World Cup 2026 — Hydration Breaks

Football · World Cup 2026 · The Cooling-Break Effect

The hidden seven minutes

Every World Cup 2026 match is played in a North-American summer, so FIFA’s hot-weather protocol brings a cooling break in each half. They don’t appear in any feed as an event, but they leave a fingerprint: a long gap in the ball-in-play data where nobody touches the ball. Reading those gaps across the whole tournament, here’s when the breaks come, how long they last, and what they do to the game.

When do they come?

The textbook says cooling breaks fall around the 30th and 75th minute. World Cup 2026’s don’t. They cluster tightly earlier, at the first stoppage after roughly 20′ and 65′. Each row below is one match (in kickoff order); each bar is a break.

How they change the game

The ball goes dead through the break (the gap below). Does the game come back busier, with the play-minutes after a break out-shooting the minutes before it? Refreshed legs, or a coach’s reset working? Or does it wash back out to about the same rate? Toggle the half to check.

Does the break tilt the play?

The pooled view above counts both teams together. Splitting them into home vs away shots/min in the 10 minutes before vs after the break shows whether the reset hands the initiative to one side, or just lifts the tempo for both.

Did it look different in Qatar?

World Cup 2026 makes a cooling break mandatory in every half of every match. That’s new. Before 2026 the protocol was heat-triggered, not scheduled: a break only fires if the wet-bulb temperature crosses a threshold. Qatar 2022 is the clean contrast: a November World Cup, played in air-conditioned stadiums, in a country that built the tournament around avoiding the summer heat this protocol exists for. Running the exact same gap-detector over Qatar’s event stream turns up far fewer breaks, and that scarcity is itself the finding, not a bug.

Method & why Qatar looks so different

Qatar 2022 was a November World Cup, deliberately scheduled outside the Gulf summer specifically so the tournament wouldn’t need this protocol, and played in air-conditioned stadiums besides. Cooling breaks there were heat-triggered (wet-bulb globe temperature over a threshold), exactly like every pre-2026 edition, not mandatory the way WC2026’s are. So a match with no detected break isn’t a detector miss; it means conditions never crossed the trigger. The same gap-detector used everywhere else on this page finds only a handful across the whole tournament. The exact count is in the tiles above, and that low count is the finding.

Read this panel as descriptive, not inferential. A single-digit break count is too thin for a before/after difference — home vs away, or shots-lift — to carry any statistical weight; it’s shown as a snapshot of what little happened, not a claim about what a Qatar-style break does. The cross-tournament climate gradient below pools Qatar into a larger, more defensible multi-edition comparison.

Does the break flip momentum?

The folklore is that a cooling break rescues the team being battered: a coach’s chance to reset, a chance for the dominated side to draw breath. We can test it directly, and to avoid leaning on a single metric we run two independent momentum signals side by side. For each break we measure who was on top in the ~5 minutes before it and who was on top in the ~5 minutes after; a flip is the team that led before now trailing after.

There’s a confound, though, and it’s the whole ball game. Breaks aren’t called at random: most of them start the instant the ball goes out of play, after a turnover, a possession that’s just changed hands. Momentum would tend to flip at any such moment, break or no break, simply because the ball changed teams. So a flip after a break isn’t automatically the break’s doing. To separate the two we use two controls for each signal: the unconditioned random-minute baseline (does a break beat an arbitrary minute?), and — the honest one — a ball-out / turnover baseline drawn from non-break possession-changes in the same matches, which shares the break’s exact context. The fair comparison is break vs ball-out: only if a break flips momentum more than an ordinary ball-out does is the break itself doing anything. If the two are level, the “reset” is just the turnover.

The two signals capture different things:

Net EPV — home − away per-action expected-possession-value (the diverging signal the in-match Momentum timeline draws as net EPV): who is creating more danger.
Territorial pressure — every touch weighted by how deep it is (distance to the goal², home positive, away negative; the same weighting the timeline’s Momentum tab uses): who is camped in the other half.

If both flip more after breaks, that’s stronger, corroborated evidence; if they disagree, we say so and treat the result with caution. (World Cup 2026 only; earlier tournaments carry no per-event EPV or chain coordinates.)

The break’s bookends

A cooling break also freezes the game at a particular moment. What was the last thing that happened before play stopped, and how does play restart when it resumes? Aggregating the event right before every break and the restart that follows shows both what kind of event bookends the pause and where on the pitch it stops and restarts. The restart breakdown (throw-in, goal-kick, corner, free-kick, kick-off or open play) is inferred from what stopped play and where the restart sits (chains-WC carries no set-piece qualifier on passes, so we can’t read it directly). (World Cup 2026 only.)

Does it scale with the heat?

Cooling breaks aren’t new, but until this tournament they were heat-triggered, not mandatory. They first appeared at the 2014 World Cup, and only when the wet-bulb temperature crossed a threshold. So if the breaks really track climate, hot tournaments should be full of them and cool ones almost empty. Running the same event-gap detector across every tournament edition we have event data for — World Cups back to 2002, plus recent Euros, AFCON and Copa América. That is exactly the gradient that falls out.

When the shots come: 2026 against World Cup history

Line up shots by the minute they’re taken and a World Cup match has a shape: a quiet opening, a build through each half, a late surge. Six past tournaments of event data (2002–2022, ~380 matches) draw that shape as a baseline, and we can lay 2026 on top. The dashed lines mark the typical cooling-break minute in each half. The question the break poses: does the game come back hotter than the natural rhythm at that point? (Values are shots per 10 clock-minutes, both teams, not adjusted for the dead-ball time a break itself removes from a bucket it overlaps.)

Does the break cause this, or is minute 25 just busier than minute 20?

Honestly, not settled yet, and that’s the interesting part. To prove the break itself lifts the game you need a control: unplanned stoppages of the same length. Within the World Cup there are too few of them to compare cleanly (only a handful of injury / VAR breaks, against ~205 cooling breaks across the full tournament), so the same-tournament test is thin. A wider look hints the planned reset is doing real work: in cool-climate club football, injury breaks come back if anything slightly colder. But that’s a different competition, so it’s suggestive, not conclusive. The climate gradient above settles one half of it: the breaks are unmistakably real and heat-driven, not a detector artefact. The half still open is the causal one: cooling-break vs injury-break after-effects side by side across that same climate range. That’s a direct, data-driven answer to ESPN’s “are these momentum breaks?” question, coming once the tournament event data lands.

Method, the protocol & what’s next

The protocol. Cooling breaks debuted at the 2014 World Cup in Brazil, heat-triggered when the wet-bulb globe temperature (WBGT) topped ~32 °C, so historically they showed up only in the hottest matches. WC 2026 is the first tournament to make them mandatory in every match and half, which is exactly why 100% of matches here show the pattern, and why older / cooler tournaments instead scale with heat (the cross-tournament study, below).

Detection. A cooling break is read as the single longest gap in the ball-in-play event stream (≥150 s) within each half’s window (1st half 18–35′, 2nd half 58–78′). This isolates the planned break from goal celebrations / injuries / VAR (shorter, or outside the window), and is the same rule the in-match Momentum timeline draws as “WB” bands, so the two agree.

Controls. The shot-rate effect is shown against the within-WC injury-break control above: same tournament, same ~3-minute length, no competition confound. The next layer is a cross-tournament comparison (WC 2022’s cool Qatar winter as the ideal control, plus Euro / AFCON / Copa across the climate range), which needs upstream event data and is in progress.