They get there.
Then they stop.

This one started, as they all do, with a single line of the table.

7th · Pohang Steelers · 27 matches · 10W 5D 12L · 35 points · 24 scored, 31 conceded

Something catches there. Twenty-four goals is 11th in the league. Only Gwangju, on 18, have scored fewer. And yet Pohang sit 7th, and they have won ten matches.

A team with the second-worst attack in the division wins ten games. Usually that means the defence carries it. And the defence is fine — 31 conceded is 7th, 1.15 a game. But that alone does not get you there. To win ten with the 11th attack in the league, you have to take what you get, and then hold it.

So the question set itself. Why does this team have so few goals, and how are they 7th anyway?

01The first answer — “the goals vanish at home”

and why it is wrong

The first split was home and away. The result was startling.

• 0.36 goals a game at home is last in the league. Gwangju, 11th, are on 0.79 — more than double.
• Seven of eleven home matches ended goalless, and Pohang have never scored twice at home all season (one is the maximum).

Written as a sentence it is perfect. “Pohang cannot score at the Steelyard.”

It is wrong. Lay the fixtures out by date and it falls apart.

Pohang’s home matches are bunched at the two ends of the season, with nine consecutive away games in between. That away run is the best stretch of their season — 16 points, 14 goals.

02The second answer — “opponents sit deeper at the Steelyard”

also wrong

Still, something nagged. At home Pohang shot more and scored less. Why? Taken as season totals, an answer seemed to be there.

All four point the same way — at home the opponent drops off, Pohang hold the ball, and shoot into a crowd.

Except every one of those numbers is contaminated by the scoreboard. Pohang conceded first in seven of eleven home matches. Concede first and the opponent drops off and you end up with the ball. That is a consequence, not a cause.

So the same numbers were recalculated on the window before the first goal of the match — Pohang before the scoreboard has said anything.

One difference does survive the cut.

Still level, and Pohang shoot more and shoot worse. With 53 shots at home and 60 away, this settles nothing on its own. Note the direction and move on — the problem is not the number of shots but where they are taken from. That sentence returns later.

03Try the time axis instead

Pohang’s three blocks

Drop the venue. Cut the season into blocks of nine matches by date. The picture changes completely.

The point is that blocks 1 and 2 have essentially the same xG per match — 1.44 against 1.41. They created the same amount. What changed is whether it went in (0.39 → 1.11).

In block 1 Pohang accumulated 13.0 xG and scored five. That was not a level of performance, it was a number waiting to come back, and it came back. The 14 goals of block 2 are not a team suddenly improving. They are a team returning to what it had been doing all along.

Block 3 is a different kind of thing. Goals÷xG falling back to 0.60 is a problem, but the real problem sits upstream. xG per match fell 34%, from 1.41 to 0.93. The creation itself collapsed — and at the same time xGA rose 41% (0.91 → 1.28).

Which narrows the question to one thing. What changed at the end of July?

0429 July

the day the team had to redesign itself

On 25 July, at home to Jeonbuk, Pohang’s first-choice striker Lee Ho-jae was withdrawn on 59 minutes. It was his last appearance. Four days later his transfer to Darmstadt of the German second division was announced. From 1 August he is not in the squad.

This is where the piece proper begins, and where the focus has to move. “They lost their striker so they stopped scoring” is true of any team anywhere, and it is not analysis. The question worth asking is this. Why is one transfer an event of this size at this particular club?

Start with the shape of the result.

1. The shots did not stop. There are slightly more of them (10.3 → 10.6).
2. The entries did not stop either. There are more of those too (118 → 131). They still get there.
3. What moved is where the shot is struck from. The share taken in the box collapsed from 71% to 58%.
4. So xG per shot fell 39%, and xG per match went from 1.38 to 0.86.

The route forward belonged to the team. The finishing position did not.

That is the point at which the structure of this side becomes visible. And the structure was already there in the data from before the transfer.

05Only one player lived inside the box

18 matches, 132 shots, broken down

One in four of the team’s shots inside the box belongs to one man, and the next name is a centre-back.

And that centre-back’s shots are a different kind of shot. Of Park Chan-yong’s 25 shots this season, eight (32%) came within twelve seconds of a set-piece delivery. The team average is 10% (28 of 282). Park is not someone who lives in the box; he is someone who goes up for set pieces. Take him out and the number of players standing inside the box in open play is, effectively, one.

The season totals say the same. Lee Ho-jae accounts for 5.53 of the team’s 30.08 xG (18.4%) and 8 of their 24 goals (33%). His per-90 profile is unmatched in the squad — 82.9% of his shots were taken inside the box, and he contested 150 aerial duels and won 51% of them, at 191cm.

A share that size says the player was very good. It also says the club never spread that share around.

06Not a replacement — a redesign

what Pohang actually did

The front three Pohang have started in the seven matches since the transfer.

Four different combinations in seven matches, settling from 25 August on Won Ki-jong, Cho Sang-hyuk and Hwang Seo-woong. Hwang has spent most of the season as a midfielder. Pohang are not replacing Lee Ho-jae; they are changing the shape of the front line.

Put the candidates side by side on a per-90 basis and you can see why the answer is hard to find.

On profile the closest to Lee Ho-jae are An Jae-jun and Jorge. In the seven matches since the transfer, An has one start and 50 minutes; Jorge has no starts and 187 minutes off the bench. The most-used combination instead pairs Tranziska, who has the lowest share of shots in the box (59.3%), with a converted midfielder.

