Clay Tech

"clay-works make things real"

translated from clazytech.com

Japan's Formation Fell Behind The World's Tactical Trend

original Football

Last time, I wrote a piece called Ancelotti’s Homework. In Japan’s World Cup match against Brazil, what could Japan have done in response to Ancelotti’s tactical changes in the second half? After thinking about it for a day, I landed on a back four, a 4-5-1.

I offered that as my own personal idea. I even admitted myself it was a stretch.

But as I kept watching the knockout rounds, it stopped feeling like an idle idea. Wasn’t a back four actually the trend of the whole tournament?

So I counted.

FIFA has posted analysis data for all 104 matches, for free

I want to flag this first: FIFA’s Technical Study Group has published a Post Match Summary Report, in PDF, for all 104 matches of the 2026 tournament. No login, no payment required. I think this is surprisingly little known.

The contents are genuinely impressive. The first page shows both teams’ formation labels and starting lineups, and after that comes all of this:

The URL follows a fixed pattern, like PMSR-M76-BRA-V-JPN.pdf. A dataset of 208 records — 104 matches times 2 teams — can be built in a single day.

There’s one snag in extraction. The characters for “FORMATION” inside the PDF are set vertically, so when you extract the text they come out scattered, like F O R M A TIO N. A straightforward search won’t catch it. A rough regular expression that picks up a run of digits and hyphens near “FORMA” gets you through.

For all 48 teams, the Associated Press had already tallied first-round formations based on FIFA’s match reports, so I used that. I fixed each team’s formation label at its opening match and let only “who survived” vary as the stage advanced. That way, changes in formation and survival don’t get mixed together, and only the selection pressure is visible.

Whether it’s a back three or a back five sometimes splits depending on how you count

Before I get to the numbers, there’s one thing I want to flag.

Whether it’s a back three or a back five is usually obvious just from watching. But the count sometimes splits depending on the counting method. Anyone used to watching football knows this, but once you’re compiling statistics, you need to decide in advance which way you’re counting.

Bosnia and Herzegovina’s Round of 32 match against the United States makes this clear. When I first checked with a commercial app, it displayed as a back five. Indeed, the starting lineup registers five defenders. But FIFA’s official label was 3-5-2.

Counting Bosnia’s back line along two axes, the same eleven players can be a back five if you count by registered defender count, or a 3-5-2 if you count by wing-back behavior.

Counting by headcount

Counting by behavior

Five defenders registered

Back five

Both flanks were pushing up

3-5-2

Dedić’s overlapping runs: 33. Not a player glued to a deep position.

Same eleven players, different answer depending on how you count

Same match, same eleven players. This isn’t a case of one side being wrong. The app counts headcount; FIFA counts behavior. That’s the only difference.

Looking at the data, FIFA’s basis for counting three defenders clicks into place. Dedić on the right didn’t stay glued to a deep position. The number of overlapping runs makes that clear. So FIFA judged that this wasn’t a back five, but three center backs plus two wing-backs.

In principle, the reverse could also happen. Even with four registered defenders in the starting lineup, if one full-back constantly sits high, that could conceivably be counted as three.

So in this piece, I’ll count by the leading digit of the FIFA label. It’s consistent, and readers can open the same PDF and check for themselves. One note: in this classification, the only case where the two methods diverged was this Bosnia match, and even there both methods agree it was “not a back four.” The conclusion doesn’t break.

I counted every back line at kickoff

StageNBack fourBack threeBack fiveBack-four share
Overall48358572.9%
Round of 3232263381.3%
Round of 16161600100%
Quarterfinals8800100%

Back-four share by stage: across all 48 teams the back four is 72.9%, at the Round of 32 it’s 81.3%, and by the Round of 16 and quarterfinals it’s 100%.

Back four

Three-center-back systems

All 48 teams

72.9% (35/48)

27.1%

Round of 32

81.3% (26/32)

18.8%

Round of 16

100% (16/16)

Quarterfinals

100% (8/8)

As the stages advance, three-center-back systems disappear

Three-center-back systems — the back three and back five combined — go from 27.1%, to 18.8%, and then to 0%.

