1X2: The Match Result Market

Every football bettor starts here. It is the market on the front of the coupon, the one the odds in a news article refer to, and the reference point every other market is derived from. It is also, for exactly those reasons, the worst place to look for a mispriced number.

What 1 / X / 2 actually settle on#

• 1 — the home team leads when the referee ends the match.
• X — the score is level.
• 2 — the away team leads.

The window is 90 minutes plus stoppage time. Extra time and penalty shoot-outs do not count. A cup semi-final that ends 1-1 after 90, goes to extra time, and is won 3-1 settles as a draw on 1X2. This trips people up constantly, and it is not a quirk — nearly every standard football market (totals, both teams to score, handicaps) uses the same 90-minute window. If a bet is meant to include extra time it will say so in the market name: "to qualify", "to lift the trophy", "to reach the final".

On a neutral ground there is still a nominal home team, and 1X2 follows the fixture listing, not the stadium.

Why the draw makes football different#

Most big betting sports have no draw. Basketball, tennis, most American sports, and almost every knockout format resolve to a binary. Football does not, and the draw is roughly a quarter of all results across major leagues.

That third outcome does two things.

First, it makes the market genuinely three-dimensional. A model that gets both teams' strength right can still be badly wrong about how often they cancel each other out.

Second, and more awkwardly, the draw is the hardest of the three to price. In a lopsided match it is not the most likely outcome. In an even match it is still usually only 26–32% — high enough to matter, never high enough to be the favourite. So the draw price is almost never anchored by confident opinion the way a strong favourite's price is. It gets set mostly by the model and moved mostly by hedging flow, and the public rarely backs it, which is precisely why it is not soft: nobody is pushing it away from fair.

This is also why so many derived markets exist purely to deal with the draw. Double chance lets you cover it, and draw no bet lets you have it refunded. Both are just 1X2 with the draw handled differently, and both are priced from the same three numbers.

The three prices sum to more than 100%#

Convert each decimal price to its implied probability with 1 ÷ odds, then add them up. Take a realistic line:

47.6 + 29.4 + 27.8 = 104.8%. The 4.8% above 100 is the overround — the bookmaker's built-in cut. It is not a fee you pay separately; it is baked into all three prices at once, which is why you cannot see it by looking at any one of them.

To recover what the market actually thinks, you strip that margin out. The simple version scales each implied probability down proportionally:

Outcome	Odds	Implied	Fair (margin removed)	Fair odds
Home	2.10	47.6%	45.4%	2.20
Draw	3.40	29.4%	28.1%	3.56
Away	3.60	27.8%	26.5%	3.77

Proportional scaling is the easy method, not the best one. It removes the same relative amount from a 1.20 favourite and a 15.00 outsider, and real bookmaker margin is not distributed that way — longshots carry more of it. Tofiko removes margin with the power method instead, which fits an exponent so the probabilities sum to one while loading more of the margin onto the long prices. The fair odds guide works through why that matters, and the fair odds calculator will do it on any three prices you paste in.

How a goal model derives 1X2#

Tofiko does not estimate the three 1X2 numbers directly. Nothing sensible does. Instead the goal-based models estimate expected goals for each side, build the full scoreline grid from those two numbers, and then add up cells.

The grid is a table of every plausible scoreline — 0-0, 1-0, 0-1, 2-1, and so on out to about 8-8 — each with a probability attached. Then:

• P(home win) = sum of every cell where home goals > away goals
• P(draw) = sum of the diagonal: 0-0, 1-1, 2-2, 3-3…
• P(away win) = sum of every cell below the diagonal

The three sums add to 1 by construction, so a goal model never produces an incoherent 1X2 line. It also gets totals, both teams to score and correct score out of the same grid for free — one estimate, every market, all mutually consistent. That is the main argument for pricing football from goals rather than from results.

The engine underneath is a Poisson process with the Dixon-Coles correction, which nudges the low-scoring cells (0-0, 1-0, 0-1, 1-1) because raw Poisson underrates draws and tight games. The Poisson distribution guide explains the maths, and how our models work covers how the goal models get blended with an Elo strength rating into the ensemble that actually ships.

Try it yourself. Put two expected-goal numbers in and watch the grid produce a whole 1X2 line:

Change the home expected goals from 1.4 to 1.8 and watch what happens to the draw. It barely moves, while the home price moves a lot. That asymmetry is the draw's whole personality.

Why this is the hardest market to beat#

1X2 takes more money than every other football market combined. It is the market every trading desk staffs, every syndicate models, and every price-comparison site tracks. Attention is what makes a price accurate, and no football market gets more of it.

The practical consequence: the closing 1X2 price on a top-flight match is a very good probability estimate. Better, on average, than any public model. If you think you have found a 10% edge on Premier League match odds, the overwhelmingly likely explanation is that your model is wrong, not that the market is.

If you want to bet the match result but the price is unattractive, the answer is usually a different shape, not a different opinion: Asian handicap removes the draw and lets you buy or sell goals at a finer grain, draw no bet refunds it, and double chance absorbs it. All three are priced off the same 1X2 grid, so none of them creates value on its own — they only change how the same estimate is packaged.