Agent Lab — a live forecast exhibit · Applied AI
← Schedule
Group F · Matchday 4 · 2026-06-14

Sweden vs Tunisia

Guadalupe, Mexico · neutral venue

Full time SWE 5–1 TUN
The agent called Sweden (50%) — Better than the model call right

The forecast below is what the agent published before kickoff.

Most likely scoreline
1–0 13.1%
46.9%
SWE win
26.5%
Draw
26.6%
TUN win
Top scorelines
  • 1–0 13.1%
  • 0–0 11.8%
  • 1–1 10.5%
  • 0–1 9.5%
Total goals
0 to 10 goals · most likely → tallest

The agent’s call

3-agent ensemble · deepseek-chat
SWE win
50.0%
Draw
27.3%
TUN win
22.7%

The model's prior (41% Sweden win) already reflects Sweden's edge, but the key-player gap is stark: Sweden's attacking duo of Gyökeres and Isak are world-class, while Tunisia's key players are at mid-tier clubs. Tunisia's recent form is worse, and they under-perform expectations more severely than Sweden. I raise Sweden's win probability to 50% and lower Tunisia's to 23% to account for this quality differential. The draw stays near the model's level (27%) as World Cup openers can be tight. Sweden's expected goals (1.17) and Tunisia's (0.92) suggest a low-scoring match, but Sweden's attacking quality pushes the expected total slightly higher. The most likely scoreline is 1-0 (14%), followed by 2-0 (12%) as Sweden's defense (not explicitly weak) should contain Tunisia's modest attack. A 1-1 draw (11%) is plausible if Tunisia's organized defense holds, but 0-0 (10%) is less likely given Sweden's firepower.

What the agent weighed
What the agent chose to investigate
The agent decided what to check on tavily — 9 sources kept:
  • “Sweden vs Tunisia 2026-06-15 lineup injuries” match · 3 hits
    Late fitness or lineup changes could significantly alter match outcome.
  • “Sweden squad 2026 World Cup injuries suspensions” SWE · 3 hits
    Key player absences would shift Sweden's strength.
  • “Tunisia squad 2026 World Cup injuries suspensions” TUN · 3 hits
    Key player absences would shift Tunisia's strength.
Agent scorelines
generated 2026-06-14
Last 5 — SWE
-0.10 pts/match vs model expectation · opp. avg #33
  • SLO #50 -0.9
  • UKR #27 +1.6
  • POL #32 +1.2
  • NOR #19 -1.0
  • GRE #36 -0.7
GF / GA last 10: 13 / 20
Last 5 — TUN
-0.24 pts/match vs model expectation · opp. avg #43
  • MAL #45 -0.2
  • HAI #96 +1.0
  • CAN #41 +0.0
  • AUT #23 -0.8
  • BEL #10 -0.4
GF / GA last 10: 13 / 14
Head-to-head (last 5)
  • 2003-02-12 Tunisia 1–0 Sweden
  • 1999-02-10 Tunisia 0–1 Sweden
  • 1992-04-22 Tunisia 0–1 Sweden
  • 1976-02-28 Tunisia 1–1 Sweden
▸ Full score grid (7×7)
P(home goals × away goals), %
0
1
2
3
4
5
6
0
11.8
9.5
4.6
1.7
0.6
0.2
0.0
1
13.1
10.5
5.1
1.9
0.6
0.2
0.0
2
8.8
7.1
3.4
1.3
0.4
0.1
0.0
3
4.6
3.7
1.8
0.7
0.2
0.1
0.0
4
2.1
1.6
0.8
0.3
0.1
0.0
0.0
5
0.8
0.7
0.3
0.1
0.0
0.0
0.0
6
0.3
0.2
0.1
0.0
0.0
0.0
0.0

Rows = SWE goals; columns = TUN 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)
1.44
SWE
0.96
TUN

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.