Agent Lab — a live forecast exhibit · Applied AI
← Schedule
Group G · Matchday 5 · 2026-06-15

Iran vs New Zealand

Inglewood, United States · neutral venue

Full time IRN 2–2 NZL
The agent called Iran (62%) — Worse than the model call wrong

The forecast below is what the agent published before kickoff.

Most likely scoreline
1–0 17.9%
56.4%
IRN win
27.4%
Draw
16.2%
NZL win
Top scorelines
  • 1–0 17.9%
  • 0–0 15.7%
  • 2–0 12.3%
  • 1–1 9.4%
Total goals
0 to 10 goals · most likely → tallest

The agent’s call

3-agent ensemble · deepseek-chat
IRN win
62.3%
Draw
24.0%
NZL win
13.7%

The model's prior (57% home win) is already high, and the supplied intel reinforces Iran's advantage. Iran's recent form (+0.62 pts/match vs expectation) against weaker sides shows they are in good rhythm, while New Zealand's form is poor (-0.16) against stronger opposition. Iran's key players, especially Taremi and Jahanbakhsh, provide more attacking depth than NZL's Chris Wood-led attack. The head-to-head record is sparse but favours Iran. No injuries or extreme conditions alter the baseline. I nudge the home win probability up slightly to 62% to reflect these factors, with draw and away win reduced accordingly. Iran's expected goals of 1.44 vs NZL's 0.54 suggest a low-scoring affair. Iran's defence is solid, and NZL's attack relies heavily on Wood. The most likely scorelines are 1-0 (18%) and 2-0 (14%), with a non-negligible chance of 1-1 (10%) if NZL can capitalize on a set piece. The model's top scorelines are retained with minor adjustments.

What the agent weighed
What the agent chose to investigate
The agent decided what to check on tavily — 9 sources kept:
  • “Iran vs New Zealand 2026-06-16 lineup injuries” match · 3 hits
    Late fitness or lineup changes could shift the forecast.
  • “Iran squad news June 2026” IRN · 3 hits
    Key player availability or suspensions may affect team strength.
  • “New Zealand squad news June 2026” NZL · 3 hits
    Key player availability or suspensions may affect team strength.
Agent scorelines
generated 2026-06-15
Last 5 — IRN
+0.64 pts/match vs model expectation · opp. avg #59
  • UZB #54 -0.6
  • NIG #40 -1.4
  • COS #68 +1.2
  • GAM #88 +0.8
  • MAL #46 +1.3
GF / GA last 10: 17 / 6
Last 5 — NZL
-0.17 pts/match vs model expectation · opp. avg #43
  • ECU #17 -0.6
  • FIN #70 -1.3
  • CHI #28 +2.2
  • HAI #97 -1.3
  • ENG #4 -0.2
GF / GA last 10: 7 / 18
Head-to-head (last 5)
  • 2003-10-12 Iran 3–0 New Zealand
  • 1973-08-12 New Zealand 0–0 Iran
▸ Full score grid (7×7)
P(home goals × away goals), %
0
1
2
3
4
5
6
0
15.7
8.2
2.6
0.6
0.1
0.0
0.0
1
17.9
9.4
3.0
0.7
0.2
0.0
0.0
2
12.3
6.4
2.0
0.5
0.1
0.0
0.0
3
6.6
3.4
1.1
0.3
0.1
0.0
0.0
4
3.0
1.6
0.5
0.1
0.0
0.0
0.0
5
1.2
0.7
0.2
0.1
0.0
0.0
0.0
6
0.5
0.3
0.1
0.0
0.0
0.0
0.0

Rows = IRN goals; columns = NZL 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)
1.49
IRN
0.59
NZL

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.