Minute 118 at Lusail Stadium, December 18, 2026. Szymon Marciniak stands offset to the left of the Argentina penalty area, raises his right arm, brings it down, and points straight at the spot. It is the third time in a World Cup final that the Polish referee must make a decision nobody on the pitch wants to receive. One line in my notebook for Argentina versus France was written at 4:12 a.m. Penang time: "96 seconds, from the moment the ball struck Gonzalo Montiel's hand to the whistle."

I have rewatched that passage no fewer than thirty times, at 0.5x and then 0.25x, stopping on every frame around the point of contact. The answer to whether the ball hit the hand arrives by the sixth frame. The question the world actually argued about was a different one: how long it takes for a clear image to become an official decision, and who carries responsibility for the time in between.

The debate after that final turned on a single word: correct. Correct or incorrect, and if incorrect, whose fault. The real variables of the match sat in another sequence entirely: time, the standard applied, and the person operating the equipment. In 2026 I was nineteen, filing three pieces for a Malaysian sports outlet during that World Cup, and an editor asked me to soften the tone so it would read more easily. I kept every number intact and accepted the later publication.

Four years on, the 2026 major-tournament cycle has closed with 48 teams and 104 matches. The gap I saw at Lusail remains open: VAR became mandatory procedure, SAOT became standard infrastructure, and the temperature of refereeing disputes never dropped. Every passage of play is a line in a ledger, and I leave none of them out.

The foundation: a 47-page notebook and a summer without crowds

In 2026 I was fourteen and living in Penang. What irritated me each morning while scrolling was not the goals but the absence of anyone writing about referees. Social football pages reposted goal clips, argued about strikers, and occasionally named a referee as someone to blame before moving on. No outlet had a dedicated column for the person with the whistle.

In the summer of 2026, across 64 matches at the World Cup in Russia, I recorded every decision in a school exercise book. By the end I had 286 yellow cards, four red cards and 22 penalties, plus 1,208 decisions classified under a form I drew myself: foul type, location, minute, score at that moment, the assistant referee's body position, and the distance from the referee to the point of contact. The France versus Croatia final finished 4-2, referee Nestor Pitana awarded 11 fouls in the first half, and I wrote a line in the margin: "I will have to do this every day."

By August 2026 the notebook was 47 pages thick and complete. It taught me one thing I still repeat in every piece I write: stay silent until the evidence is in front of you.

In 2026 the pandemic emptied stadiums across Malaysia. I was seventeen, studying journalism at Universiti Sains Malaysia, and rewatched 43 Malaysia Super League matches played without spectators. I compared the share of decisions favouring the home side between the 2026 and 2026 seasons. Home favouritism in my data fell 18.2 percent. That result does not prove referees are biased, and it does not prove the opposite either. It shows that the crowd is a variable inside the decision-making process, not decoration around the pitch.

In June 2026 the first piece on that finding went up on my blog, timed with the rescheduled Euro 2026 and the Tokyo Olympics. My analysis of the penalty in the England versus Denmark semi-final, officiated by Dutch referee Danny Makkelie, drew 3,200 reads overnight. Blog traffic jumped from 70 to 2,100 visits per week. That is where I learned to place data on a relative scale: a conclusion only means something when compared across seasons, across competitions, across referee pools.

In November 2026 I took three writing contracts with a Malaysian sports outlet for the World Cup in Qatar and was assigned to track referee Szymon Marciniak. In the final I counted 28 fouls, six yellow cards and three penalties, two of them inside regulation time. While the media lined up to praise semi-automated offside technology, I opened my group-stage data and found that four of 25 offside decisions took more than 80 seconds to resolve. The counter-piece drew 6,400 reads. An editor called and asked me to go easier. I said the numbers should stay exactly as they were.

By Euro 2026 I was twenty-one and working as a trainee commentator on an English-language podcast in Penang. After Lamine Yamal scored against France in the semi-final, colleagues across the studio called him the generational breakthrough. I quietly collected data from 50 Yamal matches for Barcelona in the 2026-24 season and set it against Messi in 2026, Mbappe in 2026 and Pedri in 2026. My 2,300-word piece concluded that at least 50 more high-density matches were needed before the generational label could be settled. It was cited by four outlets and it ran exactly three days after everyone else's.

