If you watch the closing twenty minutes of every Division 2 match in 2025, you see one team doing something the rest of the league isn't.
Kintetsu's net points in the 60-to-80 minute period averaged plus 8.1 per match across the season. The next-best team in the league, in any team's best period, sat at plus 5.4. Six of the eight clubs were net negative in the closing twenty.
The chart that anchors this is a single grid: eight teams, four periods each, net points per match in the cell. There is one cell that the eye lands on first.
That cell, top right, plus 11.5 points scored and 3.4 conceded for a net of plus 8.1, is the largest single positive cell on the board. The next-largest is plus 5.4 in a different team's middle period. The closing twenty is where the gap shows up.
What this is, and what it isn't
Net points is not a fitness metric on its own. A team can outscore the opposition late because they were better, because the opposition gave up, because the bench arrived heavy, or because the score state had already tilted. Any one match could carry any of those explanations.
Across fourteen matches per club, against eight different opponents, that variance averages out. What's left is a structural pattern about which teams are doing the most work in the period of the game where most teams are doing the least.
This piece is the points-only version of that case. The training side, periodisation, weekly volumes, and the kind of work that holds a squad to plus 11.5 in the closing twenty, sits in the companion post.
The closing twenty, ranked
Kintetsu plus 8.1, more than 50% larger than the next-best team. Six of the eight clubs are net negative in this window. That's the league's structural fade. Most teams concede more points than they score in the closing twenty, and the fade is consistent enough across the season that it shows up in averages.
Kintetsu doesn't fade. The 60-80 minute period is the team's highest-scoring period and its lowest-conceding period. That's not the league pattern. That's the inverse of it.
What "fittest" means here
I'm using "fittest" in the literal coaching sense. The team that can do the most physical work in the period when the rest of the league is running out. Three observations make me confident the closing-twenty signal is doing the work it looks like it's doing.
The first observation is that the gap between Kintetsu's closing-twenty number and its own first-twenty number is the only positive gap of that direction in the league. Most teams score and concede roughly evenly across the four periods, with a soft fade at the end. Kintetsu's curve climbs.
The second is that opposition concessions in the closing twenty drop to 3.4. The defensive number falls along with the offensive number rising. Teams that fade only on attack but hold defensively are doing one half of the work; this is both halves.
The third is that the pattern is consistent across the fourteen rounds. It isn't carried by one or two outlier matches. The closing-twenty net was positive in eleven of fourteen, including against the league's other top-three sides.
What I'm not claiming
I'm not claiming this is causal. The post is descriptive. Teams that win the closing twenty win more matches, which contributes to scoring more points overall, which influences scorelines, which feeds back into the closing-twenty number. Some of that loop is conditioning, some of it is the score state tactical choices that come from being ahead. Untangling that is a different post.
I'm also not claiming the league-wide context will hold for 2026. Division 2 in 2025 was eight clubs over fourteen rounds with sample sizes that, while better than single-team analyses, are still small for any one period. The next-best closing-twenty number, GR Tokatsu's plus 5.1, is close enough to Kintetsu's that a stronger 2026 from one of them could compress the gap quickly.
What I am claiming is the descriptive headline. Across the 2025 D2 season, Kintetsu won the closing twenty minutes by a margin no other team came close to. That's the dataset.
What a GM would want to do with this
Two things, both about prioritisation rather than recruitment.
For pre-season planning, the closing-twenty net is the single number I'd anchor a conditioning programme to. It's downstream of nearly every fitness adaptation a coach can train, and it's the one number that aggregates them into a match outcome. If the goal is to move that number, the programme is doing the work; if the number is flat across pre-season, the programme isn't.
For in-game decisions, the league fade is real enough to plan around. Six of eight clubs concede more than they score in the closing twenty. A team that knows it can outwork the league in that window changes its bench timing, its tactical posture in the 50-to-60 transition, and the kind of replacements it brings on. It's not a complicated insight. It is one most teams aren't acting on, because they don't have the closing-twenty number on their dashboard.
What the next post is about
The closing-twenty number is the outcome. The training side is the cause, or at least a candidate cause, and the part I want to show in detail.
In the companion post, I cover the annual periodisation curve, weekly volumes by phase, position-group splits, and one finding I almost didn't publish because it cuts against the textbook expectation. Squad-mean training distance, week by week across the season, the in-season retention of pre-season volume, and the relationship between deload weeks and injury rate. The honest version of "how we trained to stop fading," with the negative finding included.
Coming next: the training programme.