Five Substitutions: The 15-Minute Window That Shapes the Final 20 Minutes

In my tracking notebook, the data line that made me stop longest during the 2026 Project Restart sits in the 60th-to-75th-minute window. Liverpool raised their pressing intensity exactly when opponents sent on three substitutes at once, and their expected goals (xG) rose by 0.23 after each substitution. I logged 20 matches, broke each one into 15-minute cells, and cross-referenced substitution timings with xG swings. The scoreline never says what I saw: the five-substitution rule pulled the centre of gravity of a match away from the first half and pushed it toward the final 20 minutes, where tactics meet stamina and where squad depth turns into a structural advantage. I was twenty-one that year, working as a research assistant at a university in England. The professor in charge judged the research direction too small to become a topic. I finished the log anyway, sent it to a student journal, and months later received a letter from a Burnley analyst. He asked for the raw dataset of all 20 matches. From then on, I stopped reading a football match as one continuous 90-minute block. I read it as six time cells, each with its own rhythm, its own attrition rate, and its own spatial limits. ONE LINE OF LAW, TWO PURPOSES In May 2026, IFAB permitted five substitutions on a temporary basis to ease the load on players after the pandemic break. In June 2026, the law was made permanent. The Premier League only adopted the full five substitutions from the 2026–23 season, after a campaign of debate about fairness. The wording looks purely technical: a maximum of five players, three stoppage opportunities outside half-time. In practice it is a tactical lever. The mechanism is so simple it is easily overlooked. A pressing team needs three things: energy, structure, and a replacement. Before 2026, energy was a constant that declined linearly with time. After 2026, it became a manageable variable. Coaches no longer had to choose between pressing and surviving to the 90th minute. They bring on two or three players around the 60th minute, and the press is recharged. A rule changes one line; football philosophy changes a generation. The bench shifted from a shelter for the injured into half a tactical squad. The market value of the specialist who plays the final 30 minutes went up. And most importantly, the match split into two zones with different rules, the first half for building position, the final 20 minutes for deciding through attrition. I learned to read pressing at eighteen, after RB Leipzig beat Freiburg 4–1. I analysed Ralph Hasenhüttl's 4-2-2-2 and the way Timo Werner moved into the space behind the full-backs. A male journalist left a comment asking what a girl knew about pressing. I rewatched 14 matches, counted 212 pressing sequences, built heat maps and published them. The only response I accept is letting the data speak. SIX 15-MINUTE CELLS AND THE BREAK POINTS I split every match into six cells and measure three indicators inside each: PPDA, xG and the number of possession losses in a team's own half. PPDA is the number of passes an opponent is allowed before each defensive action; the lower it is, the higher the pressing intensity. xG measures chance quality, not outcomes. The 0–15 cell is almost always the thinnest for xG. It is the probing phase, both teams carry a high PPDA, and the ball circulates in midfield. Viewers call it a slow start. To me, it is when two defensive lines read each other's build-up structure. The 15–30 cell is the first pressing wave. PPDA falls clearly, especially for the home side. Possession losses in one's own half rise, and most of the dangerous chances of the whole match begin forming in this very cell. The 30–45 cell is the pre-half-time window. Micro-fitness bottoms out for the first time, set pieces thicken, and that is why this stretch concentrates more conceded goals than spectators' intuition suggests. I once tracked 47 set pieces taken by France in the 2026 World Cup group stage before their quarter-final against Uruguay. When France won 2–0, with Raphaël Varane opening the scoring from Antoine Griezmann's free kick, nobody asked again why I had spent three days on one free kick. Every number is a testimony. My job is to make sure they cannot lie. The 45–60 cell is the dead zone of the match. Tempo drops, half-time adjustments have faded after roughly twelve minutes, and xG per minute hits its lowest level. Many teams lose control here without ever knowing. The 60–75 cell is the leverage window, where I spend most of my research time. The two teams enter it with different fitness states, and coaches begin to intervene. In my tracked sample, substitution density in this cell runs many times higher than in any other. Liverpool in the 2026 season is the clearest example: they usually sent on two or three players at once, and their xG rose by 0.23 compared with the preceding stretch. What stands out is that the increase did not come from the individual quality of those coming on. It came from opponents losing their pressing reference during the first five to eight minutes after the change. The 75–90 cell is a war of attrition. This is where squad depth speaks. A team with five quality options can sustain its pressing structure; a team with three options is forced to drop its block and accept living inside its own penalty area. The fitness gap does not show up in the scoreline immediately. It shows up in clearance counts, in the distance between lines, in full-backs no longer pushing high after the 80th minute. There is one technical detail that ordinary stat sheets miss: a substitution does not only change people, it changes spatial reference points. A team pressing high uses the opposition striker's position to trigger its traps. When that striker leaves the pitch, the whole system misreads for a few minutes. That is the golden window, and it is also why good coaches substitute by schedule rather than by instinct. Three stoppage opportunities outside half-time turn substitutions into a scarce resource. Sending on two players in the same opportunity means trading away all reactive capacity for the rest of the match. Over the long run, this law operates like a tax on squad depth: any team without enough quality bodies to split the pressing load into two shifts pays the price in the sixth cell, every matchweek, every season. A COUNTER-INTUITIVE ANGLE: THE BLIND SPOT INSIDE MY OWN MODEL When I presented the Liverpool data, I received a hard question: how can you be sure the 0.23 increase came from the substitutions and not from Liverpool chasing a deficit and being forced to attack. This is a genuine blind spot, and I have to admit it. Post-substitution xG is contaminated by score state. I had to split the sample into two groups: leading and trailing. When I did, the increase survived but shrank, ranging roughly between 0.11 and 0.15. The larger lesson lies elsewhere. The advantage of five substitutions belongs not to the richest club but to the fastest decision-maker. A double change in the 58th minute creates an entirely different pressure from the same double change in the 72nd. In my sample, the team that substitutes before its opponent has a clearly higher probability of keeping a clean sheet over the following 15 minutes. The gap between clubs in this category does not sit in the transfer budget. There is one more paradox worth noting: the fifth substitution is often left unused. Most teams in my tracked sample use only three or four, keeping the last one free until the 85th minute in case of injury. On paper there are five cards; on the table only three get played. Bias is simply noise data that the market has not yet learned to process. WHAT TO TRACK NEXT I do not predict. I only read the data one beat faster than everyone else. Next season, three indicators will reveal where this law is heading: the usage rate of the fifth substitution, the average substitution minute of the leading team, and PPDA fluctuation inside the 75–90 cell. If fifth-substitution usage rises and the average substitution minute drifts toward the 65th, football will have accepted that the match has shortened to 75 minutes plus three timeouts. At that point the question is no longer who has the deeper squad. The question is who understands that the final 20 minutes have become a separate match, with separate rules, prepared for before the ball ever rolls.