When the Stands Go Quiet: 141 Empty-Stadium K League Matches and Which Part of Home Advantage Actually Disappeared

On 8 May 2026, at Jeonju World Cup Stadium, the first half of the K League opener between Jeonbuk Hyundai Motors and Suwon Samsung Bluewings unfolded in a strange kind of sound. The away goalkeeper shouted instructions to his back line, and the shout carried so clearly that the press seats on the east stand could hear every syllable. I sat there with my notebook open and logged a detail that had nothing to do with football: in the first twelve minutes, the referee blew his whistle four times — fewer than the average I had recorded for this exact fixture in the 2026 season. That figure appeared in no news bulletin. It opened a project I followed for the entire season: 141 K League matches played without spectators, and one narrow question — when the noise disappears, which part of home advantage disappears with it? The 2026 season kicked off nearly three months late, cut from 38 rounds to 27, split into two phases, with three-match weeks. The early phase was played behind closed doors. In late July some stadiums reopened at limited capacity, then closed again as the second wave returned. For anyone working with data, this was a rare natural experiment: same league, same laws, same pitches, one variable changed. In 2026, the K League 1 home win rate was 46.3%. Across the 141 empty-stadium matches I tracked, it fell to 34.7%. The draw rate rose 7.2 percentage points. Total goals per match dropped slightly, but the distribution by minute shifted far more: goals from the 75th minute onward fell, goals in the first half rose. The home side, in other words, lost the weapon usually deployed late — the final push. Three layers of explanation emerged once I broke the data down by group. The first layer is refereeing. Yellow cards per match for away teams were 2.41 in 2026; in the empty-stadium sample they fell to 2.08. Second-half stoppage time dropped by an average of 34 seconds. VAR interventions in penalty-area disputes fell by nearly a third. There is no evidence that Korean referees became worse or better; what changed was crowd pressure — a variable every coach knows exists but nobody could measure until the stands emptied. This implicates no individual. It says only that a marginal decision, in a single moment, is made by a human being who is either hearing fifteen thousand voices or hearing silence. The second layer is communication. In an empty stadium, the goalkeeper's shout rings out like a tactical manifesto. Away back lines could hear each other call markers, hear the full-back telling the centre-back to step up, hear the coach on the touchline. In 2026, crowd noise was a free screen for the home team. In 2026 that screen was removed. The measurable consequence: away teams were caught on the counter after pushing up less often, and their successful line-breaking passes increased. The third layer is routine. Travel, hotels, familiar pitches, familiar kick-off times, familiar pre-match meals — none of that disappeared. This is why home advantage fell but did not reach zero. Home teams still won more than away teams; they simply no longer won as much. The data that caught my attention most sat in set pieces. A goal from a free kick is the product of ten seconds of preparation nobody sees — and in a season without crowds, those ten seconds were heard. Teams use short code words to call routines: a shouted change of running direction, a clap, a signal phrase mixing Korean and broken English. With the stands empty, the defending wall could hear those signals. In my sample, goals from direct set pieces fell from 0.31 to 0.24 per match. The teams most dependent on rehearsed dead-ball routines suffered most — and that group coincided with the teams whose home win rate dropped hardest, not with the weakest squads. The title race that season followed the same data line. Jeonbuk and Ulsan stayed neck and neck to the final round; Ulsan's Junior Negrão took the golden boot with 26 goals, while Jeonbuk won the title on goal difference and consistency in decisive matches. Reading the table, you see a tight race. Reading the home-advantage data, you see that Jeonbuk lost less of it than the rest of the league — because their game model rests on structured possession rather than on momentum from the stands. This is where the popular story goes off course. The familiar telling — no fans, no fortress — lumps every team into one basket. The data disagrees. The teams that live on home emotion, that start slowly and erupt in the second half under crowd pressure, lost the most. Starting 0.05 seconds late is sometimes the way to finish earlier; and when the stands go quiet, that 0.05 seconds is no longer paid back in roars. The second angle is more uncomfortable still: some away teams genuinely benefited, not because they improved, but because their communication costs fell. Zonal defensive sides that must constantly shout their shifts were the most liberated by silence. Off the pitch, the money story runs the other way. Seongnam FC recorded a 23% drop in sponsorship during the spectator-free season. For a club on a thin budget, losing fans is not just losing the twelve players on the pitch — it is losing the revenue line the board uses to negotiate next year's sponsorship. Home advantage can be measured in win rates. The value of home must be measured on the balance sheet. COVID-19 taught football that noise is not the crowd, and the crowd is not the noise. When the stands returned, K League home win rates crept back toward the old level, but not all the way. That leaves a question for analysts: if part of home advantage is manufactured by marginal decisions and drowned-out signals, then measuring that part properly will change how we judge a good coach far more than staring at the home record will. I am keeping this dataset, not to prove that crowds matter — everyone knows that — but to answer the harder question: at which layer do crowds matter, and can that layer be trained?