The Market Pays for Ball Carriers; Tactics Are Won by Off-Ball Movers

I have an odd habit during every transfer window: while negotiations hang in the air, I reopen old match footage. This week I rewatched the 2026 World Cup semi-final between Croatia and England, and my coding sheet still holds the number from that night – 47 times Ivan Rakitic left his central midfield position across 90 minutes. Not one of those runs appeared on the passing graphic broadcasters showed after the match. Not one of them was priced by the transfer market. On my desk in Shanghai there are always two sheets of A4 paper: one listing the fees of the deals currently burning, one listing the off-ball movement counts of players nobody asks the price of. The two sheets almost never agree.

The market pays for the man on the ball

In the summer of 2026, the Premier League spent more than 2.36 billion pounds on transfers, according to Deloitte's end-of-window figures. Moisés Caicedo moved from Brighton to Chelsea for 115 million pounds. Enzo Fernández arrived at Stamford Bridge in January of the same year for 106.8 million pounds. Both are central midfielders, and both were paid for what they do with the ball: progressive passes, carries across the halfway line, successful escapes from pressure. In thirty-nine years of reading scouting reports, I have never seen a column that prices an off-ball run.

It sounds like a small detail. It is not small.

In 2026, when I joined the tactical analysis unit at Shanghai SIPG, I sat through the derby against Shenhua and counted something nobody counts. Oscar – the most expensive midfielder in the league at the time – touched the ball an average of three times in the first five minutes of each half during spells of high pressing. Three times. I used video coding software to log every lateral pass from the opposition back line and stitched them into a map. The result: SIPG switched from a 4-2-3-1 to a 3-5-2 mid-match and won 3-1. The coaching staff asked what my evidence was. I handed them that small number. Every revolution starts with a number left forgotten on a desk.

From that day I stopped trusting the feel of the pitch. I trust the map.

Croatia 2026 and the geometry of intersections

In July 2026 I sat in the analysis room in Shanghai watching Croatia against England. In the fifth minute, Kieran Trippier scored from a free kick. England led 1-0, and almost immediately the story was told in a familiar mould: Croatia had plenty of the ball but no teeth, England defended scientifically and would go through.

I did not see it that way. I saw Rakitic repeatedly dropping level with the centre-backs and dragging Harry Kane out of the central corridor. Kane followed. That was the only mistake Croatia needed. With Kane out of the middle, Ivan Perisic had room to tuck inside, and Mario Mandzukic had room to attack the far post. Perisic equalised on 68 minutes. Mandzukic made it 2-1 on 109. Croatia went to the final.

Croatia in 2026 did not chase the ball. They drew a matrix, then let the ball find the intersection.

I counted 47 times Rakitic left his position across 90 minutes, and had to rewatch the high-angle footage to confirm each one. None of them was a touch. All of them were a space created. This is the blind spot of every modern data model: we count the man who receives the ball, but not the man who cleared the road for that pass.

In tactics, a position is only a node. The real value lies in the timing of the intersections. Croatia in 2026 did not play three midfielders or four; they played a set of precisely timed junctions. Rakitic dropping, Perisic tucking in, Mandzukic pinning the centre-back, Ante Rebic stretching the flank. Four men, four directions, one axis the ball had to travel through.

There is a technical detail that makes this finding hard to verify. Broadcast cameras follow the ball, not the player. When Rakitic dropped, the frame sat on the feet of the man in possession; his 47 runs lived outside the edge of the screen. To see them I had to pull the wide-angle feed, slow it to quarter speed and mark them by hand. One match took six hours. That is why most viewers, including people who watch a great deal of football, have never actually seen the thing that decides matches.

Jorginho's reverse shuttle

Three years later, in July 2026, I watched the Euro quarter-final between Italy and Belgium and noticed Roberto Mancini doing something counter-intuitive. He stationed Jorginho deeper than usual, but not to defend. He used Jorginho as a shuttle running backwards: the player voluntarily dropped into the highest-pressure zone to drag the opposing forwards out of midfield and open room for the runners above him. Italy won 2-1.

