Betting 101

Multi goals betting explained: how to price a 2-3 goals band

By FootInsights · Published 22 Sep 2026 · 6 min read

Most goals markets ask you a yes-or-no question. Multi goals betting asks a different one: not “will there be more than 2.5 goals?” but “which band will the final total land in?” — 1-2, 2-3, 2-4, 1-5 and so on. The market looks friendlier than Over/Under because a band feels like a wider target, and the prices look generous next to a half-goal line. Both impressions are produced by the same thing, and it is not generosity. Here is how a band is actually priced, and the two arithmetic traps that make it hard to tell a fair band from an expensive one.

What multi goals betting settles on

A multi goals selection wins if the total goals in the match — both teams, 90 minutes plus stoppage time, own goals included, extra time and penalties excluded — falls inside the quoted range, endpoints included. A bet on 2-3 goals wins on 1-1, 2-0, 2-1 and 3-0. It loses on 1-0 and on 2-2.

That is the whole rule. Everything else about this market is the price, and the price comes from a distribution.

Every band is just a sum of cells

A goal model does not produce a band. It produces a probability for each possible match total, and every goals market — half-goal lines, Asian totals, multi goals — is a different way of slicing that same object. Take a hypothetical match where a model expects 2.7 goals and treats totals as roughly Poisson-shaped:

Pricing any band is now addition. Add the cells the band covers, then invert. A 2-3 band is 24.5% + 22.0% = 46.5%, so its fair price is 1 ÷ 0.465 = 2.15.

Run that across the board a bookmaker typically offers:

Two things fall out of that table immediately. The first is that there is no “best band” in the abstract — a wider band is not a better bet, it is simply more probability bought at a proportionally lower price, and the ranking of bands by value depends entirely on the prices offered against them. The second is the answer to why 2-3 is the market’s most popular selection: at this expected total it is the single most likely two-goal window, covering the two most common match totals at once.

The trap: you cannot build a band by multiplying two lines

The obvious way to check a band price is to rebuild it from markets you trust. A 2-3 band is exactly “Over 1.5 and Under 3.5” — the intersection of two half-goal lines, with no gap and no overlap. So multiply them.

From the distribution above, Over 1.5 wins 75.1% of the time and Under 3.5 wins 71.4% of the time. Multiply:

0.751 x 0.714 = 0.537  ->  fair odds 1.86

The true answer is 46.5% and 2.15. The shortcut overstates the win probability by more than 7 percentage points and understates the fair price by 15%.

It fails because the two legs are not independent — they are strongly negatively correlated. A match that comfortably clears Over 1.5 is, by that very fact, a higher-scoring match and therefore less likely to stay Under 3.5. Multiplying treats each leg as if the other had told you nothing, when in reality each one is evidence against the other. This is the same correlation that makes same-match legs misbehave in a bet builder, and it runs in the direction that matters most: anyone benchmarking bands this way will conclude that every band on the board is a rip-off, because their reference price is 15% too low before the bookmaker has taken anything at all.

There is no shortcut. To price a band you need the distribution.

Where the margin actually hides in multi goals betting

Now the part the market guides skip. Compare two selections drawn from the same hypothetical distribution.

A 2-3 band quoted at 1.95 implies 51.3% against a fair 46.5% — the price is shaded by about 10% in probability terms. An Over 2.5 line quoted at 1.91 implies 52.4% against a fair 50.6% — shaded by about 3.4%.

Same match, same model, same underlying goal distribution. Roughly three times the margin on the banded version.

That gap is structural, not a quirk of these particular numbers. Half-goal totals are the most heavily traded football market in existence, and competition compresses the margin on them. Multi goals is a derived, lower-liquidity market with six or eight selections rather than two, which gives a book more places to hold margin and far less pressure to price any single band sharply. The prices look bigger because the events are less likely; the value is smaller because the fee is larger.

The practical consequence is worth stating plainly: if you can express your opinion on a half-goal line, that is usually the cheaper way to buy it. The band is the convenient version, and convenience is priced.

The 0-0 blind spot

One more feature of these boards deserves attention, because it is where “covering yourself” quietly fails.

Most multi goals boards begin at 1. A bettor spreading across 1-2, 1-3 and 1-4 feels well protected — three selections, three chances. Against a goalless draw, all three lose together. At the expected total used above, a 0-0 arrives in about one match in fifteen, and in a low-scoring fixture — a relegation six-pointer, a cautious away leg — that share climbs steeply.

The bands are not independent bets. They are nested slices of one distribution, and every band that starts at 1 shares the same dead zone. If the goalless outcome is the risk you actually care about, the band market does not address it; a market that includes zero does.

What this market is good for

Multi goals is a legitimate way to express a shaped view. If your reading of a match is “goals, but not a shootout”, no half-goal line says that in a single bet — Over 1.5 does not cap the top end and Under 3.5 does not set a floor. A 2-3 band does both. That specificity is real, and it is the market’s only genuine advantage.

Whether it is worth the extra margin comes down to a comparison you now have the arithmetic to make: build your own distribution over match totals, sum the cells the band covers, invert, and put your number next to the board’s. If the shading is 10% and the same view costs 3% elsewhere, the answer is usually elsewhere.

None of which matters if the distribution itself is wrong. The arithmetic here is exact — it converts probabilities into prices faithfully and has no opinion on whether those probabilities are any good. That question is settled only by scoring forecasts against results across a long run of matches and publishing what comes back, which is what a public track record exists to do. A precise price built on a poorly calibrated model is just a confident mistake.