Both teams to score (BTTS): a probability guide

Published

“Both teams to score” asks a deceptively simple question: will neither side keep a clean sheet? In our 14,772-match dataset (eight leagues, 2021–2026), the answer was yes 53.6% of the time — another near coin-flip market, which is why bookmakers love offering it.

BTTS is not Over/Under in disguise

The two markets overlap but measure different things, and conflating them is the most common BTTS mistake:

  • 3-0 is Over 2.5 but not BTTS. A dominant win with a clean sheet splits the markets.
  • 1-1 is BTTS but not Over 2.5. A cagey draw splits them the other way.

The correlation is strong, not perfect. BTTS is really a question about the weaker scoring side: it needs the underdog to find a goal. A mismatch fixture can carry a high Over probability and a mediocre BTTS one at the same time — you’ll see exactly that pattern on some of our match pages.

What drives it

  1. The weaker attack vs the stronger defence. The favourite scoring is usually near-given; BTTS lives or dies on whether the other side breaks through. Underdog attacking quality matters more here than in any other mainstream market.
  2. Defensive fragility on both sides. Two open teams push BTTS up fast; one elite defence drags it down even against good attacks.
  3. Base rates by league. Like totals, BTTS rates vary by competition — another reason our model is fitted per league rather than globally.

How we price it

Same engine as everything else: the Dixon-Coles scoreline grid. BTTS “yes” is the sum of every cell where both teams have at least one goal — equivalently, 1 minus the probability of either side blanking (counting 0-0 once). One grid, every market, no contradictions between our BTTS, totals and 1X2 numbers.

A worked intuition

Take a match where we expect the home side to score 1.8 goals and the away side 0.9. Poisson says the home team blanks about 17% of the time and the away team about 41%. Naively adding those overstates it (both can blank in the same match); the grid handles the overlap properly and lands BTTS “yes” at roughly even money — just under 50% — despite the fixture being lopsided. That counter-intuitive gap between “who wins” and “who scores” is where BTTS value hides, when it exists at all.

Judge the tips, not the vibes

BTTS content online is dominated by streak-scanning (“BTTS landed in 5 of their last 6!”). Streaks are small samples of a noisy coin — the model needs seasons, not weeks. Whatever source you use, apply the standard test: probabilities shown, every call graded, sample sizes visible. Our BTTS record sits on the track record page with the rest, wins and misses alike.