Agent Lab — a live forecast exhibit · Applied AI
Scorecard · the agent

Where the agent breaks from the model

A small LLM ensemble reads each fixture and emits its own W/D/L. It is handed the statistical model as its prior, the betting markets, and structured intel — recent form, head-to-head, key players, injuries, match-day conditions, and fresh news it pulls from a live web search per team. It treats the model as an informed prior, not a target: it holds where the facts are neutral but commits to the divergence the facts justify — and names the fact either way. The per-fixture call sits inline on the schedule and each group page; this is the cross-tournament cut — its biggest divergences, and how they scored.

Agent calls
72
Calls graded
16
6 right · 10 wrong
Outright flips
4
agent's favourite ≠ model's
Ensemble
3 agents
deepseek-chat

Upcoming · 56

Calls on matches still to play, soonest kickoff first.

France v Senegal
Δ 11.4 pts
FRA
Draw
SEN
Model
50.3%
28.6%
21.1%
Agent
61.7%
23.7%
14.7%

The model's prior (50% France win) is a reasonable starting point, but the squad market value gap (France 2nd, Senegal 12th) and recent form (France in line, Senegal underperforming) indicate France's talent edge is stronger than the model captures. Key players like Mbappé and Olise give France a clear attacking advantage, while Senegal's key attacker Mané is past his peak. No injuries or conditions offset this. I raise France's win probability to 60%, lowering draw and away win accordingly. France's expected goals (1.36) and Senegal's (0.72) suggest a low-scoring match, but France's superior attack should generate multiple chances. The most likely scorelines are 1-0 and 2-0, with 2-1 also plausible if Senegal's defense cracks. A draw (0-0 or 1-1) is possible but less likely given France's form and talent edge.

Group I· Matchday 6 · ·8 factors ·9 web sources ·called 2026-06-16
Iraq v Norway
Δ 8.4 pts
IRQ
Draw
NOR
Model
16.9%
25.1%
58.0%
Agent
13.0%
20.7%
66.3%

The statistical model already favours Norway (58% away win), and the squad market value gap (Norway ~28.8x more valuable) reinforces that Norway's talent is far superior. Iraq's recent form is mediocre against weak opposition, while Norway has over-performed against strong opponents. Key player quality heavily tilts toward Norway with Haaland. I increase Norway's win probability to 68% to reflect the talent disparity not fully captured by the model, while keeping a modest draw chance (20%) as Iraq may defend deep in a World Cup debut. Norway's expected goals (1.66) and Haaland's presence point to a likely 1-0 or 2-0 win. Iraq's low xG (0.68) and defensive setup suggest they may struggle to score, but a 0-0 or 1-1 draw is possible if Norway is wasteful. The most probable scorelines are 0-1, 0-2, and 0-3, with a non-negligible chance of 1-1.

Group I· Matchday 6 · ·8 factors ·9 web sources ·called 2026-06-16
Argentina v Algeria
Δ 3.5 pts
ARG
Draw
ALG
Model
62.5%
23.0%
14.2%
Agent
66.0%
21.7%
12.3%

The statistical model already reflects Argentina's strong form and talent advantage, and the supplied intel—squad value gap, key player quality (Messi, Lautaro), and Argentina's perfect recent run—all point in the same direction. Algeria's over-performance against stronger opponents is a mild note but insufficient to close the talent chasm. I hold the prior essentially unchanged, nudging home win slightly from 63% to 65% to account for the squad value and key player edge not fully captured by the model's results-based rating. Argentina's expected goals (1.84) and Algeria's (0.63) from the model suggest a low-scoring affair with Argentina controlling. The most likely scorelines are 1-0, 2-0, and 0-0, reflecting Argentina's defensive solidity and Algeria's likely defensive posture. A 1-1 draw is possible if Algeria capitalizes on a set piece or counter, but Argentina's quality should see them through.

Group J· Matchday 6 · ·8 factors ·9 web sources ·called 2026-06-16
Austria v Jordan
Δ 10.8 pts
AUT
Draw
JOR
Model
54.9%
25.2%
19.8%
Agent
65.7%
21.7%
12.7%

The prior model already favours Austria (55% win), and the supplied intel reinforces that: Austria's recent form is excellent against weaker opposition, while Jordan is in poor form against stronger teams. The massive squad value gap (19.6x) and superior key players (Arnautovic, Sabitzer, Alaba) further tilt the match toward Austria. The only counterweight is a minor midfield injury (Baumgartner out, Ljubicic in), but it is not enough to close the gap. I raised Austria's win probability to 65% to reflect the cumulative edge from form, squad value, and key-player quality. Austria's xG of 1.62 and Jordan's 0.78 suggest a low-scoring match with Austria controlling play. The most likely scorelines are 1-0 and 2-0, consistent with a solid but not dominant Austrian performance. Jordan's poor attack (0.78 xG) makes a draw less likely, but a 1-1 or 0-0 cannot be ruled out given Austria's occasional defensive lapses.

Group J· Matchday 6 · ·8 factors ·3 web sources ·called 2026-06-16
Portugal v DR Congo
Δ 9.0 pts
POR
Draw
COD
Model
65.0%
22.5%
12.3%
Agent
74.0%
17.0%
9.0%

The statistical model already gives Portugal a strong 65% win probability, but the supplied intel points to an even larger gap. Portugal's squad is vastly more valuable (94th vs 32nd percentile), their key players are in excellent form, and they have over-performed expectations recently. DR Congo, meanwhile, faced disruption from an Ebola-related isolation order and cancelled training camp, which likely hampers preparation. The climate-controlled dome neutralizes any weather factor. I raise Portugal's win probability to 72%, lowering draw and away win accordingly. Portugal's expected goals (1.88) and DR Congo's low xG (0.56) suggest a relatively low-scoring match with Portugal controlling. The most likely scorelines are 2-0 and 1-0, reflecting Portugal's defensive solidity and DR Congo's limited attack. A 3-0 is possible if Portugal's stars click, while a 1-1 draw is less likely but not negligible given DR Congo's occasional counter threat from Wissa.

Group K· Matchday 7 · ·8 factors ·9 web sources ·called 2026-06-16
England v Croatia
Δ 5.7 pts
ENG
Draw
CRO
Model
48.3%
28.8%
22.9%
Agent
54.0%
26.3%
19.7%

The model gives England a 48% win probability, but the squad value gap (England #1 vs Croatia #16, ~3.5x more valuable) and superior key-player quality push England's chances higher. Croatia's recent form is slightly better but against stronger opponents, and head-to-head is mixed. The climate-controlled roof neutralizes conditions. I raise England's win probability to 55%, reducing draw and away win accordingly. England's expected goals from the model (1.33) and Croatia's (0.77) suggest a low-scoring match. The most likely scoreline is 1-0 (14%), followed by 2-0 (12%) and 1-1 (11%). The 0-0 draw (10%) and 0-1 away win (8%) are less likely due to England's attacking talent.

