The FIFA ranking: how the points are worked out and what each fixture is worth

In the edition published on 20 July 2026, the most recent one available when this piece was written, on 7 September 2026, Brazil stood fifth in FIFA's men's ranking on 1.804,92 points. Morocco, immediately behind them, had 1.803,99. The gap between fifth and sixth in the world was therefore 0,93 of a point.

That number is worth holding on to, because it makes everything that follows concrete. Between 25 September and 3 October 2026, Brazil play three friendlies inside the international match calendar window. Apply the formula FIFA sets out in its method document to the points from the edition of 20 July 2026 and a draw with Australia costs the Brazilians 2,0 points. A single draw in Townsville or Brisbane would wipe out more than the whole of the gap to Morocco.

That is the sort of conclusion the FIFA ranking allows, and one that almost never gets explained. The table is published, the world argues about the positions, and the mechanism that produces the positions stays off the page. This guide sets the mechanism out: the formula, the weight of each type of fixture, the expected result and the two rules behind many of the oddities a reader finds in the table without knowing where they come from.

Before we start, a necessary distinction. This paper has already published a guide to the UEFA coefficient, which measures clubs and associations, accumulates points across seasons and serves to hand out places in the European competitions. The FIFA ranking is a different animal: it measures national sides, works as an Elo system that adds and subtracts points fixture by fixture, and hands out no places at all. Two systems, two sums, two different uses. Confusing them is the most common mistake made by anyone coming to the subject for the first time.

What the ranking measures, and what it does not

FIFA's men's ranking is not a trophy count, it is not a seasonal average and it is not a committee's opinion. It is a running balance. Every side carries a points total, and that total moves only when the side plays. A national side that goes six months without a fixture ends those six months on exactly the points they started with, even if their position has changed in the meantime because everyone else has been playing.

The current model, which FIFA calls SUM, replaced a system that worked out an annual average of points. The difference is bigger than it looks. Under the old model the value of a result was diluted over time, a side's position wobbled from day to day without anyone having kicked a ball, and there was a well-known route to climbing the table without improving the football: avoid friendlies. FIFA lists those three among the problems the review fixed, and it names outright the ability to influence a ranking position by avoiding friendlies.

A fourth element disappeared in the review as well: the extra weighting by confederation, which meant the same win was worth different amounts depending on the continent and made climbing harder for anyone outside Europe or South America. Under the current model, the continent a fixture is played on never enters the sum.

The document these rules live in is worth a word of its own. The official text carrying the formula is not recent: it was written to describe the review that came into force in 2018, and it remains the material FIFA points to for anyone who wants to know the procedure. The governing body's method page links to it without reproducing it, which explains why almost all coverage of the ranking quotes the table of coefficients at second hand.

The sum, line by line

The formula FIFA publishes is this: P = Pbefore + I x (W - We). Turning each letter into plain English, P is the points total after the fixture and Pbefore is the total the side brought into it. I is the importance coefficient, a fixed number that depends only on the type of fixture and runs from 5 to 60. W is the result obtained, worth 1 for a win, 0,5 for a draw and 0 for a defeat. We is the expected result, a number between 0 and 1 worked out before kick-off from the points gap between the two sides.

The heart of the formula is the subtraction of We from W. It does not measure whether the side won: it measures how far above or below expectation they finished in that particular fixture. A heavy favourite that wins has beaten expectation by very little, and the gain is small. An underdog that wins has beaten it by a long way, and the gain is large. The coefficient I comes in afterwards, as a multiplier: it decides the size of the effect, not its sign.

FIFA publishes a worked example in the document, and the arithmetic is worth following. A side on 1.300 points meets a side on 1.500 points in a qualifier, which carries an I of 25. The side on 1.300 wins. By the formula they gain 17 points and move to 1.317. Their opponents lose exactly the same 17 and drop to 1.483.

Two important properties show up in that example. The first is that, in fixtures where only the general rules apply, the arithmetic is zero-sum: what one side gains the other loses, in the same quantity, and the two exceptions, the penalty shoot-out and the knockout stage of a final tournament, come later in this guide. The second is that the amount transferred depends on where the two sides stood before the fixture. There is no tariff of points per win; there is a calculation done case by case.

