With the finalists of the Qatar 2022 men’s World Cup decided, it’s a perfect time to look back on some of the incredible moments that shaped the tournament: Morocco’s stoppage-time brace against Belgium that set the tone for their run to the semifinals, Kylian Mbappe’s sparkling second goal against Poland that put him in the lead for the Golden Boot and sealed a knockout-round win in the 91st minute, and Argentina blowing a 2-0 quarterfinal lead to the Netherlands 11 minutes past the end of regulation time (before forcing, and winning, a penalty-kick shootout that kept Lionel Messi’s World Cup title hopes alive).

Wait a minute — just how much of this World Cup has happened during stoppage time?

From the opening match, it was clear that this competition’s official time was going to be unusually kept; Ecuador’s 2-0 cakewalk against their Qatari hosts went on 10 minutes and 18 seconds longer than expected. FIFA referees committee chairman Pierluigi Collina soon confirmed that throughout this World Cup, officials would be adding much more time than usual, at least in part to punish teams that deploy time-wasting tactics.

With all but two matches¹ in the books, the record is clear: These games have been running far longer, on average, than in any previous World Cup. But according to ESPN’s Stats & Information Group, it isn’t a one-off. It’s an acceleration of a decadeslong trend:

In Russia 2018, there was a then-record average of 7.3 added minutes per regulation game (i.e., excluding overtime periods). But that was still, according to FiveThirtyEight’s research, nowhere near how much time should have been added if officials were really trying to account for every second the ball was out of play. Thirty-one of the 32 games FiveThirtyEight measured undercounted stoppages by at least two full minutes. Twenty-one undercounted by at least five minutes, and 11 undercounted by an incredible eight minutes or more. Had 2018’s officials accurately tracked every lost second, an average of 13.2 minutes would have been added to every game — nearly double the actual average.

The 2022 added-time average, 11.6 minutes, isn’t quite that high. But it’s a 59 percent increase over 2018, a 93 percent increase over the prior record, set just four years before that, and a 136 percent increase over 2010. All that added time doesn’t just stretch out the length of the game, it means a larger share of all minutes being played are ones that weren’t originally on the game clock. And with stoppage time accounting for so much of these games’ total time, it’s no surprise that a decent chunk of the tournament’s most meaningful soccer has been played beyond the 45th and 90th minutes.

Some of that is because a record 11.5 percent of all the regular-time soccer at this World Cup has been played after the referee’s watch has already run for 90 minutes. But even after accounting for all that added time, a disproportionate amount of offensive action has still been concentrated within it. In Qatar, 12.6 percent of all goals have been scored — and a record 13.8 percent of all expected goals have been generated — during stoppage time.

It’s not unusual to see such a flurry of activity during the extra minutes officials add to the clock, since those moments are usually among the most frenetic in any game — particularly at the end of the second half, where the majority of regulation stoppage time takes place. (The average 2022 World Cup game has seen 7.6 extra minutes tacked onto the end of the second half, nearly double the 4.1-minute average added to the end of the first half and more than the average added to entire games in 2018.) Only one World Cup since 1986 — the 2010 tournament in South Africa — didn’t feature a disproportionate share of xG taking place during stoppage time, relative to its share of all minutes. The desperation of a close game in its final minutes is a good recipe for creating scoring chances.

But because there has been a concerted effort to add more time than what the game clock prescribes — and it’s clearly working — that means stoppage time is taking on a more prominent role in determining the outcomes of World Cup games. And that trend might continue: The last time there was such a significant jump in stoppage, from 1990 to 1994, there was a one-cycle-later spike in the share of goals and xG generated during those minutes. In 1998, the share of minutes played in stoppage time only increased from 5.4 percent to 5.8 percent — but the share of xG generated in those minutes jumped from 6.0 percent to 10.1 percent, presumably because teams could actually plan tactics for what to do with all that extra time. If this trend repeats itself, the expanded field of 48 teams might show up to the 2026 men’s World Cup co-hosted by the United States, Canada and Mexico prepared to make even better use of their games’ extended final minutes.

That’s not necessarily a bad thing. One might even argue it makes for a more exciting product, despite taking longer to consume. Actual xG generated during non-stoppage time has been roughly flat over the last three World Cups, so it’s not like the regular-time action has been more boring. But it’s definitely a different thing than we’ve seen in World Cups past. And if the trend toward ever-increasing amounts of stoppage time continues, it’s a thing soccer fans should get used to seeing more of at the game’s most famous tournament.

Neil Paine contributed research.

Check out our latest World Cup predictions.