Group L· Matchday 7 · ·8 factors ·9 web sources ·called 2026-06-16
Ghana v Panama
flips the pick Δ 6.5 pts
GHA
Draw
PAN
Model
40.8%
28.5%
30.6%
Agent
34.3%
30.0%
35.7%

The model prior favours Ghana (41%) but the critical news of Thomas Partey's visa denial removes Ghana's midfield anchor, significantly weakening their spine. Ghana's recent form is poor (LLLLD) against strong opponents, while Panama is in decent form (DWLWD) against weaker sides. The squad value gap (5.5x) is real but Partey's absence and Ghana's form slump narrow it. I shift the prior toward a more even match, slightly favouring Panama due to the Partey blow and Ghana's underperformance. Outcome probabilities: Ghana 35%, draw 30%, Panama 35%. The model's expected goals (Ghana 1.22, Panama 0.99) suggest a low-scoring affair, but Partey's absence likely reduces Ghana's attacking output and defensive solidity. Panama's recent form suggests they can score. Most likely scorelines are 1-1, 0-0, 1-0, and 0-1, with a slight tilt toward draws and narrow margins.

Group L· Matchday 7 · ·8 factors ·8 web sources ·called 2026-06-16
Uzbekistan v Colombia
Δ 2.3 pts
UZB
Draw
COL
Model
15.0%
26.9%
58.0%
Agent
13.3%
26.3%
60.3%

The statistical model already gives Colombia a clear edge (58% win probability), and all supplied intel reinforces that lean without creating a reason to diverge. Colombia's superior recent form (vs stronger opponents), much higher squad market value (5.2x), and key players at elite clubs contrast sharply with Uzbekistan's underperformance and weaker talent base. The high altitude of Mexico City may further benefit Colombia, who are more accustomed to such conditions. Therefore, I keep the model's prior unchanged. The model expects a low-scoring match (Uzbekistan 0.56 xG, Colombia 1.53 xG), and the intel does not suggest a goal-fest. Colombia's defensive solidity and Uzbekistan's limited attack point to a 1-0 or 2-0 win for Colombia, with a non-negligible chance of a 0-0 draw if Uzbekistan defends deep. The top scorelines reflect this: 0-1, 0-0, 0-2.

Group K· Matchday 7 · ·8 factors ·9 web sources ·called 2026-06-16
Czech Republic v South Africa
Δ 8.1 pts
CZE
Draw
RSA
Model
45.2%
28.9%
25.8%
Agent
53.3%
27.7%
19.0%

The statistical model gives Czech Republic a 45% win probability, but the squad market value gap (3.2x) and recent form (South Africa underperforming by -0.94 pts/match) point to a stronger edge. Czech Republic's key players operate at higher club levels (Schick at Leverkusen, Soucek at West Ham) and have better recent ratings than South Africa's (e.g., Zwane at 5.2). The head-to-head is ancient and irrelevant. Conditions are neutral. I raise the home win probability to 55%, lowering away win to 18% and draw to 27%. Czech Republic's expected goals (1.27) and South Africa's (0.84) suggest a low-scoring match. The most likely scorelines are 1-0 and 2-0, reflecting Czech superiority but not a rout. 1-1 and 0-0 are plausible given South Africa's defensive potential. 0-1 is less likely due to Czech Republic's form advantage.

Group A· Matchday 8 · ·8 factors ·9 web sources ·called 2026-06-16
Switzerland v Bosnia and Herzegovina
Δ 2.1 pts
SUI
Draw
BIH
Model
58.4%
23.7%
17.8%
Agent
60.0%
24.3%
15.7%

The model's prior (58% home win) is supported by a clear squad value gap (~3x) and superior key-player depth, especially in attack with Embolo, Amdouni, and Xhaka. Bosnia's recent form is flat (five draws) and they rely on an aging Dzeko. No injuries or conditions create a reason to move away from the model; the only head-to-head is ancient. I nudge the home win slightly to 60% to reflect the talent edge, but keep the draw at 24% and away at 16% as the model's baseline is already reasonable. Switzerland's expected goals (1.77) suggest they score 1-2 goals; Bosnia's low xG (0.76) limits them to 0-1. The most likely scorelines mirror the model: 1-0 (14%), 2-0 (12%), with draws like 1-1 (10%) and 0-0 (9%) plausible given Bosnia's defensive resilience in recent draws.

Group B· Matchday 8 · ·8 factors ·9 web sources ·called 2026-06-16
Canada v Qatar
Δ 5.1 pts
CAN
Draw
QAT
Model
61.6%
20.6%
17.4%
Agent
66.7%
20.7%
12.7%

Canada are clearly the stronger side based on squad value (7.6x), recent form (in line vs underperforming), and head-to-head. The model's 62% home win is a reasonable baseline, but the squad gap and Qatar's poor form justify a slight increase to 65%. Davies' limited fitness is a minor concern but not enough to shift the balance significantly. Qatar's late draw with Switzerland shows resilience, but they remain heavy underdogs. Canada's expected goals (2.10) and Qatar's (0.88) suggest a low-scoring win for Canada. The most likely scorelines are 1-0 and 2-0, with 1-1 also possible if Qatar's defense holds. Canada's attack (David, Larin) should be enough to break through Qatar's defense, but clean sheets are likely given Qatar's offensive struggles.

Group B· Matchday 8 · ·8 factors ·9 web sources ·called 2026-06-16
Mexico v South Korea
Δ 3.1 pts
MEX
Draw
KOR
Model
47.6%
26.9%
25.4%
Agent
50.7%
27.0%
22.3%

The model's prior (48% home win) is already reasonable given Mexico's superior recent form and head-to-head record. Squad value and key player quality reinforce this edge, though recent news of a punishment and South Korea's strong win over Czech Republic introduce minor counterweights. I adjust home win probability slightly upward to 50% to reflect the consistent form and historical advantage, while keeping draw and away win near model levels. Mexico's expected goals (1.42) and South Korea's (0.91) suggest a low-scoring match. The most likely scorelines are 1-0 (14%) and 2-1 (12%), with a 1-1 draw (11%) also plausible given South Korea's recent scoring ability. The distribution remains tight around 2-3 total goals.

Group A· Matchday 8 · ·8 factors ·9 web sources ·called 2026-06-16
United States v Australia
Δ 4.0 pts
USA
Draw
AUS
Model
39.7%
27.3%
33.0%
Agent
43.0%
28.0%
29.0%

The model's prior (USA 41% win) is already reasonable given the key player quality gap (USA's stars at top European clubs vs Australia's mostly domestic or lower-tier players) and a slight head-to-head edge for USA. However, Australia's recent form is marginally better (2 wins vs USA's 2 wins in last 5), which tempers the home advantage slightly. No injuries or extreme conditions shift the balance. I keep the probabilities close to the model, nudging USA down a point and Australia up a point to reflect the form difference. Expected goals (USA 1.28, AUS 1.08) suggest a low-scoring affair. The most likely scorelines are 1-0, 1-1, and 0-0, consistent with the model. The key player quality gap may produce a narrow USA win, but Australia's form and defensive solidity (Souttar) keep it tight.