I went home and compared Jorginho's data across 60 matches spanning three seasons. The metric I cared about was not passing volume. It was the number of receptions under direct pressure. Jorginho led Europe at that moment, and the gap to the group behind him was not small. I drew a diagram I called the borderless rhombus: arrows moving continuously, attached to no fixed position. That piece drew 500,000 reads on a Chinese sports platform.

What I kept from it was not the read count. It was a question: why does the transfer market not pay for this skill?

The empty-stadium experiment

In March 2026, global football stopped. I was 49, sitting in Shanghai, doing what an ISTP tends to do: downloading data.

I took 120 Bundesliga matches from the 2026-20 season and built a model around stadium sound. My hypothesis: teams press harder when they are cheered. I simulated crowd noise and compared it with matches played in empty grounds. The hypothesis was wrong. Home teams did not press harder with noise behind them.

But the model returned something else, and that was the real finding: with no crowd, the weaker-rated sides pushed up 12 per cent more than usual. Psychological pressure from the stands is a tactical variable, and it disappears when the ground is empty. I wrote a 3,000-word study, sent it to a European tactics journal, and it was published as reference material.

An empty stadium does not strip a match of its soul. It only removes the make-up and reveals the tactical skeleton.

I wrote "I was wrong" inside the piece itself. That is the most important sentence I have ever written in a research article.

Mistakes packaged as data

Back to the transfer window. Clubs are buying centre-backs who can pass, midfielders who can carry, full-backs who can cross. All of these are ball-linked metrics. Nobody buys off-ball movement, because it has no column in the spreadsheet. The result is a paradox: a club pays 100 million for a player who has not yet played 50 top-level matches, then grows frustrated that he makes no difference – while the difference sits with someone else, ten times cheaper, who knows how to run into the right space at the right second.

And this is where the back three comes in.

Since 2026, every analysis I write follows the same five-layer architecture: shape, mechanism, timing, blind spot, price. I applied that architecture to the returning trend of the back three, and the result irritated me. Most teams switching to three centre-backs did not discover a new mechanism. They switched because their back four was being cut open too often, and three centre-backs is the cheapest way to reduce goals conceded before the press starts counting the coach's defeats.

That is a transfer of risk rather than a step forward. A coach adding a centre-back means removing a midfielder, which means less capacity to control off-ball movement in midfield – the exact thing Croatia used to win in 2026. The team looks more solid on the scoreboard, but the mechanism has lost a link. Against an opponent who knows how to drag players, the back three breaks in exactly the same way as a back four, only a few seconds later.

The execution blind spot

My counter-intuitive angle is this: the problem with the back three is not the back line. It is the two central midfielders ahead of it. In a theoretical back-three system, those two must compensate for the space by moving constantly without the ball. In practice, most players in that role stand still waiting for the ball and expose half a second of slowness – enough for a forward who can read space to exploit it.

I tested this by re-coding the goals conceded by back-three teams between 2026 and 2026. Most of those goals did not come from an error by the third centre-back. They came from no one holding the central axis in midfield while the ball was being circulated from flank to flank. That is an execution error, not a design error. And that is why buying another expensive centre-back solves nothing.

What the viewer sees on screen is not the real match. The real match happens in the gaps the camera forgets.

Verification in the next match

The first three seconds of a match do not lie. The rest is just how we fool ourselves.

In the next match you watch, count one thing only: how many times the deepest central midfielder of each side leaves his zone in the first three minutes. If that number is zero, the team is playing with shape, not with mechanism. If that number rises over time, you are watching a matrix being drawn in real time.

And if a club pays 100 million for a young player this summer, try asking what nobody in the meeting room asks: how many off-ball runs did he make last season? If nobody has the answer, then that fee is being paid for an advertisement, not yet for a tactic.