Suwon FC have scored more goals than anyone in K League 2. Yet they reach the opponent’s final third less often than most — 13th of 17. What happens next is different: no team turns those entries into shots more often. I put all 212 league matches on one scale and followed that single line — when it works, and when it stops.
Suwon FC scored 50 goals in 25 matches, the most of the 17 K League 2 clubs. They are 2nd, 5 points behind the leaders, Suwon Samsung.
Rate every match on the expected goals (xG) of each shot and the order changes. Their non-penalty expected goal difference per match is +0.41 — 4th in the league, behind Suwon Samsung (+0.63), Seoul E-Land (+0.47) and Daegu (+0.43).
Replay every shot 20,000 times at its xG and Suwon FC’s expected points come to 41.7; they have 47. +5.3 points — the third-largest overperformance in the league. Their goals also run +6.7 above the 41.3 xG of their 324 shots, the largest excess in the league.
So is it luck? Re-take the same shots at their xG 100,000 times and 48 or more goals come up 12% of the time. Too large to wave away as luck, too small to call skill. So I changed the question — where is this team different?
| Club | Points | Expected | Diff. |
|---|---|---|---|
| Suwon Samsung | 53 | 46.1 | +6.9 |
| Seoul E-Land | 48 | 43.3 | +4.7 |
| Suwon FC | 47 | 41.7 | +5.3 |
| Daegu | 46 | 43.2 | +2.8 |
| Hwaseong | 43 | 39.7 | +3.3 |
| Busan | 41 | 34.1 | +6.9 |
The top scorers in a league usually live in the opponent’s half. Suwon FC do the opposite.
Build-up, progression, entry — all three below the league average. Yet this team scored the most. Only one place is left where the difference can come from: what happens after the threshold.
I counted, for all 17 teams in the same way, how often an attack that reached the final third ended in a shot.
Where you draw the “final third” line could change the ranking. I drew it in three places and counted again (open-play chains only).
| Final-third line | Entries/90 | Entry rank | Entry → shot | Shot rank | League avg. |
|---|---|---|---|---|---|
| x ≥ 60 | 34.1 | 13th | 25.3% | 1st | 21.4% |
| x ≥ 66.7 (default) | 28.1 | 13th | 30.5% | 1st | 25.4% |
| x ≥ 75 | 23.0 | 12th | 36.2% | 1st | 31.0% |
Apply the same yardstick to the attacks Suwon FC allow and the same shape appears, as if in a mirror.
Final-third entries run 28.9 to 31.9 — opponents get in more often. Yet expected goals run 1.43 to 1.01 in Suwon FC’s favour. Of the shots they allow, 42.3% come from outside the box, 3rd highest in the league, and the xG they allow inside the box, 0.82 per match, is 3rd lowest.
Attack and defence reduce to the same sentence. They give up space, and win beyond the threshold.
One reason they shoot so often once in is speed. No team generates more xG from shots within ten seconds of winning the ball.
| Time to first shot after regain | Suwon FC xG/match | League rank |
|---|---|---|
| Within 10 s | 0.325 | 1st |
| Within 15 s | 0.405 | 3rd |
| Within 20 s | 0.460 | 4th |
Widen the window to 15 or 20 seconds and they drop to 3rd and 4th. Suwon FC’s weapon is not the counter-attack but the very fast counter-attack. Of their 7.8 ten-second transition xG, 3.69 (47%) comes from Ha Jeong-woo alone, and he also starts more of those chains than anyone (10).
I split the same phase metrics by score state. Of 2441 minutes, Suwon FC spent 1058 (43%) in front — 1st in the league.
Brackets show the league rank: attack from highest, allowed from lowest.
| Metric | Leading | Level | Trailing |
|---|---|---|---|
| Suwon FC minutes | 1058 | 919 | 464 |
| Attack · final-third entries/90 | 26.7 (8) | 30.4 (8) | 31.1 (17) |
| Attack · entry → shot | 32.8% (2) | 33.1% (1) | 21.2% (15) |
| Attack · 10 s transition xG/90 | 0.37 (3) | 0.32 (1) | 0.05 (16) |
| Attack · open-play xG/90 | 1.25 (2) | 1.07 (2) | 0.99 (8) |
| Allowed · xG per entering chain | 0.018 (1) | 0.027 (11) | 0.043 (15) |
| Allowed · open-play xG/90 | 0.65 (1) | 0.83 (9) | 1.08 (11) |
Entry → shot 33.1%. Leaving penalty metrics aside, no phase-by-state value in this piece sits further from the league average (z = +2.5). Ten-second transition xG is 1st too. Meanwhile opponents cannot get out: their build-up reaches midfield 55.7% of the time, 2nd lowest in the league.
In front, they allow 0.65 open-play xG per 90, and an opponent chain that enters their third is worth 0.018 xG. Both are the lowest of the 17. Nor are they sitting back: their own open-play xG in the same minutes is 1.25 per 90, 2nd in the league.
