They Get In and Shoot —
and So Does Everyone Else

Daegu FC have 13 wins, 7 draws and 6 defeats from 26 matches — 46 points and 4th place. Count the shots, though, and the story changes.

14.3 shots and 1.84 expected goals (xG) per match — both 1st in the league. On the volume of chances created, no one in K League 2 is ahead of Daegu.

But they give back almost as much. They allow 12.4 shots per match — the 5th most in the league — and 1.38 xG, the 7th most. Subtract what they concede and what is left, non-penalty expected goal difference per match, is +0.43, 3rd in the league.

It is not finishing, good or bad. Their 371 shots were worth 47.9 xG and produced 47 goals (−0.9). Replay those shots 100,000 times at their xG and you get at least that many goals 59% of the time. Split by score state, though, one cell on the scoring side falls short: when leading, Daegu turned 18.5 xG into only 13 goals — losing roughly what they gained when level (+4.0). We come back to that cell later.

I split attacking play into phases and put all 17 clubs on the same scale. Daegu stand out in one place — after they reach the final third.

They reach the final third 31.2 times per 90, 7th in the league — ordinary. This is not a team that gets in often. But the share of those entries that end in a shot is 29.4%, 2nd in the league.

To check whether it depends on where the “final third” line is drawn, I drew it in three places and counted again.

Wherever the line goes, they are 2nd or 3rd. And each entry is the biggest in the league: one chain that reaches the final third is worth 0.040 xG, 1st; open-play xG per 90 is 1.28, also 1st.

How do they finish so often? Daegu attacks are short. The share of ball-winning chains that reach a shot within 10 seconds is 4.6%, 1st in the league, and they average 5.5 passes from regain to first shot, the 6th fewest. From wide areas they go straight in — 21.8 crosses per match (2nd) and 12.1 box entries (3rd).

Now the same scale for the opponent: when a team facing Daegu reaches the final third, how often does that move end in a shot?

The good news first. Opponents rarely get into Daegu’s third: 28.5 entries allowed per 90, the 5th fewest in the league. The problem is what happens next.

The moves that get in are big, too. Opponents’ xG per entry chain is 0.034, the 2nd most in the league, and Daegu allow 7.4 shots inside the box per match, the 6th most. Dead balls point the same way. Set-piece xG allowed is 0.12 per match, the 3rd fewest — yet they have conceded 6 set-piece goals (xG 3.0). At expected rates, conceding that many or more happens 7% of the time.

Put the two rates in one picture and Daegu’s place becomes clear. Across: how often they shoot once in. Up: how often they get shot at once opponents are in.

This is what a Daegu match is. Whoever has the ball, reaching the final third means a shot. Shots per match in Daegu games, both teams combined, come to 26.7 — 1st in the league.

At the entrance, though, it is different. Daegu keep opponents out (entries allowed the 5th fewest) and get in an ordinary number of times themselves (7th). Daegu win the fight up to the door; at the door, both sides shoot. Daegu’s edge in expected goal difference comes not from the shot rate but from the number of entries and the size of each move.

Split the same metrics by score state and you can see when both boxes run hottest. The answer: when Daegu are leading.

Ahead, Daegu do not stop. Open-play xG while leading is 1.72 per 90, 1st in the league; xG created within 10 seconds of a regain, 0.45, also 1st. The shot rate of their entries rises to 34.1% when leading — 1st again.

The price arrives in the same minutes. When Daegu lead, the share of opponent regains that reach a shot within 10 seconds is 4.3% — 1st in the league. The moment they lose the ball, the space behind opens.

This resolves the odd cell from the first section. In 814 minutes ahead, Daegu turned 18.5 xG into 13 goals and conceded 11 from 14.6 xGA. It is the stretch in which they create most and concede plenty — yet goals went in below expectation. They still held on to points: in the 17 matches in which they led at any point, they took 2.47 points per match (3rd in the league), dropping 9 points after leading.

When level they are at their most efficient: open-play goals 1.32 per 90 and +0.29 over expectation — both 1st. When trailing they shut the door. Opponents reach the final third 22.9 times per 90, the 3rd fewest, and Daegu’s penalty xG is 2nd.

I cut each match into six 15-minute segments and compared all 17 clubs. Daegu’s boxes heat up as the second half goes on.

It is not only goals. Non-penalty xG created in the first and second 15 minutes after the break is 2.02 and 2.20 per 90 — both 1st. And the conceding side grows as well: xG allowed in the last segment (76th minute to the end) is 1.50 per 90, the highest of any segment in Daegu matches. Both ends heat up together.

Of the 16 goals by substitutes, 8 belong to one man. Edgar came off the bench in all 13 of his appearances, played 505 minutes and scored 8.

All 8 came after the 66th minute, and 6 were headers. Daegu deliver the second-most crosses in the league and win aerial duels at the 2nd-best rate. When the ball arrived in the air into a box that had heated up after the hour, Edgar was at the end of it.

So did Edgar make Daegu a team of the last half hour? I compared only the same period (after the 60th minute).

The volume is about the same without him. What Edgar changed was the finish: his 22 shots were worth 3.2 xG and 8 went in (+4.8). Header-heavy scoring at that rate is hard to sustain. The lower xG conceded with him on may also reflect that he often came on with Daegu ahead (50% of his minutes were while leading).

If Edgar is the end of the box, who opens the entrance? I counted who passes to whom, and who takes part in moves that end in a shot.

How did the team look without them? Substitutes tend to enter at particular scores, so I removed the team’s average in each score state and kept only the difference that remains.

Serafim handles both the entrance and the end. He is joint top scorer (8 goals), his 10-second counter shots carry the most xG in the squad (1.8), and he starts the most of those counters with a regain (8). Jaewon Hwang carries it to the threshold, Serafim takes it over, Edgar finishes it.

Using the player pages on the official K League website, I checked the nationality of all 563 players who have appeared in K League 2. Daegu used 7 foreign players — Serafim (Brazil, 2320 min) · Cesinha (Brazil, 1263 min) · Dackers (England, 1042 min) · Edgar (Brazil, 510 min) · Danrlei (Brazil, 327 min) · Geovani (Brazil, 9 min) · Carlos (Brazil, 8 min).

Metrics where Daegu sit at or near either end of the 17-club table and which did not make it into the story above.

The matches were chosen by rules fixed in advance — ① the match closest to the season average ② the match where the title’s shape (both sides shoot once in) showed most strongly ③ the match where the result most exceeded the performance ④ the match where the performance most exceeded the result.

Hypotheses I formed during the analysis that the data did not support stay on the page rather than being deleted. All of them are mine.

The diagnosis narrows to one line — in Daegu matches, the final third ends in a shot at both ends. Daegu win at the entrance, get hotter after the hour, and their foreign forwards stand at the end of it.