Group D· Matchday 9 · ·7 factors ·6 web sources ·called 2026-06-11
Scotland v Morocco
Δ 1.7 pts
SCO
Draw
MAR
Model
21.0%
31.3%
47.7%
Agent
22.7%
30.0%
47.3%

The model and all supplied intel consistently point to Morocco as the stronger side. Recent form (unbeaten in 5), a lopsided head-to-head victory, and a clear advantage in key player quality (Brahim Díaz, Hakimi, El Kaabi vs Scotland's mid-tier Premier League/Serie A players) justify a slight increase in Morocco's win probability from the model's 47% to 45%? Actually the model already had Morocco at 47%, and the news of potential Moroccan injuries (Mazraoui, Ezzalzouli) slightly tempers that, so I keep away win near the model at 45%. Scotland's home advantage is minimal in a neutral venue, and their recent form is inconsistent. Draw remains plausible given Scotland's defensive organization, but Morocco's attacking quality should prevail. Morocco's expected goals of 1.20 and Scotland's 0.66 suggest a low-scoring match with Morocco likely scoring once. The most probable exact scorelines are 0-1 (model: 18%) and 0-0 (18%), reflecting Morocco's defensive solidity and Scotland's limited attack. A 1-1 draw is also possible if Scotland capitalizes on a set piece. Higher-scoring outcomes like 0-2 or 1-2 are less likely but still in the distribution given Morocco's firepower.

Group C· Matchday 9 · ·7 factors ·6 web sources ·called 2026-06-11
Brazil v Haiti
Δ 5.1 pts
BRA
Draw
HAI
Model
82.5%
10.2%
3.8%
Agent
87.7%
8.7%
3.7%

The model already gives Brazil an 83% win probability, and the supplied intel strongly reinforces that. The head-to-head is lopsided (17-1 aggregate), Brazil's recent form is solid (3 wins in last 5), and the key players comparison shows a chasm in quality: Brazil's attackers play for Real Madrid, Barcelona, etc., while Haiti's best players are at Esteghlal and AEK Athens. Neymar's potential absence is the only mild uncertainty, but Brazil's depth (Vinícius, Raphinha, Paquetá) is still overwhelming. I have nudged Brazil's win probability up to 88% to reflect the extreme mismatch, but kept some draw/away probability for the slight Neymar doubt and the fact that Haiti is not completely hopeless (they have some attacking threat). Brazil's expected goals of 3.53 and Haiti's 0.41 from the model are reasonable. Given Brazil's attacking firepower and Haiti's defensive limitations (conceded 7 in last H2H), a 3-0 or 4-0 win is very plausible. The most likely scorelines reflect Brazil dominating possession and creating many chances, while Haiti struggles to score. The model's top scorelines (2-0, 3-0, 4-0) are kept but with slightly higher probabilities for 3-0 and 4-0 due to the quality gap.

Group C· Matchday 9 · ·7 factors ·6 web sources ·called 2026-06-11
Turkey v Paraguay
Δ 3.0 pts
TUR
Draw
PAR
Model
37.5%
28.7%
33.7%
Agent
39.3%
30.0%
30.7%

The model's near-even forecast (home 36%, draw 29%, away 35%) is a reasonable starting point, but the supplied intel shows a clear squad quality gap: Turkey's key players (Çalhanoglu, Güler, Aktürkoğlu) are at top European clubs with strong recent ratings, while Paraguay's key players (Almirón, Sanabria, Enciso) are at lower-tier clubs with weaker ratings. Turkey's recent form (WDWWW) is also slightly better than Paraguay's (LWWLW). News articles reinforce Turkey as dark horses and note Paraguay's long World Cup absence. I therefore increase Turkey's win probability to 40%, reduce draw to 30%, and away win to 30%. The model's low expected goals (Turkey 1.08, Paraguay 1.05) suggest a low-scoring match, with 0-0 and 1-0/0-1 as top scorelines. Given Turkey's slight edge in quality, 1-0 is the most likely single outcome, but draws (0-0, 1-1) also remain probable. I keep the scoreline distribution close to the model's, with a slight shift toward Turkey scoring.

Group D· Matchday 9 · ·7 factors ·6 web sources ·called 2026-06-11
Netherlands v Sweden
Δ 3.4 pts
NED
Draw
SWE
Model
51.6%
23.1%
25.1%
Agent
55.0%
23.3%
21.7%

The model's prior (53% home win) is a reasonable starting point. Head-to-head history and key player analysis both favour Netherlands, but the gap is not overwhelming. Sweden's attack is potent with Gyökeres and Isak, which prevents a large shift. I slightly increase home win probability to 55% and reduce away win to 22%, reflecting the mild edge from H2H and squad depth. The draw stays near the model's estimate. Netherlands are expected to control possession and create more chances, leading to a likely 1-0 or 2-0 win. However, Sweden's counter-attacking threat means a 1-1 draw or even a 0-0 is plausible. The most probable scorelines cluster around low-scoring outcomes because both defenses are solid and the match is a World Cup opener, often cautious.

Group F· Matchday 10 · ·7 factors ·6 web sources ·called 2026-06-11
Germany v Ivory Coast
Δ 6.0 pts
GER
Draw
CIV
Model
55.7%
23.9%
20.3%
Agent
61.7%
21.7%
16.7%

Germany's perfect recent form and superior key-player quality (Havertz, Wirtz, Kimmich at elite clubs vs Ivory Coast's mid-tier stars) justify a slight upward adjustment from the model's 56% to 60%. The head-to-head is ancient, conditions are neutral, and no injuries affect key personnel, so the move is modest. Ivory Coast's strong recent form (4 wins) prevents a larger shift. Germany's expected goals (1.76) and Ivory Coast's (0.83) suggest a low-to-moderate scoring match. The most likely scorelines reflect Germany dominance: 1-0, 2-0, 0-0, 1-1. The 0-0 draw retains some probability due to Ivory Coast's defensive potential, but Germany's attacking quality (Havertz, Wirtz) pushes toward at least one goal.

Group E· Matchday 10 · ·7 factors ·6 web sources ·called 2026-06-11
Ecuador v Curaçao
Δ 7.4 pts
ECU
Draw
CUW
Model
74.3%
18.2%
6.9%
Agent
81.7%
13.7%
4.7%

The model already heavily favours Ecuador (73%), and the supplied intel reinforces that. Curaçao's recent form is poor (1 win in 5), they are missing two key players (Gorre and St Jago), and their squad is composed of players from weaker clubs compared to Ecuador's stars (Caicedo, Estupiñán, Valencia). These factors justify raising Ecuador's win probability from 73% to 82%, while reducing the draw and away win chances. Ecuador's expected goals (2.12) and Curaçao's (0.39) suggest a low-scoring game for the underdog. With Curaçao's key absences, Ecuador should control the match and score multiple goals. The most likely scorelines are 2-0 and 3-0, with a non-negligible chance of 1-0 or 2-1. A clean sheet for Ecuador is likely given Curaçao's weak attack.