Behind, final-third entries rank 17th, chains started by a regain 17th, ten-second transition xG 16th. Teams that trail usually push in more often (league average 36.8 entries per 90). Suwon FC rise to 31.1, but that is the fewest of the 17, and shot conversion after entry is 15th. One reading is that an opponent who sits deep leaves fewer balls to win and less space to run into — the event data does not show that directly.
Yet in matches where they conceded first, Suwon FC took 1.60 points per game, 2nd in the league. How, with every weapon switched off?
Of the 10 goals they scored while trailing, 3 were penalties and 1 came from set pieces. At the other end, opponents in those minutes created 1.08 open-play xG per 90 and scored only 0.58.
The other end looks the same. Suwon FC allowed 31 set-piece shots worth 2.57 xG and conceded 6 — also the largest excess in the league (3.4% chance by luck alone).
Long spells in front require scoring first. Suwon FC scored first in 58% of the matches that had a goal (5th in the league). I split their goals into 15-minute blocks.
The same shape repeats inside those minutes. Final-third entries per 90 rank 15th in the first half and 16th in the second, yet expected goals rank 2nd in both. In contrast, their xG in minutes 16–30 of the first half, 0.47 per 90, is the lowest in the league. A surge straight after kick-off, then a lull.
Suwon FC’s last defeat came on 23 May, R13 away at Gyeongnam. Since then: 13 matches, 7 wins, 6 draws, no losses.
| R1–13 | R14–27 | |
|---|---|---|
| Matches · W-D-L | 12 · 6-3-3 | 13 · 7-6-0 |
| Points / match | 1.75 | 2.08 |
| Scored : conceded | 24 : 18 | 26 : 11 |
| Final-third entries / match (events: 12 · 12 matches) | 31.8 | 31.8 |
| Entry → shot | 24.6% | 36.5% |
| Non-penalty xG : xGA / match | 1.31 : 1.07 | 1.54 : 0.95 |
| 10 s transition xG / match | 0.25 | 0.40 |
Entries stayed at 31.8 to 31.8 per match. What changed was the conversion to shots (match-level Mann–Whitney P = 0.004). During the season this team did not go in more often; it shot more often from the same number of entries.
Opponent strength in the second half is not adjusted for. An easier run of fixtures is not ruled out by this table.
The matches were picked by rules set in advance, not by impression — ① the match closest to the season average ② the match where this piece’s formula (reach less, shoot more) showed most strongly ③ the match where the result most outran the performance ④ the match where the performance most outran the result.
Six metrics sit closest to the season average (z-distance 1.22). An opponent own goal in first-half stoppage time put them ahead; goals in the 52nd and 55th minutes turned it round. The equaliser came from Frizzo’s penalty four minutes into second-half stoppage time. A lead, a reversal, and an equaliser at the moment the ball stopped — the arc of this piece in one match.
Only 21 final-third entries — equal to their season low. 42.9% of them ended in a shot. The opponents got in 40 times and shot from 15.0%. Frizzo in the 44th minute, Ha Jeong-woo in the 88th.
Three points from 1.19 expected (+1.81, the season’s largest). 43% possession and 24 entries to 42, but 45.8% of entries ended in a shot. Both goals were headers — Jo Jin-woo from a 10th-minute corner, Matheus Babi on 31 minutes.
63% possession and 40 entries, among their most. But only 17.5% of entries ended in a shot. 2.41 expected points became zero (−2.41, the season’s biggest shortfall). Their only goal was a Frizzo penalty in the 53rd minute. Reach more, shoot less, lose — the exact opposite of ②.
Hypotheses and readings from this analysis that the data did not support stay in, undeleted. All of them were mine. A rejected hypothesis draws the boundary of the conclusions that survived.
Revised. Ten-second transition xG is 1st, but at 15 and 20 seconds they are 3rd and 4th. Goals from moves that began with a ball win are 47.9% of their total, below the league average (53.7%). They have the fastest counter-attack, not the most counter-attacking goals.
Rejected. Trailing, they rank 17th for final-third entries and 17th for chains started by a regain. Points won from behind came from dead balls (4 of 10 goals while trailing) and opponents’ poor finishing.
Rejected. Substitutes scored 5 of 49 goals (10.2%), below the league average of 22.9%. Correction — my first count was 11. I missed that on goal rows of the live feed, the “sub in” field holds the assisting player.
Rejected. Regains in the opponent’s half (x ≥ 60) come 6.75 times per match, 14th in the league; their PPDA of 5.80 (higher = less pressing) is 4th highest. They win the ball in their own half (77% of regains).
Unresolved. 48 goals from 324 shots worth 41.3 xG (+6.7) is the league’s biggest excess, but chance alone produces it 12% of the time. The largest single piece of the excess is set pieces (+4.6).
The diagnosis narrows to one line — they reach less, shoot the most often, and close up when ahead. Behind, the formula stops, and the moment the ball stops becomes the lifeline.
Whether this shape is the head coach’s design (event data records outcomes, not intent) · whether the set-piece and penalty excess is skill (small samples, found by scanning many metrics) · whether the unbeaten run owes to the fixture list (opponent strength not adjusted) · what happened in R25 away at Yongin (no event data on the portal).