The eight quarterfinal teams were France, Spain, England, Argentina, Morocco, Norway, Belgium, and Switzerland. All of them played a back four. And for these eight teams, whether you count by their opening-match formation or by that particular match’s starting lineup, it’s 8/8 either way.

Not a single three-center-back team reached the Round of 16

This is the crux of it.

Thirteen teams used a three-center-back system as their base formation. Seven fell in the group stage: South Korea, Czechia, Tunisia, and four unspecified teams that used a 3-4-3. Six fell in the Round of 32: Germany, Japan, Sweden, South Africa, DR Congo, and Australia. That’s thirteen in total. In other words, all of them.

Germany and Australia even went to a penalty shootout. They weren’t blown out. Even so, not one of them got through.

The 3-4-3 in particular was used by seven teams; five went out in the group stage, and the remaining two went out in the Round of 32. Not a single 3-4-3 team remains in the Round of 16.

The structure is two-stage. Already at the population level there’s a bias toward the back four at over 70%, and on top of that, decisive selection pressure hit in the first knockout round.

The teams expected to come out in a back three didn’t

Another interesting point: the very teams tipped before the tournament to come out in a back three almost all started with a back four instead. The Netherlands played 4-1-2-3, Morocco 4-2-3-1, Argentina 4-3-3, Scotland 4-4-2.

For me, the symbolic case was Morocco. This is the same Morocco that went to the semifinals at the 2022 Qatar tournament with a tight back-three block. They became a subject of analysis worldwide, a textbook case. And that Morocco, after Regragui stepped down as manager three months before the tournament, came out with a back four. In front of Bounou: Mazraoui, Shadi Riad, Issa Diop, Hakimi. And they went on to reach the quarterfinals.

The side that wrote the textbook left it behind, under a new manager.

But this is not a story about “the back four being stronger”

This is the heart of the matter.

The theme FIFA’s Technical Study Group put forward in its review at the end of the first round was, precisely, “building up in a back three.” The example cited is France: in the first half against Senegal, they switched from a back-four build-up to a back three. This took the form of dropping the anchor Rabiot between the two center backs. The TSG explains that this gave France greater control in possession and resistance to Senegal’s high press. The TSG also organizes this three-man build-up into three types: two center backs plus a defensive midfielder, two center backs plus a deep full-back, and three center backs.

A phase-dependent variable defense: back four out of possession, back three in possession. Out of possession, four players stand in a flat line. In possession, the anchor drops between the two center backs to form a back three, and both full-backs push up.

Out of possession: back four

In possession: back three

Anchor

Flat line of four

Full-backs push up

Anchor drops in to make three

Defense is a back four, possession is a back three

In other words, this is what’s going on. The “variability” of 2026 isn’t a switch between a back three and a back four. It’s phase-dependent: back four out of possession, back three in possession.

So what lies behind the nominal dominance of the back four isn’t that the back four is stronger. It’s that setting the back four as the baseline, then varying the number of players on the back line by phase, has become standard equipment.

A mid-match switch from three to four: I didn’t find one

Having gotten this far, I looked for one more thing. Was there any case of a team, based on in-game judgment, switching from a back three to a back four?

I couldn’t find one.

What I found was, in every case, either the reverse direction, or a reshuffling within a back four. Argentina in the final started with a 4-4-2, went to a 4-4-1 after Enzo Fernández’s red card, and brought on Senesi to go to a 5-3-1 in extra time’s second half. England in the semifinal, after taking the lead, made defensive substitutions to line up six defenders, ceded control, and got overturned. Brazil against Haiti switched from a flat 4-4-2 to a 4-3-3 mid-block, but that didn’t change the number of players on the back line.

I don’t think this is a coincidence. I think it’s a structural asymmetry.

When a back-three team wants to go on the attack, there’s no need to reshuffle the back line to four. Push the wing-backs up and it becomes a 3-2-4-1. The nominal back line stays at three, while the number of players in front increases. Conversely, when a back-four team wants to close out a match, all it takes is dropping the full-backs to make five.

