Four wins, eighteen draws, seven defeats in 29 matches. This is a side that draws, not one that collapses. The way in is not open play — it is the dead ball, and the minutes when they are ahead. A pre-match briefing built from 41,879 raw events parsed from the K League Data Portal.
Gimcheon Sangmu is a club whose squad changes on discharge dates — dates that have nothing to do with the transfer window, and no notice attached. In this season alone, three regulars played their last match on a single day, 29 August, one of them the first-choice goalkeeper who had started 25 of the previous 26 rounds. That was ten days after the summer registration period closed.
One more thing worth knowing. Gimcheon's 2027 place in K League 2 is already fixed, regardless of where they finish. It follows from the expiry of their hosting agreement, not from results. So the collective incentive in their remaining fixtures is not the table. Do not read that as apathy, though. Every player in that squad is auditioning for a return to his parent club and, in many cases, for the contract that defines his career. Nothing collective is at stake and everything personal is.
This runs against the instinct, and it is the single most exploitable thing about them. Across the whole season they conceded 36 goals against 41.1 expected — five to the good. They give that entire surplus back in one score state.
Level and behind, they concede less than expected. Which means a match in which they lead and you chase is the shape that favours you. Do not treat going 0–1 down as a disaster. Treat it as a rehearsed scenario.
489 minutes leading — 7.73 goals expected, 14 conceded. Poisson P(X≥14) = 0.027. Conceded per 90: leading 2.58 / level 0.74 / behind 1.28. The mean expected value of the 14 goals conceded while leading was 0.294, lower than the 0.355 of everything else — they were not being opened up, bad chances were going in.
Their entire −8 goal difference comes from stationary play. Set pieces 2 for, 13 against. Open play 26 for, 23 against. In the run of play this is a winning team. Do not let the play run.
Concretely: winning a corner or a wide free kick is itself a scoring chance. And finish on the first ball. There is very little need to prepare second-phase routines.
Of the eight set-piece goals conceded (penalties excluded), seven were finished within three seconds of the dead ball and four on the first contact. Across 111 corners faced, the first Gimcheon defensive touch averaged 6.1 from their own goal line — the edge of the six-yard box — and 42% of those touches were clearances. They concede on 3.6% of corners faced.
In 29 matches they have conceded six penalties and won none. A differential of −6 is the worst of the twelve clubs, and they are the only side on zero won. There is no reason to avoid a physical contest in their area.
The timing is consistent too. Three of the six came in added time at the end of a half. A cross or cutback into their box in the closing minutes is worth more than it usually is.
Penalties conceded 6 / won 0 (league average conceded 3.5). Three in stoppage time — 45+3, 45+4 and 90+3. The fouls were committed by five different players, so this is not one individual. It is structural.
Gimcheon's aerial ability is not distributed across the team. It sits with two centre-backs. Lee Jung-taek and Byun Jun-soo win in the mid-60s in their own half and in midfield. Every other outfield regular is at 43% or below.
In practice: aim near and far post and miss the centre of the six-yard box where those two stand. And the best moment of all is while both of them are up the pitch for a Gimcheon set piece — the combination left behind is the weakest aerial group they can field.
Team aerial win rate 46.3% (467 won, 542 lost). By zone: own third 47.8%, middle 50.0%, attacking third 39.1%. Lee Jung-taek wins 67.3% in midfield and Byun Jun-soo 66.7%, while Lee Gun-hee (forward, 186 cm) runs 37–42% in every zone, Ko Jae-hyun 38–42%, and Park Se-jin 18.8% in midfield.
The axis is not left and right. It is depth. Crosses hit early against Gimcheon are worth almost nothing; crosses from the byline are three times as dangerous.
And a cross delivered straight from a set piece is far more dangerous than one from open play — which converges on instruction 2.
Across the 496 crosses Gimcheon faced: byline 4.2% conversion, deep positions 2.2%, early positions 1.4%. Set-piece deliveries are 83 crosses (17%) and produced 6 of the 16 goals (38%); open-play fast breaks are 257 crosses and 5 goals. Per cross, 3.8 times more dangerous.
