Daejeon enter the final third more than all but two clubs in the division and finish sixth for shots. Nine of their 27 matches ended goalless, and all nine fall before R19. This is their 2026 attack read from one end to the other.
In the previous episode Gwangju were a team that could not get into the final third. Daejeon are the inverse. Third in the league for getting there, sixth from bottom for shooting once they do.
Converted to shots per final-third action it comes to 0.059, 11th of 12. Gwangju were last for arriving but 5th for shots per arrival. Two clubs build a bottom-half attack for exactly opposite reasons.
The odd part is that when they do shoot, it goes in. The share of their shots taken inside the box is 64.6%, 2nd in the league, and they score on 36.9% of shots on target against a league baseline of 29.6%. Their 38 goals beat their 35.9 xG.
The quality is good and the quantity is not. What Daejeon lack is not good chances but the number of them.
Nine of 27 matches produced nothing. One in three. And all nine of them fall before R19. That is too clean a split to shrug at.
The first suspicion is that opponents sat deep and locked the game. They did not. The average height of opponents' defensive actions was 67.4 in the goalless matches and 68.6 in the rest — effectively identical (measured in Daejeon's attacking direction, higher means a deeper block).
The comparison carries two traps, so it was controlled twice. First, the nine goalless matches cluster in the first half of the season, so they are confounded with time — the comparison was rerun within the first half only. Second, comparing "the 0-0 phase" is circular: in a match they eventually win, that phase ends the moment the goal goes in, so by construction it contains the passage that produced it. Instead the comparison uses a fixed window: the opening 15 minutes, whose length cannot depend on the outcome.
Passes differ by 11% and entries by 11%, while left-side key passes run 0.67 against 6.00. The nine silences had already begun by the fifteenth minute, and they began in one place.
1.11 goals per match across the first 19. 2.12 across the last 8. Goalless: 0.
The first question is whether the opponents got easier. The average league position of their opponents was 5.8 before and 6.6 after — near enough the same. More strictly, restricting to the seven clubs they played in both halves (Anyang, Gimcheon, Gangwon, Seoul, Jeju, Ulsan, Gwangju) leaves the result intact.
| Same seven opponents | First half, 11 | Second half, 8 | Change |
|---|---|---|---|
| Goals per match | 1.36 | 2.12 | +56% |
| Final-third actions | 175.1 | 212.4 | +21% |
| Key passes | 8.64 | 10.12 | +17% |
| of which on the left | 3.64 | 4.50 | +24% |
| Shots | 11.45 | 11.75 | +3% |
| xG per shot | 0.114 | 0.144 | +26% |
The line to watch is that the number of shots barely moved (11.45 → 11.75). What rose was their quality: the same volume, taken from better places. Splitting the shots by distance shows exactly which places.
This is what the rise in xG per shot actually is. Daejeon did not talk themselves out of long shots — from beyond 25m the count went up, from 1.42 to 1.62. What changed is that chances to shoot from inside 12m grew by 81%.
Every player whose minutes jumped in the second half was already at the club in R1. Nobody arrived. What changed is that they started appearing every week.
| Player | First half | Second half | Full matches (95m+) |
|---|---|---|---|
| Lee Myung-jae (16) | 12 / 19 | 8 / 8 | 42% → 88% |
| Cho Seong-kwon (4) | 13 / 19 | 8 / 8 | 63% → 100% |
| Anton (98) | 10 / 19 | 7 / 8 | 47% → 63% |
| Massa (7) | 10 / 19 | 8 / 8 | 25.5 → 62.0 min/match |
| Ju Min-kyu (10) | 18 / 19 | 8 / 8 | 48.7 → 73.5 min/match |
| Eom Won-sang (11) | 8 / 19 | 5 / 8 | 18.9 → 44.2 min/match |
| Gone over the same span — João Victor (77) 15 matches→0 · Kim Min-deok (3) 9→0 · Park Gyu-hyun (2) 7→0 | |||
What happened on the left is the clearest of it. In the first half of the season five players split the left-side key passes and the leader held 27%. In the second, Lee Myung-jae (16) alone holds 42%.
Splitting where key passes originate into 5 lanes by 3 thirds shows the weight shifting out of the centre and onto both flanks.
A cross only becomes a key pass if somebody is there to meet it. In the second half the people standing in the box settled as well — the number of players taking at least 0.25 shots per match from inside the area went from 8 to 10, and box shots themselves from 6.68 to 8.50.
To see what those numbers looked like as football, four matches — chosen not by result but because each carries one of this piece's claims.
Jeju give the same pairing. Losing 0-1 at home in R9 they took 9 shots worth 0.61 xG at an average of 18.4m. Away in R26 they took 10 shots worth 1.96 xG at 10.5m — almost the same count and three times the xG. They lost that one 1-3, and what they built was nothing alike.
The natural explanation here is "playing every week brought their sharpness back". That explanation was tested.
Across three seasons of K League 1, 2024–2026, 12 clubs, 24,894 player-matches, the comparison was made within each player: matches he entered having played 180+ minutes across the previous three rounds, against matches he entered on under 120. Differences in ability are controlled away automatically.
So no individual's passing became more accurate. Lee Myung-jae (16) did not start striking the ball better. He was in the same place every week, and the same people were standing in the box, and the crosses began finding somebody.
Form did not attach to the players. It attached to the connection.
The distinction matters in practice. If the players have come good, you wait. If a combination has settled, the job is not to break it.
In one sentence: Daejeon's 2026 attack was never about the ability to arrive but about what happened after arriving, and that showed up as nine goalless matches before disappearing after R20. It disappeared not because of new players, but because the ball coming in from the flanks finally began to find somebody.