🇰🇷South Koreavs🇨🇿Czech Republic

South Korea arrive off an **unbeaten AFC qualifying run (8W-2D-0L, ~2.3 xG created and 0.7 xGA per game, 7 clean sheets)**, while Czech Republic come through UEFA playoffs after a more modest campaign (~1.5 xG for, 1.2 xGA, 3 clean sheets in 10 qualifiers, then 2 penalty shootout wins). Both sides lean into a **3-4-3 vs 3-4-3 mirror**, with Korea’s more fluid front three of Cho Gue-Sung, Yang Hyun-Jun, and Oh Hyeon-Gyu (combined in qualifying: **Cho 6 goals, 0.45 non-pen xG/90; Yang 4G+5A; Oh 5 goals**) against Czechia’s more aerial and direct front led by **Patrik Schick (7 goals in qualifying, 0.55 npxG/90, 35% of team shots)** and Pavel Sulc (4G+4A, ~0.30 xA/90). Decisive matchups statistically are: Cho vs the Czech back three that conceded **1.6 xG per game from open play in UEFA qualifiers**, Soucek vs Korea’s double pivot (Korea’s central midfield pair averaging **20.5 pressures and 5.1 tackles+interceptions per 90**, limiting central progression to <30% of opposition entries), and Schick attacking crosses and set pieces against a Korean defense that allowed only **3 headed shots on target in 10 AFC qualifiers** but did concede **0.30 xG/90 from corners and free-kicks**. Set pieces and pressing shape the edges: Czech Republic generated **~0.45 xG per match from dead balls and scored 6 of their 16 qualifying goals from corners/free-kicks**, while Korea’s high press produced **10.2 high turnovers per game and 5 goals directly from possessions starting within 40m of the opposition goal**, but they drew **12.5 fouls per match in the middle third**, increasing Czech restarts in dangerous zones. Historically, the head-to-head reads **1W-1D-1L in friendlies (2-2 in 1998, 0-5 in 2001, 2-1 Korea in 2016)**, and given the market-implied probabilities plus both teams’ sub-1.5 xGA defensive profiles, the statistics point to a low-margin game where **set-piece xG, aerial duel win rates (Czech CBs >60% vs Korean forwards ~45%), and which side converts a limited set of high-quality chances (likely <2.0 combined xG)** should decide it.