Group E· Matchday 10 · ·7 factors ·6 web sources ·called 2026-06-11
Tunisia v Japan
Δ 4.9 pts
TUN
Draw
JPN
Model
23.0%
28.8%
48.1%
Agent
19.3%
27.7%
53.0%

Japan's dominant recent form (5 wins), strong head-to-head record (4-1), superior key player quality and recent production, plus Tunisia's key injury to Ezzalzouli all point to a Japan win. The model already favours Japan (45% away win), but the intel strengthens that lean, so I increase Japan's win probability to 52% and reduce Tunisia's win and draw probabilities accordingly. Japan's attack is prolific (Ueda, Ito, Kamada, Doan, Ogawa) while Tunisia's defence is average. The most likely scoreline is 0-1, consistent with Japan's solid defence and Tunisia's limited scoring threat. Higher-scoring Japan wins (0-2, 1-2) are also plausible, while a draw or Tunisia win is less likely given the form and injury gap.

Group F· Matchday 10 · ·7 factors ·6 web sources ·called 2026-06-11
Spain v Saudi Arabia
Δ 6.9 pts
ESP
Draw
KSA
Model
73.7%
18.3%
7.5%
Agent
80.7%
13.7%
5.7%

The model already heavily favours Spain (75%), and every supplied factor reinforces that: Spain's recent form is solid (unbeaten in five), they have a perfect head-to-head record (3-0), their key players are elite European talent while Saudi Arabia's are domestic, and recent news shows Spain beat a stronger opponent in their warm-up. No injuries or conditions disrupt Spain's advantage. I raise Spain's win probability slightly from 75% to 80% to reflect the consistent strength of the evidence, but keep the move modest because the prior was already high and Saudi Arabia's recent win (over Puerto Rico) shows some resilience. Spain's expected goals (2.34) suggest 2-3 goals, while Saudi Arabia's (0.43) suggest 0-1. The most likely scorelines are 2-0 and 3-0, reflecting Spain's defensive solidity (clean sheets in 3 of last 5) and Saudi Arabia's low scoring threat. A 1-0 is also plausible if Spain underperform, but the historical 5-0 suggests potential for a bigger margin. I keep the scoreline distribution close to the model's top lines, slightly increasing 2-0 and 3-0 probabilities.

Group H· Matchday 11 · ·7 factors ·6 web sources ·called 2026-06-11
Belgium v Iran
Δ 4.7 pts
BEL
Draw
IRN
Model
52.0%
25.6%
22.4%
Agent
56.7%
25.3%
18.0%

Belgium's recent form is excellent and their key players (De Bruyne, Tielemans, Lukaku) are world-class, while Iran's best attacker Taremi lacks comparable support. The model's 51% home win is a reasonable baseline, but the squad quality gap justifies a modest upward adjustment to 55%. Draw and away win probabilities are slightly reduced accordingly. Belgium's expected goals (1.52) and Iran's (0.84) suggest a low-to-moderate scoring match. The most likely scorelines are 1-0 and 2-0, reflecting Belgium's defensive solidity and Iran's limited attacking threat. A 1-1 draw is possible but less likely given Belgium's form.

Group G· Matchday 11 · ·7 factors ·6 web sources ·called 2026-06-11
Uruguay v Cape Verde
Δ 12.7 pts
URU
Draw
CPV
Model
60.6%
26.7%
12.6%
Agent
73.3%
19.0%
7.7%

The statistical model already reflects a large quality gap, and the key-player analysis strongly reinforces it: Uruguay's core plays for elite European and South American clubs, while Cape Verde's key players are at mid-level clubs. Cape Verde's recent form (8 pts) is slightly better than Uruguay's (5 pts), and they are a motivated debutant, but the squad quality disparity is too wide. I have increased Uruguay's win probability slightly from 64% to 70% to reflect the player-level gap, while keeping a modest draw chance (22%) because Cape Verde's recent form suggests they can be competitive. Uruguay's expected goals (1.71) and Cape Verde's (0.46) point to a low-scoring match. The most likely scoreline is 1-0 (18%), followed by 2-0 (16%). Cape Verde's defense has been solid in recent matches, but Uruguay's attack should find the net. A clean sheet for Uruguay is probable given Cape Verde's limited attacking threat against a top defense.

Group H· Matchday 11 · ·7 factors ·6 web sources ·called 2026-06-11
New Zealand v Egypt
Δ 0.5 pts
NZL
Draw
EGY
Model
17.1%
30.7%
52.1%
Agent
17.7%
30.3%
52.0%

The model and all supplied intel converge on Egypt as clear favourites. Recent form (Egypt DWDWL vs NZL LLWLL), head-to-head (3-0-1 Egypt), and a major gap in key-player quality (Salah, Marmoush vs Wood) all point away. No countervailing factors (injuries, conditions, news) justify moving off the prior, so I keep Egypt's win probability near 52% and New Zealand's low at 18%. Egypt's expected goals (1.23) and New Zealand's (0.53) imply a low-scoring match with Egypt likely to score 1-2 goals and NZL struggling to score. The most probable exact scorelines are 0-1 (20%) and 0-0 (19%), consistent with the model's top two outcomes. The distribution is slightly shifted toward Egypt clean sheets given their defensive solidity (Rami Rabia high rating) and NZL's poor form.

Group G· Matchday 11 · ·7 factors ·6 web sources ·called 2026-06-11
Argentina v Austria
Δ 7.9 pts
ARG
Draw
AUT
Model
58.7%
25.3%
15.9%
Agent
66.7%
21.3%
12.0%

The model prior gives Argentina a 59% win probability, which is already high. The supplied intel reinforces this: Argentina's perfect recent form, lopsided head-to-head, and vastly superior key players (Messi, Lautaro, Álvarez vs Austria's less stellar squad) all point to a clear advantage. No injuries or conditions offset this. I increase home win probability to 65% and reduce draw/away accordingly, reflecting the strong squad gap and Argentina's status as defending champions. Argentina's expected goals (1.67) and Austria's low (0.63) suggest a low-scoring match for Austria. Argentina's attack is potent, but Austria's defence (led by Alaba) may keep it to 2-0 or 2-1. The top model scoreline (1-0) is plausible, but I slightly raise 2-0 and 2-1 probabilities given Argentina's attacking quality and Austria's ability to score (recent form WDWWW).

Group J· Matchday 12 · ·7 factors ·6 web sources ·called 2026-06-11
France v Iraq
Δ 12.5 pts
FRA
Draw
IRQ
Model
69.8%
20.7%
9.2%
Agent
82.3%
13.3%
4.3%

The model already heavily favours France (70% win, 2.05-0.45 xG), and the supplied intel only strengthens that case. France's recent form (4 wins in 5) and elite key players like Mbappé and Olise, who play at top clubs, dwarf Iraq's squad, which features lower-level players and a key forward (Aymen Hussein) who faced a 7-hour interrogation upon entering the USA, potentially affecting his focus. With no injuries and Dembélé returning, France's depth is even stronger. I moved France's win probability upward by 12 percentage points to 82%, reflecting the clear quality gap and the Hussein distraction. France's expected goals of 2.05 suggest they should score 2-3 goals against a weaker Iraq side. Iraq's xG of 0.45 indicates they are unlikely to score, though a consolation goal is possible. The most likely scorelines are 2-0 and 3-0, with 1-0 also plausible if France are wasteful. Clean sheet for France is probable given the disparity.