07The tools are unchanged. There is nothing to aim at.

the clearest sign of a side out of tune

The crosses went up. The aerial duels fell by nearly a third. The 191cm target man is gone and the balls still go in — more of them, in fact. They are being delivered to nobody.

And this is where the real argument of the piece arrives.

Even when Lee Ho-jae was there, Pohang never turned him into a structure.

• 109 corners — 8th in the league — have produced one goal. With a 191cm target man in the side, all season (5 May away at Gangwon: Ki Sung-yueng’s corner headed in by Cho Sang-hyuk). In the other direction, seven of the 31 they have conceded came from set pieces, and headers run 3 scored to 9 conceded.
• Crosses run at 4.30 a match, 9th in the league (completed). That is not the number of a side built around a target man.
• His eight goals look less like the output of a team pattern and more like the output of one player standing in the right place.

So what Pohang faced when he left was not “irreplaceable” but “never defined in the first place.” Nothing had been built as a pattern, so there was nothing to hand over. What remains is the habit of crossing.

08How far has the redesign got?

four matches of searching, three of settling

The defence has finished resetting. xGA is down from 1.27 to 0.87 — better than the 1.03 of the pre-transfer team. Shots on target are back up from 1.5 to 4.3.

The attacking redesign has not. xG per match has gone 0.85 to 0.88, which is nowhere, and the share of shots taken in the box has fallen further, from 62% to 54%. They shoot more, and from further out.

Those three settled matches read Jeju 2-1 (W), Gangwon 0-0, Jeonbuk 0-2 — four points. That is why. The stopping came back. The scoring did not.

09And yet they are 7th

On the evidence so far this reads like a relegation side. Pohang are 7th. That is in the data too.

• In the 11 matches where they have led, they average 2.73 points (91% of what was available). Once ahead, they rarely let go.
• Five 1-0 wins. 21 of 27 matches have been decided by a single goal or fewer. This is a side that manages tight games.
• 1.15 conceded, 7th, with clearances 4th, blocks 5th and aerial duels 5th. They do not win the ball high (tackles 10th, interceptions 10th, fouls 11th) but the mid/low block works.

Pohang hold on to what they score. There is exactly one problem, and it is how often they score.

10Four matches

the same team, four different shapes

The argument reduces to two axes — how much they get forward (final third passes per match) and how good the shot is once they are there (xG per shot). Plot all 27 matches on those two and the season fits on one chart.

11Where the risk sits

• 15 goals conceded between the 60th and 90th minute — 48% of the 31 total (9 between 60 and 75, 6 between 75 and 90). They have scored 7 in the same window.
• The goalkeeper change: Hong Seong-min, 9 matches, 9.20 xGA and 12 conceded (+2.80), save rate 65.7%. Hwang In-jae, 18 matches, 19.56 xGA and 19 conceded (−0.56), save rate 67.2%. There is a direction here, but it is nine matches. It will not carry a verdict.

12What the data does not say

1. Why the selections were made. Nothing here explains why An Jae-jun and Jorge have played so little. Fitness, injury and tactical fit are outside the portal.
2. Pressure on the shot. There is no defender-distance data. “They shoot from further out” is established by coordinates; “they shoot under pressure” is inferred from block rates, not measured.
3. Variables that changed together. The transfer, the goalkeeper change and a run of home matches all fall in the same August. Naming one as the cause is where correlation and causation get mixed. What this piece claims is only that the shape of the result — entries held, shot positions retreating — fits the absence of a reference point in the box better than anything else.
4. Sample. Seven matches since the transfer, three since the line-up settled. Every interval here is wide.

13Predictions that can be checked

rounds 28–33

Six matches left — 9 Sep Gimcheon (H) · 12 Sep Daejeon (A) · 20 Sep Seoul (H) · 9 Oct Incheon (A) · 18 Oct Anyang (H) · 24 Oct Ulsan (A).

1. From the first match in which the share of shots taken in the box recovers to 65% or higher, Pohang’s xG per match clears 1.2. If the share stays below 60%, xG stays below 1.0.
2. Six to nine goals across the six matches. The low end if the settled-period rate (0.67 a game) holds; the high end if the finishing regresses to the mean.
3. At least one of the three home matches produces two or more goals (the block-1 home drought was always a number due to come back).
4. Set-piece goals: nought or one. That one does not regress, because there is no design behind it.
5. xGA stays at or below 1.0. The defensive reset is already done.

14Playing against Pohang / Pohang’s to-do list

If you are playing them

• Do not concede first. A Pohang side in front takes 91% of the available points.
• Do not empty the box — but the angles to defend are in front of it. Half of Pohang’s shots now come from out there.
• Give up corners freely. One goal from 109. Conversely your set pieces will work — Pohang have conceded seven from dead balls and nine to headers.
• Go at them after the hour. Half of everything they concede arrives there.

Pohang’s to-do list

• Short term — redefine who occupies the box. One player restored, or two rotating through it, either works. What cannot keep changing is the answer to “who is in there”, and right now it changes every time the front three does.
• Medium term — set pieces. 109 corners is 8th in the league for volume. The volume is there; the design is not. Three or four goals a season from that is worth about five points, and it is the one scoring route that does not depend on a particular individual.
• Structural — the final third bottleneck. Middle-third passes 3rd, defensive-third passes 4th, and attacking-third passes 10th (−21.9% against the league average). The ball circulates well. Moving it from there into the box is the real bottleneck — and that was true while Lee Ho-jae was still here.

That one transfer could be an event of this size is itself the name of the problem Pohang have to solve.