Half-Time / Full-Time Calculator
Predict HT/FT outcomes using a two-stage Poisson model. Split expected goals between halves (default 44% first half / 56% second half) and calculate all 9 HT/FT combination probabilities.
Half-Time
Full-Time (2H only)
HT/FT Combination Probabilities
D/D (Draw/Draw) is among the most common HT/FT outcomes in evenly-matched games (~15-20% in this model). H/H dominates when home xG significantly exceeds away xG.
How the HT/FT Model Works
The Half-Time/Full-Time market asks you to call two results with one bet: the state of the match at the interval and the result at the final whistle. Three half-time outcomes times three full-time outcomes gives nine combinations — H/H, H/D, H/A, D/H, D/D, D/A, A/H, A/D and A/A. Because a single price has to cover two linked events, every HT/FT price is long — and long-odds side markets are exactly where bookmaker margin and lazy pricing hide.
This calculator prices the market with a two-stage Poisson model. Each team's full-match expected goals is split between the halves. Football is not symmetric: across major European leagues roughly 44% of goals arrive before the break and 56% after, as fatigue, substitutions and trailing teams chasing the game open matches up late. The split produces four half-specific lambdas, and each half gets its own Poisson score matrix, yielding home/draw/away probabilities for the first half and for the second half in isolation.
The correct framing is conditional probability: P(HT = X and FT = Y) = P(HT = X) × P(FT = Y | HT = X). The full-time result is not independent of the half-time score — a team leading at the break only needs to avoid losing the second half by two goals. This tool uses the standard fast approximation: it multiplies the half-time result probability by the second-half result probability. That is exact for some cells (level at half-time plus a drawn second half is always a full-time draw) and approximate for others (a half-time leader that draws the second half actually wins the match). An exact model would enumerate every first-half scoreline against every second-half scoreline and classify each pair by the aggregate score. The approximation is close enough to expose badly priced cells — which is the point of the exercise.
Why bother with such an awkward market? Because HT/FT is one of football's least efficient markets. The overround is spread across nine outcomes instead of three and is typically much larger than on the 1X2. Liquidity is thin, so bookmakers copy each other rather than re-price from a model. Recreational money piles onto the favourite/favourite cell, distorting the remaining eight. And because most combination probabilities are small, a modest error in the bookmaker's probability becomes a large error in odds: a cell priced at 11.0 that should be 9.0 is a 22% pricing gap of a size you will rarely find on a main line. When you spot one, quantify it with the Expected Value calculator before staking anything.
HT/FT Formulas
1st-half λ = full-match xG × s (s ≈ 0.44) 2nd-half λ = full-match xG × (1 − s) P(k goals in a half) = e^(−λ) × λᵏ / k! P(HT = X) from the 1st-half score matrix P(2H = Y) from the 2nd-half score matrix Approximation used here: P(X/Y) ≈ P(HT = X) × P(2H = Y) Exact conditional form: P(HT = X, FT = Y) = P(HT = X) × P(FT = Y | HT = X) (sum joint 1H × 2H scorelines, classify by aggregate)
Worked Examples
With a 44% split, half-time lambdas are only 0.66 and 0.48 — so the draw leads at the break (HT: Home 34.2%, Draw 42.9%, Away 22.9%). Top combinations: D/H 16.3% (fair 6.12), D/D 15.9% (fair 6.30), H/H 13.0% (fair 7.67). A hypothetical book price of 8.00 on D/H would carry a +30% model edge — worth taking only if you trust your xG inputs.
Half-time picture: Home 50.2%, Draw 36.6%, Away 13.2%. H/H comes out at 28.1% (fair 3.55) and D/H at 20.5% (fair 4.87). D/H — the slow-starting favourite — is the classic value cell: casual money crowds H/H, and books often leave D/H a point or two longer than the model justifies.
Frequently Asked Questions
What are the 9 HT/FT combinations?
HT/FT pairs the half-time result with the full-time result: Home/Home, Home/Draw, Home/Away, Draw/Home, Draw/Draw, Draw/Away, Away/Home, Away/Draw and Away/Away. You win only if both legs are correct — a team leading at half-time that ends up drawing the match is H/D, not H/H.
How is HT/FT probability calculated?
Formally it is conditional probability: P(HT and FT) = P(HT result) × P(FT result given the HT result). This calculator approximates that by splitting each team's expected goals into two Poisson halves (default 44% / 56%), computing home/draw/away probabilities for each half separately, and multiplying the half-time result probability by the second-half result probability.
Why is the first-half goal share only 44%?
Fewer goals are scored before the break across virtually every league: teams start cautiously, while fatigue, substitutions and trailing sides chasing the game all push scoring into the second half. Roughly 44% first half / 56% second half is a sensible default; lower the share slightly for defensive leagues and cagey derbies.
Why are HT/FT odds often mispriced?
It is a low-liquidity side market: the bookmaker margin is spread across nine outcomes instead of three, books tend to copy each other's prices rather than re-model them, and recreational money concentrates on the favourite/favourite cell. Because most combinations are low-probability, a small error in the bookmaker's probability estimate becomes a large error in the odds.
Is Draw/Draw really the most likely HT/FT outcome?
In evenly matched, low-scoring fixtures, yes — each half is a short, low-lambda scoring interval where draws dominate, so D/D typically lands around 15-20% in this model. Once one side's expected goals clearly exceed the other's, Home/Home overtakes it, reaching roughly 28% for a strong favourite.