Group I· Matchday 12 · ·7 factors ·6 web sources ·called 2026-06-11
Norway v Senegal
Δ 3.2 pts
NOR
Draw
SEN
Model
36.8%
29.0%
34.1%
Agent
40.0%
28.0%
32.0%

The model and key-player analysis both point to a slight Norway advantage, driven by Erling Haaland's world-class output. No injuries, neutral conditions, and similar recent form leave the prior largely intact, but Haaland's presence justifies a modest upward adjustment to Norway's win probability from 37% to 40%, with a corresponding small drop in Senegal's chances. Expected goals (Norway 1.12, Senegal 1.04) suggest a low-scoring affair. The most likely scorelines reflect that: 1-0 Norway, 1-1 draw, 0-0, and 0-1 Senegal. Haaland's quality raises the chance of a 2-1 or 2-0 Norway win, but Senegal's defensive solidity keeps the match tight.

Group I· Matchday 12 · ·7 factors ·6 web sources ·called 2026-06-11
Jordan v Algeria
Δ 4.2 pts
JOR
Draw
ALG
Model
22.1%
25.0%
52.8%
Agent
18.7%
24.3%
57.0%

The model already favours Algeria (52% win), but the supplied intel strengthens that case: Algeria's recent form (WLWDW) is far superior to Jordan's (LDDLL), and Algeria's key players are at a higher club level (Mahrez, Amoura, Bensebaïni vs. Tamari, Olwan). The head-to-head record, though sparse, also favours Algeria. I raise Algeria's win probability to 58%, lowering Jordan's to 18% and the draw to 24%. Algeria's expected goals (1.61) and Jordan's (0.88) suggest a low-scoring match with Algeria likely scoring 1-2 goals. The most probable scorelines are 0-1 (14%) and 0-2 (12%), with 1-1 (10%) as the top draw outcome. Jordan's defensive struggles and Algeria's solid attack support a clean sheet for Algeria.

Group J· Matchday 12 · ·7 factors ·6 web sources ·called 2026-06-11
Portugal v Uzbekistan
Δ 7.4 pts
POR
Draw
UZB
Model
63.2%
23.5%
13.1%
Agent
70.7%
19.3%
10.0%

The model's prior (64% home win) already reflects Portugal's superiority, but the key players gap is even starker when comparing elite European-based stars to Uzbekistan's modest squad. Recent form (2.2 vs 1.0 pts/game) and the absence of any countervailing factors (no injuries, neutral conditions) justify raising Portugal's win probability to 70% and lowering the draw and away win. Uzbekistan's World Cup debut and lack of top-tier talent make an upset unlikely. Portugal's expected goals (1.82) and Uzbekistan's low xG (0.57) suggest a comfortable win with 2-0 or 1-0 most likely. The model's top scoreline 1-0 (17%) is plausible but I slightly favor 2-0 given Portugal's attacking quality. A 3-0 is possible but not dominant. Low-scoring draws (0-0, 1-1) are less likely due to Portugal's offensive edge.

Group K· Matchday 13 · ·7 factors ·6 web sources ·called 2026-06-11
England v Ghana
Δ 7.7 pts
ENG
Draw
GHA
Model
68.3%
21.5%
9.9%
Agent
76.0%
17.0%
7.0%

The model already gives England a 69% win probability, and the supplied intel strongly reinforces that. Ghana's dreadful form (0 wins in 5) and a massive squad quality gap — England's key players are all at elite clubs with high ratings, while Ghana's are at mid-tier clubs — push the probability even higher. The only minor counterweight is Saka's reported Achilles issue, but he is still expected to play. I raise England's win probability to 75%, lowering draw and away win accordingly. England's expected goals (1.97) and Ghana's (0.48) suggest a low-scoring match for Ghana. The most likely scorelines are 2-0 and 1-0, reflecting England's defensive solidity and Ghana's inability to score (recent form: 0 goals in last 5? Actually LLLLD implies they scored at least once in a draw, but still poor). The model's top scorelines (1-0, 2-0, 0-0, 3-0) are kept but with slight adjustments: 2-0 becomes most likely due to England's attacking quality against a weak opponent.

Group L· Matchday 13 · ·7 factors ·6 web sources ·called 2026-06-11
Panama v Croatia
Δ 0.5 pts
PAN
Draw
CRO
Model
14.2%
21.9%
63.7%
Agent
14.7%
22.0%
63.3%

The model already heavily favours Croatia (64% away win), and the key player analysis strongly supports that: Croatia's stars play at top European clubs while Panama's are mostly in lower-tier leagues. No injuries, no extreme conditions, no head-to-head history, and no relevant news alter the picture. The model's prior is well-justified, so I keep it essentially unchanged, with a tiny nudge toward Croatia due to the squad quality gap. The model expects Croatia to score around 1.95 goals and Panama 0.67, leading to low-scoring outcomes like 0-1, 0-2, or 0-0. Croatia's superior attack and Panama's limited firepower make a clean sheet for Croatia likely, while Panama's defence (with Murillo at Marseille) might keep it from being a rout. The most probable scorelines reflect a one- or two-goal Croatian win.

Group L· Matchday 13 · ·7 factors ·6 web sources ·called 2026-06-11
Colombia v DR Congo
Δ 6.2 pts
COL
Draw
COD
Model
59.8%
26.0%
14.1%
Agent
66.0%
22.0%
12.0%

The model already assigns Colombia a 61% win probability, and the supplied intel reinforces that. Colombia's key players (Díaz, Rodríguez) are at a higher club level and have superior recent output compared to DR Congo's attackers (Bakambu, Wissa). The altitude in Zapopan (1551 m) gives Colombia a slight additional edge. No injuries or head-to-head history alter the picture. I nudge the home win probability up to 65% to reflect the squad quality gap and altitude, keeping the draw and away win roughly in line with the model. The model's most likely scoreline is 1-0 (18%), reflecting Colombia's defensive solidity and DR Congo's low xG (0.53). I keep 1-0 as the top line, with 2-0 (14%) and 0-0 (13%) also plausible. The altitude and quality gap may increase Colombia's goal expectation slightly, but DR Congo's defensive organization (implied by their recent form) limits blowout potential. Scorelines with 3+ goals for Colombia are less likely given the model's low xG for both sides.

Group K· Matchday 13 · ·7 factors ·6 web sources ·called 2026-06-11
Bosnia and Herzegovina v Qatar
Δ 6.6 pts
BIH
Draw
QAT
Model
44.2%
26.1%
29.6%
Agent
50.0%
27.0%
23.0%

The model's prior (45% home) is already reasonable, but the key players factor pushes slightly upward: Bosnia's attack features Bundesliga-level talent (Dzeko, Demirovic, Tabakovic) while Qatar's key men play domestically. Recent form also slightly favours Bosnia (five draws vs Qatar's four losses in five). I adjust home win probability to 50%, draw to 27%, away to 23%, reflecting a moderate edge for Bosnia without overcommitting given the limited head-to-head and no major injury news. Bosnia's expected goals (1.45) and Qatar's (1.07) suggest a low-scoring match. The most likely scorelines are 1-0, 1-1, and 0-0, consistent with the model's top three. I keep the distribution similar to the model, as no factor strongly alters goal expectations.

