Extra Time and Penalties: What Your Bet Actually Covers

The Laws of the Game have already answered this

Three short passages of the 2026/27 Laws settle the whole question, and almost nobody writing about cup betting quotes them.

• Law 7.1 defines the match. "A match lasts for two equal halves of 45 minutes." That is the match. Extra time is not part of that definition.
• Law 10.2 defines the result. "The team scoring the greater number of goals is the winner. If both teams score no goals or an equal number of goals, the match is drawn." A cup tie level after ninety minutes is a draw — not a provisional draw, not an unfinished match. The Law then adds that "when competition rules require a winning team after a drawn match or home-and-away tie", only three procedures are permitted: the away goals rule, two equal periods of extra time of no more than 15 minutes each, and penalties. They are procedures applied to a match that is already drawn.
• Law 10.3 puts the shoot-out outside the match altogether. "Penalties (penalty shoot-out) are taken after the match has ended." Six words do all the work. A shoot-out cannot change a score that is already final, which is why a 1-1 draw settled by penalties is still 1-1 for every market that counts goals.

So the ninety-minute convention in betting is not a bookmaker preference. It is what the sport's own rulebook says a match is. The useful consequence is that you do not have to guess: you have to know which of the two questions a market is asking — who won the match, or who is still in the competition.

Which markets stop at ninety minutes, and which run to the end

The dividing line is whether the market is about the match or about the tie. Markets about the match stop when the match does. Markets about the tie run until someone is knocked out.

Two consequences are worth spelling out because they cost people money every season. First, a shoot-out produces no goals for betting purposes, so a match that ends 1-1 and is decided 5-4 on penalties settles as 1-1: under 2.5 goals is a winner, and a 5-4 correct score bet is not. Second, if a competition does use extra time, goals scored in it count for markets that explicitly cover extra time and for nothing else — which is why the same shot can win a to-qualify bet and lose a goalscorer bet struck on the same player.

The word doing the work above is "normally". These are contract terms set by each operator in its own betting rules, and Law 10.2 leaves the choice of procedure to each competition, so the away goals rule, extra time and penalties are combined differently in different cups and in different rounds of the same cup. Before you stake on a knockout tie, find two things: what the competition does with a drawn match, and what the market you are about to click says it settles on. Neither takes a minute, and the second is the one people skip. The same discipline applies to markets that reprice while a tie is running — see In-Play Betting Explained for how suspensions and the bet delay behave once extra time begins.

How often the rule actually bites

"Occasionally" is the usual answer, and it is wrong. If both sides' goals arrive independently at a constant rate — the standard first approximation for a football match — the chance of a level score after ninety minutes falls out of the expected total, and it is large.

Roughly one evenly matched knockout tie in four is still level when the referee ends the match. In those ties the qualifying question and the match question have completely different answers, and the gap between the two fair prices is about a third. That is not a rounding error; it is larger than the margin built into most football markets.

The one line that links the two markets

Take a tie level on aggregate going into the deciding match. A team qualifies if it wins inside ninety minutes, or if the match is drawn and it then wins whatever procedure the competition uses. Write p for the chance it wins in ninety minutes, x for the chance of a draw, and d for its share of the ties decided after the match. Then its chance of going through is exactly p + d·x. Set d to one half — the natural benchmark, since both sides start the extra period level — and translate into prices:

Everything you need is on one screen, in the market you are already looking at. Take an illustrative match line of 2.20 for the home side, 3.40 the draw, 3.30 the away side. Then 1/2.20 = 0.4545 and 1/6.80 = 0.1471, which add to 0.6016, so the benchmark home qualify price is 1.66. The same sum for the away side gives 2.22. Note what the ninety-minute rule has done: the home team's match price is 32% longer than its price to go through, and both refer to the same afternoon.

