Match Analysis · K League 2 2026

From Cross to Shot
Where Suwon Samsung loses value

Suwon Samsung lead K League 2 and have conceded fewer goals than any other side. Yet they have scored the fewest of the top six. Using 20 matches of data against 96 clubs from Europe's Big 5 leagues, this report narrows the gap to a single phase.

Hyuckjin Han · Tactical AnalysisAug 2026K League Data Portal (Bepro11), 20 matchesFBref / Opta, Big 5 leagues 2025-26, 96 clubsFIFA TSG

The conclusion first. Suwon take slightly more shots per match than the European average, and their conversion after the shot — shots on target, goals — is almost identical to Europe. The only phase where they differ is cross → shot.

Crosses / match
27.7
Europe 17.6 · 1.58×
Shots / match
13.6
Europe 12.6 · 1.08×
Crosses ÷ shots
2.04
Europe 1.40 · 1.46×
Shots on target
34.9%
Europe 34.3% · identical

00Scope of the data used in this report

Every figure that follows comes from exactly these four populations. Stating the boundaries of the sample first is a precondition for reading the rest.

PopulationClubsTeam-matchesCrossesShotsOn targetGoalsPeriod · notes
Big 5 (Europe)963,50261,48144,03415,1044,7132025-26 completed season · 38 rounds ENG/ESP/ITA/GER, 34 FRA (2 clubs 33)
K League 1122625,3722,9871,0083072026 season in progress · rounds 1–23 (Gangwon & Incheon 21 matches, the other ten 22)
K League 2173366,9383,9721,4084662026 season in progress · rounds 1–21 (13 clubs 20 matches, 4 clubs 19)
Suwon Samsung120554272952928 Feb – 7 Aug 2026 · round 12 was a bye (odd number of clubs)

All figures retrieved 2026-08-13. K League figures come from the Korea Professional Football League data portal (supplied by Bepro11); European figures are FBref (Opta) club season totals. The portal's xG excludes penalty-kick goals and opponent own goals. Because K League 2 has 17 clubs — an odd number — one club rests each round, so match counts differ by up to one.

01Locating the leak

Splitting cross-to-goal into three conversion steps makes it immediately clear where the value drains away. Suwon are effectively identical to the European average in the second and third steps. Only the first step is 22.5pp lower.

Conversion funnel — cross → shot → shot on target → goal

Each bar is the conversion rate against the previous step. Sample boundaries as in the table above.

cross → shotshot → on targeton target → goal
Only the cross-to-shot rate is low. Suwon 49.1% / K League 2 57.2% / K League 1 55.6% / Europe 71.6%. The two steps after the shot sit at 33–35% and 30–34% for all four populations — effectively the same.
What this narrows downEverything after the shot is created — shot accuracy, finishing — is already at European level. The thing to fix is confined to one phase: crosses that do not become shots.

02What this ratio means across 96 European clubs

Across the 96 Big 5 clubs, how many crosses a team puts in has no relationship to how much it scores (r = +0.03). The rate at which crosses become shots, by contrast, is strongly linked to scoring at r = −0.68.

How to read rr expresses, on a scale from −1 to +1, the degree to which two quantities move together. +1 means one rises exactly as the other rises; 0 means no relationship at all; −1 means one rises exactly as the other falls. Conventionally |r| around 0.1 is read as weak, around 0.3 as moderate, and 0.5 or above as strong. r says nothing about cause — which way the causation runs has to be judged by observation and design, not by r.

Crosses ÷ shots vs goals per match — 96 Big 5 clubs

Further right means more crosses spent per shot created. Hover a point for that club's figures.

96 Big 5 clubsreference clubs (labelled)Suwon Samsung
The trend line is an ordinary least-squares fit over 96 clubs. Suwon's 2.04 would be the highest of all 96, and the clubs in that range are mostly relegation candidates. At the other end (0.79–0.87) sit Barcelona, Bayern, PSG and Real Madrid.

Split into quartiles, there is no exception

The 96 clubs divided into four groups by crosses ÷ shots, with each group's mean goals per match

The quartile that crosses the most (19.3 per match) scores the least (1.05 goals). The decline is monotonic, with no exception. Suwon's 2.04 sits above even the bottom quartile's mean of 1.75.

03So does this mean crossing less — no

Clubs that cross a lot and convert well do exist. Inter Milan average 23.0 crosses, 17.1 shots and 2.24 goals per match. Their crossing volume is close to Suwon's; their conversion ratio is 1.35.

Shot output among clubs with comparable crossing volume

Crosses and shots per match. At the same crossing volume, shot output diverges sharply.

crosses / matchshots / match■ highlighted band = Suwon Samsung
Athletic Club and Newcastle cross heavily with ratios close to Suwon's (1.77–1.78). Inter and Lens turn the same heavy volume into 1.35–1.38. What separates them is conversion, not volume.
The target pointMoving Suwon halfway to the European ratio on both terms gives 23.3 crosses and 16.7 shots per match. That is almost exactly Inter Milan's measured profile (23.0 · 17.1). In other words, the target is reachable while keeping a wide-attacking identity intact.