Group B· Matchday 14 · ·7 factors ·6 web sources ·called 2026-06-11
Canada v Switzerland
Δ 1.0 pts
CAN
Draw
SUI
Model
30.3%
25.5%
44.1%
Agent
29.3%
26.3%
44.3%

The model's prior favours Switzerland (44% away win) and the supplied intel reinforces that lean. Key player analysis shows Switzerland's core (Xhaka, Ndoye, Embolo) outperforms Canada's in recent form, and the news that Alphonso Davies missed the opener is a significant blow to Canada's defence and attack. Recent form slightly favours Canada (2 wins vs 0) but Switzerland's draws include tougher opposition. Conditions are neutral. Thus I keep the model's away win probability at 44%, slightly increase draw to 26% (given Canada's resilience at home) and keep home win at 30%. The model expects low scoring (Canada 1.13, Switzerland 1.48 expected goals). Switzerland's superior key players and Canada's Davies absence suggest Switzerland is more likely to score, but Canada's home advantage and decent form keep it close. 0-1 is the top scoreline (11%), followed by 1-1 (10%) and 0-0 (10%). I keep the model's scoreline distribution as no strong reason to shift.

Group B· Matchday 14 · ·7 factors ·6 web sources ·called 2026-06-11
Morocco v Haiti
Δ 9.3 pts
MAR
Draw
HAI
Model
72.3%
19.2%
8.0%
Agent
81.7%
12.7%
5.7%

The model already assigns Morocco a high win probability (72%) based on expected goals (2.18 vs 0.45). The supplied intel strongly reinforces this: Morocco's recent form is solid (DWWWD), while Haiti's is inconsistent (WLDWL). The key player gap is enormous — Morocco's attackers play for Real Madrid, PSG, and PSV, while Haiti's top scorers are at Esteghlal, AEK Athens, and Toronto FC. No injuries or conditions alter the picture. I raised Morocco's win probability to 80% to reflect the squad quality chasm, which is not fully captured by the model's generic rating. Morocco's expected goals of 2.18 suggest they will score 2-3 goals. Haiti's expected goals of 0.45 indicate they will struggle to score. The most likely scorelines are 2-0 (18%), 3-0 (14%), and 1-0 (12%). A clean sheet for Morocco is probable given Haiti's weak attack against Morocco's strong defense (Hakimi at PSG).

Group C· Matchday 14 · ·7 factors ·6 web sources ·called 2026-06-11
Scotland v Brazil
Δ 2.8 pts
SCO
Draw
BRA
Model
13.2%
22.5%
64.2%
Agent
12.3%
20.7%
67.0%

The model already heavily favours Brazil (65% away win), and the supplied intel reinforces that lean without introducing any countervailing factor. Brazil's dominant head-to-head record (5 wins from 5), superior squad quality (Neymar, Vinícius, etc. vs. mid-tier Scotland players), and better recent form all point toward a comfortable Brazil win. No injuries or conditions mitigate this; Scotland's recent wins came against weaker opposition. Therefore I slightly increase Brazil's probability to 68%, trimming the draw and home win accordingly. Brazil's expected goals (1.95) and Scotland's low xG (0.60) suggest a low-scoring match for Scotland. Brazil's attack is potent but may not run riot; the most likely scorelines are 0-1, 0-2, and 0-3, consistent with the model's top outcomes. A draw (0-0 or 1-1) is possible but less likely given Brazil's offensive quality and Scotland's defensive limitations.

Group C· Matchday 14 · ·7 factors ·6 web sources ·called 2026-06-11
Mexico v Czech Republic
Δ 4.7 pts
MEX
Draw
CZE
Model
50.3%
25.3%
24.3%
Agent
55.0%
25.0%
20.0%

The model's prior (49% home win) already gives Mexico a slight edge, but the supplied intel reveals a key factor not fully priced in: the 2240m altitude of Estadio Azteca. Mexico is acclimated to high-altitude conditions, while Czech Republic is not, which historically provides a meaningful home advantage. I increase Mexico's win probability to 55%, drawing from draw (25%) and away (20%) to reflect this environmental edge. No other factor (form, head-to-head, injuries, key players, news) justifies a further move. The model expects a low-scoring match (Mexico 1.54 xG, Czech 0.95 xG). The altitude may suppress Czech attacking output further, making 1-0 the most likely scoreline. I keep the scoreline distribution close to the model's top lines, slightly shifting probability from 0-0 and 1-1 toward 1-0 and 2-0 to account for Mexico's altitude advantage.

Group A· Matchday 14 · ·7 factors ·6 web sources ·called 2026-06-11
South Africa v South Korea
Δ 3.2 pts
RSA
Draw
KOR
Model
24.2%
30.0%
45.8%
Agent
21.7%
29.3%
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.

Group A· Matchday 14 · ·7 factors ·6 web sources ·called 2026-06-11
Curaçao v Ivory Coast
Δ 3.8 pts
CUW
Draw
CIV
Model
10.0%
20.4%
69.2%
Agent
8.7%
18.3%
73.0%

The model already heavily favours Ivory Coast (66% away win), and the supplied intel reinforces that. Curaçao's recent form is poor (1 win in 5), they are missing key attacker Kenji Gorre, and their squad is composed of players from lower-tier clubs, while Ivory Coast boasts multiple top-level talents and recent strong form (4 wins in 5, including a draw with France). The absence of market lines leaves the model as the only prior, but the intel justifies a further shift: away win probability increased to 74%, with draws and home wins reduced accordingly. Ivory Coast's expected goals (1.87) and Curaçao's low xG (0.53) suggest a low-scoring game dominated by Ivory Coast. The most likely scorelines are 0-1, 0-2, and 0-3, consistent with the model but with slightly higher away win probabilities. Curaçao's defensive absences (St Jago) may allow Ivory Coast to score multiple goals, but the moderate wind and heat are not extreme enough to significantly alter scoring rates.

Group E· Matchday 15 · ·7 factors ·6 web sources ·called 2026-06-11
Ecuador v Germany
Δ 1.9 pts
ECU
Draw
GER
Model
26.9%
27.5%
45.5%
Agent
25.0%
27.7%
47.3%

The model already favours Germany (44% away win) and the supplied intel reinforces that lean: Germany's perfect recent form, dominant head-to-head record, and superior key players (Havertz, Wirtz, Kimmich) all point to a clear quality gap. Ecuador's form is solid but not enough to close the gap. I nudge the away probability slightly higher (to 47%) given the consistent strength of the intel, while keeping the draw at 28% and home at 25% as Ecuador can be resilient. Germany's attacking quality (Havertz, Wirtz) and Ecuador's decent defence (Estupiñán, Torres) suggest a 1-2 or 0-1 outcome is most likely. The model's top scorelines (0-1, 0-0, 1-1) are retained, but I add 1-2 and 0-2 to reflect Germany's higher expected goals (1.29 vs 0.95). Low-scoring match is expected due to Ecuador's conservative style and Germany's control.

