Agent Lab — a live forecast exhibit · Applied AI
← All fixtures
Group A · Matchday 14 · 2026-06-24

South Africa vs South Korea

Guadalupe, Mexico · neutral venue

Most likely scoreline
0–0 16.4%
24.2%
RSA win
30.0%
Draw
45.8%
KOR win
Top scorelines
  • 0–0 16.4%
  • 0–1 16.1%
  • 1–0 10.9%
  • 1–1 10.7%
Total goals
0 to 10 goals · most likely → tallest

The agent’s call

3-agent ensemble · deepseek-chat
RSA win
21.7%
Draw
29.3%
KOR win
49.0%

The model already favours South Korea (45% away win), and the supplied intel reinforces that bias. South Korea's recent form (WLLWW) is markedly better than South Africa's (LDLDD). The key player analysis shows a clear talent gap: Son Heung-Min and other Europe-based attackers outclass South Africa's domestic-based core. Recent news of South Korea's warm-up win adds confidence. I increase South Korea's probability from 45% to 50%, lowering South Africa from 25% to 20%, keeping draw at 30% as a plausible middle ground. South Korea's expected goals (1.21) and South Africa's (0.77) suggest a low-scoring match with South Korea likely to score once or twice. The most probable scorelines are 0-1 (14%), 0-2 (12%), and 1-1 (11%). South Africa's defensive record (LDLDD) and lack of attacking firepower make a clean sheet for South Korea plausible.

What the agent weighed
What the agent checked on the web
Searched tavily for RSA · KOR — 6 sources kept
Agent scorelines
generated 2026-06-11
Last 5 — RSA
-0.92 pts/match vs model expectation · opp. avg #83
  • PAN #74 -0.7
  • PAN #74 -1.7
  • NIC #151 -1.5
  • JAM #94 -0.5
  • MEX #21 -0.7
GF / GA last 10: 12 / 13
Last 5 — KOR
+0.53 pts/match vs model expectation · opp. avg #75
  • CIV #32 -1.3
  • AUT #22 -1.0
  • TRI #141 +0.5
  • EL #139 +0.4
  • CZE #40 +1.6
GF / GA last 10: 15 / 13
Head-to-head (last 5)

No prior meetings in the dataset.

▸ Full score grid (7×7)
P(home goals × away goals), %
0
1
2
3
4
5
6
0
16.4
16.1
9.6
4.4
1.7
0.6
0.2
1
10.9
10.7
6.3
2.9
1.2
0.4
0.1
2
4.3
4.3
2.5
1.2
0.5
0.2
0.1
3
1.3
1.3
0.8
0.4
0.1
0.1
0.0
4
0.4
0.3
0.2
0.1
0.0
0.0
0.0
5
0.1
0.1
0.0
0.0
0.0
0.0
0.0
6
0.0
0.0
0.0
0.0
0.0
0.0
0.0

Rows = RSA goals; columns = KOR goals. Above-diagonal cells are home wins; diagonal = draws; below = away wins. Truncated at 6 — the joint mass past that is < 1 %.

Expected goals (Poisson rate, λ)
0.76
RSA
1.23
KOR

Independent-Poisson rates. W/D/L and the score grid integrate the full 13×13 joint under these two parameters.

News context — injuries, suspensions, projected XI — is not part of the model, which sees only past results.