The 2026 cycle has closed with a 48-team, 104-match format, 62.5 percent more matches than the 32-team format. Along with that comes the number of decisions to be made, the number of video reviews to be run, and the number of silent seconds before each whistle.

The core: four layers of a refereeing decision

The time layer.

On December 18, 2026, FIFA published data showing that semi-automated offside technology cut average offside decision time from 70 seconds to 25 seconds. That improvement is real and verifiable. The system uses 12 cameras mounted under the stadium roof, tracks 29 data points per player at 50 times per second, and pairs that with a sensor-equipped ball in selected matches. The machine draws the offside line within seconds; everything after that belongs to people.

An average of 25 seconds is a measurement of the system. It tells the story of the tournament as a whole. The experience of someone in the stands or in front of a screen sits in the upper percentile instead: the passages that took 62 seconds, 80 seconds, even 96 seconds. An average conceals the tail, and in football the tail is what generates the argument.

The four offside passages exceeding 80 seconds that I logged in the Qatar 2026 group stage each fell into one of four situations: the moment of the final touch was unclear because two players touched the ball in succession; a camera was blocked by another player at exactly the decisive frame; an earlier foul in the move had been missed and the entire attacking phase had to be checked; or communication between the VAR room and the on-field referee was slowed because the referee had to reposition for a viewing angle.

The key point sits here: the machine handles geometry, people handle context. The machine does not know where an attacking phase starts, and it does not know whether a player is in a passive offside position. Every passage that runs past 80 seconds is one where the machine answered long ago while the humans were still deliberating.

The standard layer.

The IFAB Laws of the Game, 2026/26 edition, published on July 1, 2026, retains the four VAR intervention categories: goal or no goal, penalty or no penalty, direct red card, and mistaken identity. The intervention threshold remains a clear and obvious error or a serious missed incident.

| Situation category | Nature | Review time in personal records | |---|---|---| | Offside leading to a goal | Objective | 25 to 40 seconds | | Penalty for handball | Partly subjective | 55 to 95 seconds | | Direct red card | Subjective | 45 to 80 seconds | | Challenge inside the box | Subjective | 60 to 110 seconds |

I built that table from my own records at the 2026 World Cup, the 2026 World Cup and Euro 2026, sampling matches with VAR interventions. What it shows is straightforward: the first two categories can be measured with a ruler, the last two only with a threshold. With offside, the line is the truth. With a challenge inside the box, the truth lives in the degree of contact, in whether the player was searching for the collision, and in which direction the ball travelled after the touch.

The phrase "clear and obvious error" is the fault line of the whole system. It is a threshold with no unit of measurement. Within a single match, under a single referee, that threshold shifts by minute, by scoreline, and by whether the passage leads to a goal. When I compared knockout rounds against group stages at the last two World Cups, the number of on-field reviews in knockout football was markedly higher, despite a comparable number of contested passages per match.

The historical data layer.

Of my 1,208 classified decisions, 137 involve handball. The phrase "making the body unnaturally bigger" appears in 61 of those situations, producing three different outcomes: penalty, no penalty, and no penalty plus a yellow card for the reaction. One sentence in the law, three results on the pitch.

In 2026 IFAB tightened the reading of arm position; in 2026 it added guidance on handball in the build-up to a goal. After each adjustment, the handball entries in my log rose, then fell back over roughly one season. I call this the adaptation lag: the law changes in a single meeting, while the referees' tolerance window only changes after hundreds of real matches.

What the data exonerates most often are the passages fans call impossible to miss. On a fast handball, the distance from the shot to the arm is frequently under three metres, and human reaction time falls below what is needed to watch the ball, the arm position and the attacker's position simultaneously. The referee is not looking wrongly; the referee is looking at three things at once.

The human layer.