Group E· Matchday 15 · ·7 factors ·6 web sources ·called 2026-06-11
Japan v Sweden
Δ 7.2 pts
JPN
Draw
SWE
Model
41.1%
26.4%
32.5%
Agent
48.3%
26.3%
25.3%

Japan's perfect recent form (WWWWW) and Sweden's inconsistent results (DWWLD) suggest Japan has momentum, but Sweden's superior individual talent (Gyökeres, Isak at top clubs) and a historical head-to-head unbeaten streak (Japan 0 wins in 4) create a counterweight. The model's prior (42% home) is slightly elevated given Sweden's quality edge, but Japan's form justifies keeping the home win probability above 45%. I nudge home win up to 48%, keep draw at model's 27%, and reduce away win to 25%. Both teams have attacking threats, but Japan's defensive solidity (recent clean sheets) and Sweden's reliance on individual brilliance suggest a low-scoring match. The most likely scorelines are 1-0 Japan, 1-1 draw, and 0-0, consistent with the model's expected goals (Japan 1.32, Sweden 1.07).

Group F· Matchday 15 · ·7 factors ·6 web sources ·called 2026-06-11
Tunisia v Netherlands
Δ 1.3 pts
TUN
Draw
NED
Model
17.3%
24.1%
58.5%
Agent
16.0%
25.3%
58.7%

The statistical model already favours the Netherlands (56% away win), and the supplied intel strengthens that lean. Tunisia's recent form is poor (DWDLL) and they lose key attacker Ezzalzouli to injury, while Netherlands bring a deep, high-quality squad with multiple elite attackers. The head-to-head is sparse and non-competitive, and conditions are neutral. I adjust the away probability upward to 60% to reflect the injury blow and form gap, slightly reducing home and draw probabilities. Netherlands are expected to dominate possession and create chances, with Tunisia likely defending deep. The most probable scoreline is 0-1 (15%), reflecting a narrow Dutch win, but 0-2 (12%) and 1-1 (11%) are also plausible. Tunisia's expected goals are low (0.70 model), so clean-sheet outcomes (0-0, 0-1, 0-2) dominate the top of the distribution.

Group F· Matchday 15 · ·7 factors ·6 web sources ·called 2026-06-11
Paraguay v Australia
Δ 2.8 pts
PAR
Draw
AUS
Model
30.6%
32.6%
36.8%
Agent
33.0%
33.0%
34.0%

The model and all supplied intel point to an extremely balanced match. Recent form is similar, head-to-head is ancient and irrelevant, key players are of comparable quality, no injuries are confirmed despite one news report, and conditions are neutral. Therefore I keep the model's near-even split unchanged. The model expects low scoring (xG 0.89 each), with 0-0 as the most likely single scoreline. No factor suggests a deviation from this low-goal expectation, so the scoreline distribution mirrors the model's output.

Group D· Matchday 15 · ·7 factors ·6 web sources ·called 2026-06-11
United States v Turkey
flips the pick Δ 10.1 pts
USA
Draw
TUR
Model
43.4%
23.7%
32.8%
Agent
33.3%
27.7%
39.0%

The model's prior slightly favours USA (42%) but recent form (Turkey 4W in 5, USA 2W) and head-to-head (Turkey won last meeting) point the other way. Key player ratings also favour Turkey, with Güler and Çalhanoglu outperforming USA's midfield. I shift probability from USA to Turkey, making Turkey the slight favourite (38%) and reducing USA to 35%, with draw at 27%. Both teams have decent attacking talent but Turkey's recent defensive solidity (3 clean sheets in last 5?) and USA's inconsistency suggest a low-to-moderate scoring match. The most likely scorelines are 1-1, 0-1, and 1-0, reflecting the tight expected goals (USA 1.52, Turkey 1.32).

Group D· Matchday 15 · ·7 factors ·6 web sources ·called 2026-06-11
Norway v France
Δ 1.4 pts
NOR
Draw
FRA
Model
22.7%
25.2%
52.0%
Agent
21.7%
25.0%
53.3%

The model already favours France (52%) and the supplied intel reinforces that: France's recent form (4 wins in 5) is clearly superior to Norway's (2 wins), and France's squad depth — with Mbappé, Dembélé, Rabiot, Olise, Thuram — outclasses Norway's attack-heavy lineup that relies heavily on Haaland. The head-to-head record also slightly favours France. No injuries or conditions alter the picture. I have nudged France's win probability up from 52% to 55% and Norway's down from 23% to 20% to reflect the clear quality gap and form advantage, while keeping the draw near the model's 25% as Norway can be compact. France's expected goals (1.60) and Norway's (0.89) from the model suggest a low-scoring game with France likely to score 1-2 goals and Norway 0-1. The most likely scorelines are 0-1, 1-1, and 0-2, consistent with the model's top outcomes. I have kept the distribution similar to the model, slightly increasing France's clean-sheet chances given their defensive solidity.

Group I· Matchday 16 · ·7 factors ·6 web sources ·called 2026-06-11
Senegal v Iraq
Δ 7.0 pts
SEN
Draw
IRQ
Model
53.0%
29.6%
17.4%
Agent
60.0%
25.0%
15.0%

Senegal’s squad is clearly superior, with elite players like Mané and Sarr against Iraq’s modest talent. Iraq’s top scorer Aymen Hussein faced a 7-hour interrogation upon entering the USA, a potential distraction. The model’s 53% home win is a reasonable prior, but the quality gap and news push it higher to 60%. Draw and away win probabilities are reduced accordingly. Senegal’s expected goals of 1.34 and Iraq’s 0.57 from the model reflect a low-scoring affair. With Senegal’s attack and Iraq’s likely defensive approach, 1-0 and 2-0 are the most probable scorelines. The chance of a clean sheet for Senegal is elevated given Iraq’s limited attacking threat.

Group I· Matchday 16 · ·7 factors ·6 web sources ·called 2026-06-11
Cape Verde v Saudi Arabia
Δ 0.9 pts
CPV
Draw
KSA
Model
30.5%
34.5%
35.1%
Agent
30.0%
34.0%
36.0%

The model's prior (30/34/36) already reflects a slight edge for Saudi Arabia, and the key player analysis supports that lean: Saudi Arabia's attackers play at a higher club level and have stronger recent ratings than Cape Verde's. However, Cape Verde's recent form is slightly better (two wins vs one) and they have debut motivation, but that is offset by inexperience. No other factors (conditions, head-to-head, availability) move the needle. Therefore I stay close to the model, with a modest tilt toward Saudi Arabia already baked in. The model expects low scoring (Cape Verde 0.81, Saudi Arabia 0.94 expected goals), consistent with both teams' defensive profiles and the high-stakes World Cup group stage. The most likely scorelines are 0-0 (20%), 0-1 (16%), and 1-0 (14%), reflecting the tight, cautious nature of the match. The key player advantage for Saudi Arabia slightly raises their chance of scoring, but overall goal expectation remains below 2.

