The score matrix

One table — and everything else follows from it

Every probability you see on this site — who wins, whether both teams score, whether it goes over 2.5 goals, the exact final score — comes from the same place: a small table called the score matrix. Learn to read it and you understand exactly how the model thinks. It takes three minutes, and you will never look at an odds the same way again.

What each cell means

Rows are the home goals, columns the away goals. Each cell is an exact final score — and the number in it is how often that result happens. The cell where home scores 2 and away scores 1 says 8.7%. It means: of all the times this match could be played, about nine in a hundred end exactly 2–1.

Add up every cell and you get 100%. The table is simply the whole future of the match, split into every conceivable result.

Where the numbers come from

We do not guess. Behind the matrix sits a model trained on over 100,000 played matches — it has seen how teams like these, in this run of form, against this opponent, have actually performed over and over.

Ahead of a match the model does something easy to picture: it plays the match thousands of times. In each playthrough anything can happen — sometimes the home side wins big, sometimes the away side nicks it, sometimes it stays goalless. Then it just counts: how often did it end 2–1? How often 0–0? The share of all playthroughs that ended on a given result — that is the number in the cell.

The more playthroughs, the steadier the numbers. The matrix is not a formula on paper, but the result of letting the match play out thousands of times.

Beneath every cell: who does what

And now it gets really interesting. When the model plays the match thousands of times it does not stop at the final score. In each single playthrough it also settles who scores the goals, who lays on the assist, who picks up a card, how many corners there are. All in the same playthrough.

That means beneath every cell in the matrix hides a whole cast of players. The cell "2–1" is not just a number — it is thousands of playthroughs that ended exactly 2–1, and in them we know how often the striker got both, how often the winger assisted, how often the centre-back saw yellow. So the matrix is not flat. It has depth — almost three-dimensional: the final score on the surface, and beneath each cell the whole run of the match.

That is why we can honestly price even a bet like "home win + over 2.5 goals + a given goalscorer". We do not have to guess how they connect — we just look at the playthroughs where all three actually happened together, and count. Scorers, assists, cards and corners come out of the exact same simulation as the final score, never a separate calculation on the side.

Read who wins straight from the matrix

Here is the neat part. You never need a separate model for "who wins" — the answer is already in the table.

• ·Every cell where home has more goals than away, added up → probability of a home win.
• ·The cells on the diagonal (0–0, 1–1, 2–2 …) → the draw.
• ·The cells where away has more goals → away win.

Three piles of cells, three probabilities. That is the whole 1X2.

Over/under and both teams to score — same table

Want the probability of over 2.5 goals? Ring every cell where home + away is 3 or more, and add them. Under 2.5 is the rest.

Both teams to score? Take every cell where both home and away have at least 1 — that is everything except the zero row and the zero column. Correct score? That is a single cell.

Same table, different ways to mark it. Every bet you can think of is really just a question of which cells we add up.

This is why we never multiply bets together

Here is where most odds sites go wrong. Say you want to combine "home win" with "over 2.5 goals". The temptation is to take the probability of one times the other. But that is wrong — because the two are connected. A match the home side wins is more often a high-scoring match; they move together.

Because our matrix contains every final score, we can answer exactly: mark the cells where both a home win and at least three goals hold, and add just those. That gives a different — and more honest — answer than a naive multiplication. Often a better price for you.

That is the whole point of our bet builder: the slip is priced jointly from the simulations, never by multiplying loose probabilities together.

From probability to fair odds

The last step is simple. Once we know the probability of an outcome, the fair odds is just 1 divided by that probability. 25% probability → fair odds 4.00. That is the price without the bookmakers’ margin.

Find a bookmaker paying higher than that, and the bet holds mathematical value — what we flag as EV+. And because it all bottoms out in the same matrix, we can show you exactly where that value is, match by match.

Every prediction we make is frozen before kickoff and then graded openly — hit or miss. The matrix is not a black box; it is just a table, and now you can read it as well as we can.

18+ · Gamble responsibly · stodlinjen.se