04What decides conversion — Cross Freedom

So what separates the conversion rates? Combining 20 matches of data with detailed observation of two, it resolves into three elements. They do not add — they multiply. If one is zero, the whole is zero. Alongside them sits one piece of context that stays outside the formula: T (time).

Z × F × A — the three elements of cross quality, plus the context T

What each element looks like when it approaches zero, and what was observed at Suwon

FIFA's TSG reports that at the 2022 World Cup cross types shifted — driven crosses −47%, push crosses +38%, cut-back goals 2→7: a move from speed toward precision and timing.

Z — Zone: freedom and value pull against each other

Deep by the touchline the opponent has little reason to close you down, so freedom comes easily but the angle and distance make the ball cheap. In the zone adjacent to the half-space the opponent must close you down, so freedom is hard to win — but cut-backs, in-swingers and low passes all become available. FIFA describes the cut-back as “the hardest pattern for a defence to deal with.”

F — Freedom: pressure, time, options

Building on FIFA Football Language's three classes — No pressureIndirectDirect — we record pressure at the moment of impact (F1), time in possession of the ball (F2) and whether options other than the cross exist (F3), each scored 0–2.

A — Arrival: number of bodies and concentration

The number of players entering the box (N) is the necessary condition; how many of them are within 5m of the landing point (Nc) comes next, and concentration is D = Nc ÷ N. Because Nc can never exceed N, raising concentration also requires bodies. That said, FIFA reports that both 2022 World Cup finalists put fewer players in the box than the tournament average (3.3) while converting above average — there is a threshold on numbers, and above it only concentration and positioning remain.

T — Time: context, not a formula in development

T is not a fourth term multiplied alongside Z, F and A. It is treated as the context that governs how much of those three is actually realised. At the same zone (Z), the same pressure (F) and the same arrival (A), a ball that gets there one beat earlier gives the crosser time to choose and gives the players in the box time to complete their movement. The transfer of the crosser's time into the box is what T is about.

There is more than one way to produce it — a three-player connection between winger, half-space and full-back that pins the opponent on one side so the ball can be delivered a beat earlier from the other; overload → isolation, creating the 1v1 so the ball is released before contact; rapid switches that arrive before the defence is set; and a low, fast pass weight on the natural foot that cuts the ball's flight time. These are not mutually exclusive, and which combination fits Suwon is still in development.

What cannot yet be claimed about T

  1. Whether it is an independent term or part of F is unresolved. T overlaps substantially with F2 (time in possession). Until the overlap is separated, counting both is double counting.
  2. There is no measurement standard. “One beat earlier” is currently an observer's judgement; neither a threshold in seconds nor an event definition has been fixed. Quantifying it at this stage would give different values for different observers.
  3. The sample is two matches. It was pronounced in only one of them. No effect size for T can be asserted from that.
Until these three are resolved, T stays an item to record and observe. It is not promoted to a metric.

05Observation — same opponent, home and away

The two matches against Paju — at home (R18) and away (R2) — produced almost the same number of crosses, 35 and 37. The quality of the chances created diverged sharply.

Expected goals (xG) per shot — R18 home vs Paju (0-0) vs R2 away at Paju (1-0)

Each dot is one shot. Further right means a higher probability of scoring.

Away, three chances above 0.20 xG were created; at home, one. Conversely, shots at or below 0.05 xG numbered 8 at home (50% of all shots) and 4 away (31%). With the same number of crosses, xG per shot differed by a factor of 1.9 — 0.076 against 0.141.
A line from the observation notes“Even when the 1v1 is won and the cross goes in, its accuracy is bound to drop — because the point it is delivered from is too far from the half-space.” — As each match wore on, support around the crossing point thinned, and freedom that had been produced by structure was replaced by freedom dependent on an individual beating a man.

06Target figures

Metrics already at European level are marked explicitly as hold. Separating what to change from what to leave alone is the first condition of a plan.

MetricCurrentFirst targetFinal targetBasis
Crosses ÷ shots2.041.751.40K League 2 mean → European mean
Crosses / match27.725.023.3Inter Milan profile
Shots / match13.614.516.7Inter Milan profile
Crosses per goal19.116.013.0European mean
Shots on target34.9%Hold — already at the European averageEurope 34.3%

07Limitations

To be read alongside everything aboveThe definition of a cross may differ. European data is Opta; K League data is Bepro11. Three things mitigate this — ① K League 1 (1.80) and K League 2 (1.75) are internally consistent with each other. ② Suwon's 2.04 is 17% above the K League 2 mean under the same definition. ③ Recomputing with corner kicks excluded gives Suwon 1.63 against a K League 2 mean of 1.38 — the same 1.18× multiple.