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

Netherlands vs Japan

Arlington, United States · neutral venue

Full time NED 2–2 JPN
The agent called Netherlands (40%) — Better than the model call wrong

The forecast below is what the agent published before kickoff.

Most likely scoreline
1–0 12.0%
45.7%
NED win
25.7%
Draw
28.6%
JPN win
Top scorelines
  • 1–0 12.0%
  • 0–0 10.6%
  • 1–1 10.5%
  • 0–1 9.2%
Total goals
0 to 10 goals · most likely → tallest

The agent’s call

3-agent ensemble · deepseek-chat
NED win
40.0%
Draw
28.0%
JPN win
32.0%

The model's prior favours Netherlands (45%) but Japan's recent form is far superior (+1.28 pts/match overperformance vs -0.42 for Netherlands) and they have been beating stronger opposition. However, Japan's key absences (captain Endo retired, star Mitoma injured) significantly weaken their squad, tilting the balance back toward Netherlands. The head-to-head is dated and not decisive. I adjust the prior down for Netherlands and up for Japan, settling on a tighter contest: home win 40%, draw 28%, away win 32%. Both teams have attacking quality but Japan's missing creators (Mitoma, Endo) may reduce their scoring threat. Netherlands' defence led by van Dijk is solid. The most likely scoreline is 1-1 (12%) reflecting a close match, with 1-0 (10%) and 0-1 (9%) also plausible. Low-scoring draws and narrow margins dominate.

What the agent weighed
What the agent chose to investigate
The agent decided what to check on tavily — 9 sources kept:
  • “Netherlands vs Japan 2026-06-14 team news injuries suspensions” match · 3 hits
    Late fitness updates or suspensions could shift the match forecast.
  • “Netherlands squad 2026 June injuries” NED · 3 hits
    Key Dutch players' availability is unknown and could affect the forecast.
  • “Japan squad 2026 June injuries” JPN · 3 hits
    Key Japanese players' availability is unknown and could affect the forecast.
Agent scorelines
generated 2026-06-14
Last 5 — NED
-0.42 pts/match vs model expectation · opp. avg #51
  • LIT #138 +0.1
  • NOR #19 +1.0
  • ECU #17 -0.8
  • ALG #29 -2.2
  • UZB #54 +0.6
GF / GA last 10: 22 / 8
Last 5 — JPN
+1.28 pts/match vs model expectation · opp. avg #47
  • GHA #59 +0.8
  • BOL #74 +0.6
  • SCO #39 +1.5
  • ENG #4 +2.4
  • ICE #61 +0.7
GF / GA last 10: 19 / 6
Head-to-head (last 5)
  • 2013-11-16 Netherlands 2–2 Japan
  • 2010-06-19 Netherlands 1–0 Japan
  • 2009-09-05 Netherlands 3–0 Japan
▸ Full score grid (7×7)
P(home goals × away goals), %
0
1
2
3
4
5
6
0
10.6
9.2
4.8
2.0
0.7
0.2
0.1
1
12.0
10.5
5.5
2.2
0.8
0.2
0.1
2
8.2
7.1
3.7
1.5
0.5
0.2
0.0
3
4.3
3.8
2.0
0.8
0.3
0.1
0.0
4
2.0
1.7
0.9
0.4
0.1
0.0
0.0
5
0.8
0.7
0.4
0.2
0.1
0.0
0.0
6
0.3
0.3
0.1
0.1
0.0
0.0
0.0

Rows = NED goals; columns = JPN 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)
1.47
NED
1.06
JPN

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.