There is a second use for the same line, and it is the more valuable one. The two benchmark prices carry exactly the margin that was already in the match odds — in the example, 0.6016 and 0.4501 add to 1.0517, which is the 5.17% overround of the 1X2 market they were derived from. So you can check a qualification market for extra margin without knowing anything about the teams: work out both benchmark prices, add the two implied probabilities, and compare that total with the overround of the match odds. If the qualification market implies materially more, that market is carrying margin the match market is not, and the difference is coming out of your side of the bet. Our margin calculator does both totals for you, and the odds converter will put fractional prices into decimals first.

Where the rule costs most, which is not where you would guess

The instinct is that the ninety-minute rule matters most in a mismatch, because a strong favourite has the most to lose from a fluke. The arithmetic says the opposite. Keeping the expected goals at 2.6 and varying only how one-sided the tie is:

The gap between the two markets is widest — about 36% — in a level tie, and narrows to about 12% when one side is heavily favoured. The ninety-minute rule is therefore most expensive to ignore in precisely the ties that attract the most interest, the ones where nobody can separate the sides. In a mismatch the two markets nearly converge and the distinction barely matters.

What this test cannot tell you

The benchmark splits the ties decided after the match evenly. That is the right neutral assumption and it is not the truth. A shoot-out is close to a coin toss whoever is taking the kicks, but thirty minutes of extra time favours the stronger side, so a favourite's true share is above one half and an underdog's is below it. The benchmark therefore slightly understates a favourite's chance of going through, and slightly overstates an underdog's.

"Slightly" is measurable, and it is the reason the test still works. In the level tie above, moving the share from a half to 0.55 shortens the benchmark qualify price by 2.6%; moving it to 0.60 shortens it by 5.0%. Each five points of assumed extra-time dominance is worth about two and a half per cent of the price. That means a discrepancy of two or three per cent tells you nothing, and a discrepancy of fifteen or twenty per cent cannot be explained away by any plausible view of who would win extra time.

One more limit, stated plainly: this is a test of two prices against each other, not a valuation of either. It cannot tell you that a team will go through, and finding a gap is not a reason to bet. It tells you which of the two questions you are being asked, and whether the two answers on the screen are consistent with one another. The independent-goals model behind the tables is an approximation as well: real matches show mild score dependence, particularly late on, which moves the draw figures by a point or so and none of the conclusions.

Checking a knockout tie in under a minute

1. Find what the competition does with a drawn match: away goals, extra time, penalties, or a combination. Law 10.2 permits only those three, and the competition chooses.
2. Read the market name. "Match result" and "to qualify" are different bets on the same fixture, and only one of them survives a shoot-out.
3. Take the match odds and apply 1/Q = 1/P + 1/(2D) to get the benchmark qualify price for each side.
4. Add the two implied probabilities and compare with the overround of the match odds. A materially bigger number means extra margin in the qualification market.
5. Remember that goals in a shoot-out do not exist for goal markets, and that a discrepancy under about three per cent is noise.

Knockout ties are only part of the picture in a European season that now decides most qualification in a league table instead: our Champions League 2026/27 guide sets out which places go straight through and which go to a play-off, and the Premier League 2026/27 guide tracks the domestic season week by week. If you are backing a team to lift a trophy rather than to win a tie, Outright Betting Explained covers how those markets settle, and Accumulator Betting Explained sets out what happens to a multiple when one leg is a cup tie that goes the distance.

Frequently asked questions

Sources and method

The three passages quoted above are from the Laws of the Game 2026/27, Law 7 and Law 10, published by the International Football Association Board and read at source before publication.

Every probability and price in the tables was computed from an independent Poisson model of each side's goals and recalculated independently before publication. The match line of 2.20, 3.40 and 3.30 is an illustration, not a quoted price: no live odds appear on this page, because prices change hourly and the arithmetic does not. Nothing here is a selection or a recommendation. Settlement terms are set by each operator in its own betting rules and by each competition in its own regulations, and those are the only binding sources for the bet in front of you.