After Portugal scored a goal to tie the U.S. in the 50th minute of the second half of their World Cup match on Sunday, some U.S. soccer fans and players wondered why the half hadn’t ended after 49 minutes. The question can only be answered alongside the equally valid, if seemingly nonsensical, question of why the half didn’t last 68 minutes.
Each half of a soccer match lasts 45 minutes, plus however many minutes the referee decides to add to make up for time lost to delays.1 The fourth official announces how much stoppage time to add, but that’s a minimum, not an exact number.
On Sunday, the U.S. entered minute 46 of the second half with a 2-1 lead. The fourth official signaled five minutes would be added. Four minutes and 33 seconds into that time, Silvestre Varela scored for Portugal, after a bad turnover by Michael Bradley and a great pass from Cristiano Ronaldo. Less than a minute later, the match ended in a draw — costing the U.S. a guaranteed spot in the next round. The Americans will have to earn that berth in Thursday’s match against Germany (or hope for help in the simultaneous Ghana-Portugal match).
Was the U.S. robbed? Well, maybe. But if it was, it might have been the kind of larceny nobody is ever going to investigate. FIFA, which organizes the World Cup, doesn’t disclose how its referees decide to add stoppage time for any given match,2 and FIFA’s website doesn’t list on its match-stats pages how much time was added.
So what is stoppage time?
Here’s one thing it definitely isn’t: the amount of time the ball was out of play during the half. Out-of-play time is five or 10 times as long as the time referees add back on. In the average 2014 World Cup match through Monday, the ball was out of play for 42 minutes and 11 seconds, according to data provided by Prozone, one of several companies that log every play in every match. Yet just six minutes of combined stoppage time were added to the average game. The second half of the U.S.-Portugal game included 22 minutes and 50 seconds in which the ball was out of play — yet no one was expecting the referee to add 23 minutes at the end.3
“You are seeing a lot of nothing when you are watching football,” said Gabriella Lebrecht, a statistical analyst who analyzes soccer for the London-based Decision Technology.4 “The amount of time added on at the end of either half does not represent what you saw in the match. That’s where it gets really interesting, because that’s where you get the room for subjectivity.”
Referees have guidelines they are supposed to follow; the officials are equipped with stopwatches to track delays and with as much discretion as they’d like to exercise. Substitutions, goals and bookings typically count for 30 seconds. Injuries that cause delays also generally count. Other events that stop play, such as throw-ins, corners or free kicks, rarely are included in the stoppage-time tallies unless they take an unusually long time.
In European club football, Decision Technology has found — using data from Opta, a Prozone competitor — virtually no correlation between the amount of time the ball is out of play in each half and the time added on at the end.
At the World Cup, there is a relationship, according to the Prozone data, though not an entirely consistent one. There’s roughly the same correlation5 between time out of play and stoppage time for the first half and second half. However, in the first half, it takes 10 minutes and 23 seconds of delays to add up to one minute of stoppage time, whereas in the second half, five minutes and 28 seconds typically add up to one minute of stoppage time. The reluctance of officials to add much time at the end of the first half was apparent in the U.S. match, when Opta logged two delays that totaled over three minutes,6 and reporters noticed an epic water break, yet the referee added just two minutes of stoppage time. (All told, the ball was out of play for 18 minutes and 47 seconds in the first half.)
The discrepancy between stoppage time awarded in each half may be because substitutions are more common in the second half. It also could reflect the much greater importance of stoppage time in the second half, when it can make the difference between a win and a draw, as it did for the U.S.
Notwithstanding the higher number of second-half subs, in this World Cup there hasn’t been a big difference between time the ball is out of play in the first half and in the second half: 19 minutes and 43 seconds in the first, 22 minutes and 27 seconds in the second. Citing similar statistics, football writer Gabriele Marcotti has argued that officials should stop the clock when the ball is out of play and make each half 30 minutes long.
Delays are inherent to soccer, and persistent throughout any match. During the World Cup’s first 36 matches, the ball was out of play an average of 38 percent of the time during the first 15 minutes, and 47 percent of the time during the 15 minutes in the middle of the second half. Other 15-minute segments also featured more than 40 percent dead time.
Given this inconsistent state of affairs, the U.S. was unlucky, but not excessively so. Based on the usual relationship between time out of play and stoppage time, the U.S. could have expected to play an extra four minutes at the end of the match, rather than five. And several reports indicate that the referee originally signaled just that much time. Subtract a minute from the match and U.S. fans would have partied all night.
However, after the fourth official prepared to signal the amount, but before he did so, his electronic board was occupied with a U.S. substitution that went on and on. A full 82 seconds elapsed between the last soccer play logged before the substitution and the first one after it, according to Prozone. When the substitution was finally made, the official signaled five minutes. The U.S. might have gotten away with the lollygagging earlier in the half, but it was hard to ignore right at the moment when the fourth official was ready to signal the amount of stoppage time.
And even if the fourth official had signaled four minutes, the referee would have been free to let the clock run to make up for the substitution.
After Portugal tied the match, the referee was also free to add time to make up for the team’s celebration, but he didn’t let the clock run beyond five and a half minutes. U.S. coach Jurgen Klinsmann said the game could have been allowed to go on even longer than it did, which would have given the U.S. time to try to score a game-winner.
Other matches have had even more anomalous stoppage-time totals. Germany and Ghana got an extra six and a half minutes at the end of their tied match to try to get a winner, when the amount of second-half dead time suggests four minutes would have been more reasonable. Cameroon and Croatia, meanwhile, got about two fewer minutes than they could have expected, perhaps because the referee saw little reason to extend the 4-0 Croatia blowout.
The inconsistent relationship between dead time and stoppage time could incentivize time-wasting. If players know the referee won’t count a little extra time taken before a goal kick or free kick, they’d be hurting their teams if they didn’t take advantage when protecting a lead. Referees can punish such behavior by giving a yellow card to players who dawdle, though they rarely do so until late in the match, and almost never give two yellows — which would result in a red card and expulsion from the match — for the violation, Lebrecht said.
Inconsistency also leaves room for uneven application of the rules that benefits some teams more than others. In the English Premier League, this is known as Fergie Time, after the propensity of officials to give Alex Ferguson’s Manchester United teams more time at the end of home matches they were losing, and of United players’ propensity to score equalizers during that time. The Guardian and Decision Technology have analyzed the data and concluded the phenomenon is real. Researchers have found La Liga home teams enjoy the same sort of advantage in the awarding of injury time.
If you’re looking for conspiracies, consider that there was no home team in Sunday’s match. Alex Ferguson isn’t coaching in the World Cup — or anywhere else for that matter — and only Brazil is playing at home. Portugal has strong historical ties with Brazil, which could cut either way, and U.S. fans bought more tickets than those from any country other than Brazil. The referee was Argentinean. Without any clear standard for stoppage time, there’s no way to know what factors on and off the field motivated the decision to make it five minutes instead of four.