In other words, the cost of “increasing” is cheap, and the cost of “decreasing” is expensive.

Asymmetry in the cost of changing the back line: changes that increase the number of players were cheap and appeared frequently across the tournament, but changes that decrease from a back three to a back four carried a high cost, and there were zero instances of it in the tournament.

Cheap to change (many instances in the tournament)

Back four → back five: drop the full-backs

Back three → 3-2-4-1: push the wing-backs up

Expensive to change (zero instances in the tournament)

Back three → back four: everyone’s reference point gets swapped out

Increasing is cheap, decreasing is expensive

I already gave the reason the “decreasing” direction is expensive in the previous piece. Because the reference points of the center backs and full-backs all get swapped out at once, you can’t do it on the fly unless you’ve planned for it and drilled it beforehand. It’s exactly the same structure as the point about Watanabe Go — he’s played almost entirely at right center back and isn’t used to the left side.

And this tournament had, if nothing else, the most favorable conditions in history for making such a switch. At the 2026 tournament, regardless of weather conditions, every match had a three-minute hydration break at the midpoint of each half. Matches were effectively split into four quarters. That gave a manager, including halftime, three pre-arranged moments per match to deliver instructions.

The historic precedent for exactly this is well known. At the 2014 Brazil tournament, van Gaal’s Netherlands used the water break in their match against Mexico to switch from a 3-5-2 to a 4-3-3, and scored twice late to win. He himself said clearly, “we moved to Plan B during the cooling break, that’s a smart way to use the stoppage.”

There were three moments for instructions. Even so, no manager went in the decreasing direction.

The one team that went in the increasing direction was, again, Bosnia. A back four in the group stage, a 3-5-2 against the United States in the Round of 32. It was an attacking design built around wing-backs pushing forward. The result: 47.7% possession, edging out the United States’ 44%, but an xG of 0.21, 10 shots, and no goals in a 0-2 loss. What’s more, the United States had Balogun sent off in the 64th minute and played the remaining 30-plus minutes with ten men.

They had the ball, but they couldn’t create anything. They couldn’t break the opponent down even after the opponent went to ten men.

Going in the increasing direction is something you can pull off. But pulling it off is no guarantee it solves anything.

The confirmation was Paraguay

In the previous piece, I wrote this:

What Japan could have done in that moment, I think, was to build something like Paraguay’s football — sticky, tenacious, resilient defense. And then look for one shot on a long counter.

Here’s the data from that same Paraguay’s match against the United States.

It was exactly the profile I described. Don’t hold the ball, stay patient, win it back and go quickly.

And Paraguay did this out of a back four, a 4-4-2.

That same Paraguay beat Germany’s 3-4-3 on penalties in the Round of 32, and lost 0-1 to France in the Round of 16. Aside from not being able to beat France, they carried it out completely.

Which means that when I previously wrote “defense like Paraguay’s” in one paragraph, and “maybe it should have been a back four” in another, it turns out, fortunately, I was saying the same single thing twice. I had been pointing, independently, from two different angles, at the same profile of an actual, real team.

Laying it all out like this, I’ve noticed one thing.

The back four wasn’t the right answer. It’s just that the teams that could keep the back four as their baseline were the ones who had the option of variability. Teams that made the back three their baseline could increase from it, but they couldn’t decrease from it.

So what Japan lacked that day wasn’t a back four. It was the option of a back four.

And that isn’t something you can prepare in fifteen minutes at halftime.

My earlier conclusion — that I can’t bring myself to blame Manager Moriyasu’s tactical decisions — hasn’t fallen apart. If anything, it’s only been reinforced. The tournament itself wrote the answer to the homework.

References

Previous work by the author

This piece was drafted and directed by Kuzuryu, with the writing done by AI.

Originally published in Japanese at https://clazytech.com/2026/07/1716/. Translated with LLM assistance and reviewed before publication.