The table: what each fixture is worth

The importance coefficient is the easiest part to look up and the most often misquoted. The official document carries nine lines but only eight distinct values, because the coefficient 25 appears twice, covering two different situations. The table below reproduces the nine lines in FIFA's order and adds a column of this paper's own: how much a side would gain in each type of fixture by beating an opponent on exactly the same points. In that case the expected result is 0,5 for both sides, and the gain comes to half the coefficient.

The right-hand column is this paper's calculation, applying the published formula. It is there to give a sense of scale: a friendly inside the window, against an equal, moves five points, while a World Cup quarter-final, against an equal, moves thirty. That is a sixfold difference between the two useful extremes of the table, and twelvefold if the friendly played outside a window is counted in.

Two observations about the World Cup line. The document says explicitly that the coefficient 50 applies up to, but not including, the quarter-final stage. That takes in the round of 32, which made its début at the 2026 World Cup with the expansion to 48 teams: since it comes before the quarter-finals, it sits on the same tier of 50. It should be said, though, that the published table was written when the World Cup had 32 teams, and FIFA has issued no amendment to the document for the 48-team format. The reading above follows the text as it stands, not an update from the governing body.

The expected result: why beating the side ranked 138th is worth almost nothing

The expected result is worked out like this: We = 1 / (10^(-dr/600) + 1), where dr is the points gap between the two sides before the fixture, that is, side A's points minus side B's points. Anyone who dislikes exponents can ignore the expression and hold on to the behaviour, which is easy to describe.

When both sides have the same points, dr is zero and We comes out at exactly 0,5: half a point is expected from each of them, which is to say a draw. As the gap grows, the favourite's We rises towards 1 and the underdog's falls towards 0, but neither ever gets there. The constant 600 governs the speed of that climb: it sets how big a points gap has to be for the favourite's win to be treated as an obligation.

In practice the constant 600 means that a gap of about 220 points puts the favourite near 0,70 on the expected result, and a gap of 720 points takes them close to 0,94. In that second case a win confirms almost everything that was already expected, and only the small fraction left over is there to be won. That is the source of the asymmetry every supporter senses and rarely sees set out: against a far weaker opponent, a win yields next to nothing and a defeat costs almost the whole coefficient.

What is at stake for Brazil in this window

It is exactly that asymmetry that Brazil's fixture list runs into over the coming weeks. Ancelotti's side meet Australia in Townsville on 25 September 2026 and again in Brisbane on 29 September, then India in Kolkata on 3 October. All three are friendlies played inside the international match calendar window, so they carry an I of 10, the second-lowest weight in the whole table of coefficients.

In the edition of 20 July 2026 Brazil had 1.804,92 points; Australia, 28th, had 1.581,51; and India, 138th, had 1.084,93. On those figures, Brazil's expected result is 0,7021 against Australia and 0,9406 against India. The table below applies the formula to each of the three fixtures.

The last line is worth reading twice. Win all three friendlies and Brazil come out of the window 6,6 points better off. Lose all three and they come out 23,4 points worse off. The spread between the best and the worst outcome is 30 points, and it was not settled by the football that will be played: it was settled on the day the opponents were chosen. Against India in particular, a win yields 0,6 of a point and a defeat costs 9,4, a ratio of more than fifteen to one.

Which brings back the number from the top of this piece. Brazil's advantage over Morocco in the edition of 20 July 2026 was 0,93 of a point. A draw with Australia is worth minus 2,0. A defeat, minus 7,0. Any result other than a win in Townsville or Brisbane wipes out, on its own, the entire gap to sixth place, and what happens next depends on Brazil's other fixtures and on Morocco's fixtures over the same period: a draw with Australia combined with two wins, for example, still closes the window with a net gain of about 1,6 points. Winning all three, on the other hand, leaves Brazil 6,6 points stronger but still some 111 points behind England, fourth on a rounded 1.923 points in the same edition.

Three clarifications about the method behind that table, which the reader is entitled to. First: the sum is this paper's, applying the formula from FIFA's official document to the points from the edition of 20 July 2026, and it is not a projection published by the governing body. Second: the figures used here are the ones the national team pages on FIFA's site were showing on 10 September 2026, to two decimal places and under the stamp of the last official update, of 20 July 2026, and they match the ones an independent tracker, updated daily, was still showing on the same date. Third: the three-fixture total works every match out from the snapshot of 20 July 2026, whereas in FIFA's own sum the points update after each fixture played, which shifts the final total by about 0,1 of a point.