Gimcheon's pass-failure rate is lower than their opponents' in every third of the pitch — 17.2% against 22.4% in their own third, 12.8% against 15.4% in midfield, 28.4% against 31.6% in the attacking third. This team does not concede because it is pressed into mistakes. A plan built on winning the ball high has no support here. If you press, press to push the game up the pitch and generate dead balls, not to force turnovers.
They do not sit deep. The mean position of a Gimcheon defensive action is 28.5 leading, 29.1 level, 29.6 behind — a spread of 1.1 units, about one metre. The block height is effectively fixed. What changes when they lead is possession: 462 passes per 90 drops to 386, a fall of 16%. They are not retreating, they are giving the ball away. So the session to run is not "breaking a low block" — it is immediate attack from a regain.
Stated plainly because this was my own finding, and I withdrew it. In aggregate, crosses into Gimcheon's right rise from 7.60 to 11.23 per 90 when they lead, up 48%. But conversion runs the other way — 4.1% on the left against 2.8% on the right (binomial p = 0.32) — and once the phases are separated, the dominant one, fast breaks, rises to exactly 6.44 on both sides. Opponent by opponent only five of seven show the increase. It is an artefact of adding things together. Do not build a plan on a left-right bias.
There is nothing there. Across 82 Gimcheon set-piece deliveries, opponents produced zero shots within 25 seconds. Their corners being harmless and their corners being a counter-attacking opportunity are two different claims, and only the first is true — 147 corners, one goal, a 0.68% conversion rate that is the lowest of the twelve clubs. Defend them, exit safely, and hang your counter-attacking triggers somewhere else.
This is the basis for instruction 4. The team average of 46.3% does not describe this side. Split it by zone, and then by player, and the picture appears.
It is worth noting that away at Gimcheon is the better fixture. They have won once in fourteen home matches and concede 0.36 goals per match more at home than away. There is no reason to treat the trip as a difficult one. With a sample of fourteen matches, though, treat this as context rather than strong evidence.
That is the attacking half. Now the defending. The good news first: Gimcheon's set pieces are harmless. 147 corners produced one goal, a conversion rate of 0.68% that is the lowest of the twelve clubs and about a third of the league average. Conceding a corner to this side is not, in itself, a problem.
The danger is all in their 26 open-play goals, and even that is concentrated.
| Player | Minutes | Goals | Assists | Shots | Key passes |
|---|---|---|---|---|---|
| Ko Jae-hyun MF | 2,166 | 7 | 3 | 53 | 19 |
| Lee Gun-hee FW · 186cm | 2,183 | 5 | 2 | 58 | 17 |
| Park Chul-woo DF | 2,434 | 2 | 3 | 22 | 31 |
| Park Tae-jun MF | 2,293 | 2 | 1 | 18 | 40 |
| Lee Jung-taek DF | 2,483 | 1 | 3 | 3 | 12 |
| Team total | — | 28 | 20 | — | — |
The top three account for 22 of the team's 48 goal involvements — 46%. Contain Ko Jae-hyun and Lee Gun-hee and you have closed half of this side's scoring. And the key passes come from Park Tae-jun (40) and Park Chul-woo (31) — the supply runs from midfield and the back line, not from the front. Pressing the origin of the forward pass is a better use of a man than marking the striker.
Worth adding that Lee Gun-hee has taken 58 shots, the most in the squad, for five goals. Conceding the shot while denying the position is a reasonable trade. Gimcheon's shots-on-target rate is 29.5%, the only figure in the twenties among the twelve clubs.
Press high to break the build-up (their pass-failure rate is lower than yours) · assume they will sit deep (the block moves one metre) · target a left-right bias (tested and rejected) · plan counters from their corners (zero shots within 25 seconds of 82 deliveries).
This is not a bad team. In the run of play they are winning, 26 to 23. To beat them you have to stop the play running, hold the ball still, let them get in front — and attack the minutes that follow.