Based on my experience tracking matches since 2026, the same law gets applied with different tolerance windows across confederations. In European competitions, a light contact in the box is usually read through a triad: distance, speed, and whether the defender had time to withdraw the arm. In South American competitions, the emphasis tends to fall on the advantage the attacker lost after contact. In many Asian leagues, intent still carries more weight than either of those factors.

When referees from four continents assemble for a major tournament, four tolerance windows meet inside the same week. The opening three weeks are usually a noise phase, and penalty counts in that stretch run higher than in the rest of the competition. I have gone back and verified this across the last four major tournaments.

Then there is the officiating structure: one referee, two assistants, a fourth official, plus a VAR team made up of the video assistant referee, an assistant VAR, a replay operator and an offside operator. Six or seven people for one decision. Each of them is watching a different frame, and the person with the final call is the one furthest from the point of contact.

The load layer.

A 48-team, 104-match format adds 62.5 percent more volume than the 32-team format used at Russia 2026 and Qatar 2026. The match count does not scale linearly with the number of teams, but match days and travel legs compound, particularly across three host nations spanning multiple time zones.

In my notebook, decisions from the 75th minute onward account for 31 percent of all decisions in a match but close to half of the passages contested afterwards. That ratio does not mean referees get weaker late on. It means collision density, fatigue and scoreboard pressure all peak in the same window, and the decisions there are harder by nature.

Put those two facts together and the load problem becomes clear: more matches, the same pool of elite referees, and the hardest part of every match still sitting in the final 15 minutes. A 104-match cycle generates more opportunities for error, not because people got worse, but because the number of decisions made under poor conditions goes up.

The metric layer.

Analysts have repeatedly tried to build an expected-goals-style index for referees, scoring the soundness of each decision. I have followed that line of work since 2026, and it carries the same structural flaw as the overuse of expected goals in match analysis: an index can only count what happened, never what was waved away.

With a player, every shot leaves a trace. With a referee, most decisions are decisions not to blow the whistle. A collision the referee plays on does not appear in any statistical table. Every refereeing index is therefore selection-biased at the collection stage: people measure only the passages that caused controversy, and controversial passages are precisely the ones that were handled.

The only way out of that loop is to log the passages where nothing happened. In my notebook, every half has its own line for box collisions the referee waved play on. Those lines vastly outnumber the penalty lines. No broadcaster replays them, and no index measures them.

The counterintuitive angle: technology made referees less decisive, not more accurate

The widespread assumption is that technology raises refereeing accuracy. Seven years of my own records do not support that simple reading. The first thing technology changed was accountability; accuracy came second. Once every passage can be re-examined, the cost of a wrong on-field call spikes, and the rational response for the person holding the whistle is to hand the decision to the screen.

That behaviour looks exactly like a coach switching to a back three after his back four gets carved open. It is a reputation-risk measure, not a tactical advance. A referee walking to the monitor at the edge of the pitch behaves the same way: he is not double-checking what he saw, he is transferring the decision to the system so that he is not the only one carrying it.

The measurable consequence shows up clearly in my logs. The number of immediate on-field decisions in box challenges has declined across successive major tournaments, while stoppage time has risen. In knockout football the automatic intervention threshold drops lower than in the group stage, even though the law itself does not change between those phases.

This is where refereeing data serves exoneration rather than indictment. Many referees criticised for being slow are not slow because they failed to see; they are slow because they are weighing a standard nobody published to them before kick-off. Emotion can lean, but video does not. SAOT is a steel eye, but the person operating it is still a human hand.

What has to change before the next cycle

What has to change is procedural, not emotional. A decision log published within 48 hours of every match, including passages reviewed but not intervened on, would move the argument from personal blame to standard comparison. A published time threshold, say 90 seconds, would force referees to decide on the pitch and record the reasoning in the report. Aligning tolerance windows across confederations should happen before a tournament, not during the opening three weeks when referees from four continents share one pitch.

Fans remember the names of players; I remember where the assistant referee was standing. When the next cycle opens, the thing worth tracking is not how many cameras hang under the stadium roof, but which standard gets published before the first whistle sounds.