The two rules almost nobody knows

The first rule deals with penalty shoot-outs. In the FIFA ranking, winning on penalties is not a whole win. The side that wins the shoot-out records a W of 0,75, and the side that loses records a W of 0,5, the same value as a draw. Both numbers go into the same formula, against the same expected result. The rule applies only when the tie was in fact settled from the penalty spot: a fixture that ends with a winner in normal time or in extra time counts as an ordinary win and an ordinary defeat.

The practical consequence is that the side beaten on penalties is not punished as a loser. They leave the fixture treated as though they had drawn, which is a defensible reading of what happened on the pitch. Nor do the winners collect the gain of a whole win: they enter the formula with 0,75 in place of 1, so their balance comes out smaller than that of a win settled inside the ninety or the hundred and twenty minutes, and the size of that difference depends on each side's expected result.

The second rule is less well known still, and its effect is bigger. In the knockout stage of a final tournament, a defeat costs no points at all. The document is explicit: if the subtraction of We from W comes out negative at that stage, then the points total after the fixture equals the points total before. The tally simply does not move. FIFA explains that the condition exists to protect the points of sides that have reached the knockout rounds, and it applies even to a side that wins on penalties against a weaker opponent, a situation in which the 0,75 can fall below the expected result.

That rule accounts for a good share of the oddities a reader finds in the table without understanding them. A side that goes deep at a World Cup piles up gains in the earlier rounds and then meets nothing but fixtures in which the worst case is standing still. It is a ratchet: once the group stage of a final tournament is behind them, that side's balance can only rise or stay put.

Why this window moves more than the others

The international window that opens on 21 September 2026 runs to 6 October, lasts 16 days and allows up to four fixtures per side. It is the result of merging the old, separate September and October windows, which lasted nine days and allowed up to two matches each. The number of fixtures has therefore not gone up: what has changed is that they all fall at once, and they all enter the ranking sum at once.

That matters for an arithmetical reason. Since the model adds and subtracts fixture by fixture, a side that plays all four permitted matches enters the same snapshot with four sums added together rather than one, and the same holds for every other side in action at the same time. The next snapshot does not record a result: it records all of them added together.

The date of that snapshot is already fixed. FIFA says the next edition of the men's ranking will be published on 7 October 2026, a day after the window closes, and the national team pages on its site carry a countdown of the days until then.

What the edition of 20 July 2026 had already changed

The edition of 20 July 2026 was the first after the World Cup, and 105 matches fell inside its calculation period. Spain went back to first after winning the final, and Argentina dropped to second. The top of the table looked like this, with the rounded points from that edition.

By the rounded figures in that table, Argentina finished 26 points behind Spain. That is a small gap between the winners and the runners-up of a World Cup, and the reason lies in the mechanism: because the expected result between two sides on similar points sits close to 0,5, winning the final earns the champions close to half of coefficient 60, and the runners-up, having lost a knockout tie at a final tournament, lose no points at all. What separates the big climbers from the small ones at a World Cup is not the trophy on its own, it is the whole run of fixtures at coefficient 50 and 60.

The two most striking moves of the edition bear that out. Morocco reached sixth in the edition of 20 July 2026, the best placing in their history up to that publication, and Norway made the biggest jump of the same edition, 12 places, up to 19th, after reaching the World Cup quarter-finals. In both cases the movement was produced by a run of World Cup knockout fixtures, the highest tier of the coefficient table, and not by an isolated result. Brazil climbed to fifth in the same edition.

All of those numbers are the snapshot of 20 July 2026. The next one comes out on 7 October 2026, carrying whatever the September window has produced.

The FIFA ranking and the UEFA coefficient are not the same thing

A final word on the most frequent confusion. The FIFA ranking measures national sides, works by Elo and updates after every fixture played. The UEFA coefficient measures clubs and national associations, accumulates performance across seasons and serves to decide how many places each country gets in the European competitions. One is a national side's points total; the other is an allocation of club places. The two sums do not speak to each other and do not use the same ingredients.

It should also be recorded that everything described above refers to the men's ranking: the method document used in this guide deals with that ranking, and the figures here do not apply to the women's ranking.

Anyone who wants to follow the subject has a date and a test. The date is 7 October 2026. The test is the 0,93 of a point between Brazil and Morocco: with the formula on this page, the results of Brazil's three friendlies and those of Morocco's fixtures in the same window, it is possible to estimate which of the two will be ahead before FIFA publishes the answer.