on totals?
Compare model % to implied odds; use closing line history on OddsGPT.
Does this include extra time?
No — standard 90-minute match assumption like most pre-match totals markets.
Link to Poisson correct scores?
Yes — the same grid powers correct-score and BTTS on our Poisson calculator.
Over 1.5 first half?
This page uses full-match λ. Halve or model HT xG separately for first-half totals.
Asian total 2.75 vs European 2.5?
Quarter totals split stakes like Asian handicaps; European 2.5 is a single half-line.
What is the break-even probability for -110 totals?
At -110 each side needs 52.4% win rate to break even before commission — compare your model % accordingly.
How do weather and cards affect totals?
Rain and red cards shift live xG upward or downward — pre-match λ is a baseline; adjust manually for forecasts.
What is a good Over 2.5 probability to bet?
Compare model % to book implied % after vig — e.g. 58% model vs 52.4% break-even at -110 suggests edge before commission.
How does Over 2.5 differ from BTTS Yes?
Over 2.5 needs 3+ total goals (3-0 counts); BTTS Yes needs both teams to score (2-0 is Over but BTTS No). Use both calculators on OddsGPT.
Can I model first-half totals here?
This page uses full-match λ. For HT Over 1.5, halve xG or use dedicated half-time models — do not reuse full-match λ without adjustment.