Group H· Matchday 16 · ·7 factors ·6 web sources ·called 2026-06-11
Uruguay v Spain
Δ 0.3 pts
URU
Draw
ESP
Model
22.4%
28.4%
49.1%
Agent
22.7%
28.3%
49.0%

The model's prior (Spain 48%) aligns well with the supplied intel: Spain's dominant head-to-head record, slightly better recent form, and superior attacking depth (Yamal, Oyarzabal, Olmo) all point toward an away win. No strong countervailing factors (injuries, conditions) emerge to move the forecast significantly. I keep the prior nearly unchanged, nudging Spain down very slightly to 47% to reflect Uruguay's home-like advantage in the Americas and their own quality (Valverde, Núñez). The model's expected goals (Uruguay 0.78, Spain 1.35) suggest a low-scoring match, consistent with Spain's defensive solidity and Uruguay's pragmatic style. The most likely scorelines are 0-1, 0-0, and 1-1, with a slight lean toward Spain scoring exactly once. The altitude may suppress scoring slightly, but not enough to alter the distribution.

Group H· Matchday 16 · ·7 factors ·6 web sources ·called 2026-06-11
Egypt v Iran
Δ 2.0 pts
EGY
Draw
IRN
Model
32.0%
32.4%
35.6%
Agent
30.0%
33.0%
37.0%

The model prior already slightly favours Iran (36% away win, 0.96 xG) over Egypt (31% home win, 0.84 xG). Recent form reinforces that: Iran arrives on a three-match winning streak, while Egypt has been inconsistent (DWDWL). Key player analysis shows Iran's attacking core—Taremi, Jahanbakhsh, Mohebi—is in better recent form and collectively more productive than Egypt's, whose stars (Salah aside) play at lower club levels. No injuries, no extreme conditions, and no market to adjust from. I nudge the away win probability up one point to 37%, home down one to 30%, draw unchanged at 33%. The model's expected goals (Egypt 0.84, Iran 0.96) suggest a low-scoring affair. The most likely scorelines are 0-0, 0-1, and 1-0, reflecting tight defensive play typical of World Cup group openers. Iran's slight edge in attack makes 0-1 more probable than 1-0. I keep the model's scoreline distribution largely intact, with minor adjustments for Iran's better form and key-player advantage.

Group G· Matchday 16 · ·7 factors ·6 web sources ·called 2026-06-11
New Zealand v Belgium
Δ 0.6 pts
NZL
Draw
BEL
Model
8.5%
18.4%
72.5%
Agent
8.0%
19.0%
73.0%

The model already prices Belgium as heavy favourites (73%) and no supplied intel contradicts that. Belgium's recent form (WWDWW) and star-studded squad (De Bruyne, Lukaku, Tielemans) dwarf New Zealand's modest attack (Wood, Barbarouses) and poor form (LLWLL). The lack of head-to-head or injury news leaves the prior intact. I keep the model's probabilities unchanged. Belgium's expected goals (2.24) and New Zealand's (0.46) suggest a low-scoring match for the underdog. The most likely scorelines are 0-1, 0-2, 0-3, consistent with a Belgium clean sheet. New Zealand's best hope is a 0-0 draw (10%) or a 1-1 (7%) if they catch Belgium off guard.

Group G· Matchday 16 · ·7 factors ·6 web sources ·called 2026-06-11
Croatia v Ghana
Δ 6.2 pts
CRO
Draw
GHA
Model
57.2%
25.3%
17.5%
Agent
63.3%
23.0%
13.7%

Croatia's recent form (WWLLW) is far superior to Ghana's (LLLLD), and their key players are at stronger clubs (Modric, Perisic, Kramaric vs Ayew, Partey). The news of Ezzalzouli's injury further weakens Ghana's attacking depth. The model's 58% home win is a reasonable prior, but the gap in form and squad quality justifies a modest increase to 65% for Croatia, with draw and away win reduced accordingly. Croatia's expected goals (1.65) suggest they are likely to score 1-2 goals, while Ghana's low xG (0.68) points to a low-scoring output. The most probable scorelines reflect a Croatia win to nil (1-0, 2-0) or a narrow 1-1 draw. The 0-0 is less likely given Croatia's attacking quality but still possible if Ghana defend deep.

Group L· Matchday 17 · ·7 factors ·6 web sources ·called 2026-06-11
Panama v England
Δ 3.0 pts
PAN
Draw
ENG
Model
7.7%
17.6%
74.0%
Agent
7.0%
16.0%
77.0%

The model already heavily favours England (74% away win), and the supplied intel reinforces that. The head-to-head (6-1), massive gap in key-player quality (Kane vs Panama's attackers), and England's strong warm-up form all push the same direction. I increase England's win probability slightly from 74% to 77% and reduce draw/home chances accordingly, because the facts are uniformly pro-England and no countervailing factor exists. England's expected goals (2.33) and Panama's (0.47) from the model suggest a comfortable England win with limited Panama threat. The top model scorelines (0-1, 0-2, 0-3) are retained, with slight adjustments: 0-2 becomes slightly more likely given England's attacking quality and Panama's defensive record. Low-scoring draws (0-0, 1-1) are possible but unlikely given the gulf in class.

Group L· Matchday 17 · ·7 factors ·6 web sources ·called 2026-06-11
Colombia v Portugal
Δ 2.8 pts
COL
Draw
POR
Model
32.5%
28.3%
39.2%
Agent
30.0%
28.0%
42.0%

The statistical model already gives Portugal a slight edge (39% vs 33%). The key player analysis reinforces that edge: Portugal's stars consistently perform at higher club levels and with better recent ratings than Colombia's key players. Recent form is mixed but Portugal's is marginally more consistent. No other factors (injuries, conditions, news) move the needle. I therefore nudge the away win probability slightly higher to 42%, keeping the draw near the model's 28% and reducing Colombia's win chance to 30%. Both teams have moderate expected goals (COL 1.06, POR 1.18), suggesting a low-scoring affair. The most likely scorelines are 0-1, 1-1, and 0-2, consistent with Portugal's slight superiority and both sides' recent defensive solidity. The model's top scorelines are retained with minor adjustments to reflect the away lean.

Group K· Matchday 17 · ·7 factors ·6 web sources ·called 2026-06-11
DR Congo v Uzbekistan
flips the pick Δ 1.3 pts
COD
Draw
UZB
Model
31.5%
34.3%
34.3%
Agent
32.3%
34.7%
33.0%

The model and all supplied intel point to a highly balanced match with no clear advantage. Both teams have similar recent form, no head-to-head history, full availability, and comparable key-player quality. The climate-controlled venue neutralizes any weather edge. Therefore, I retain the model's near-even probabilities, with a slight tilt toward Uzbekistan due to Shomurodov's experience and Sergeev's strong recent rating, but not enough to shift the prior meaningfully. The model expects low scoring (combined xG ~1.7), and the intel does not suggest a high-scoring affair. Both defenses are likely to be compact in a World Cup opener. The most probable scorelines are 0-0, 0-1, 1-0, and 1-1, consistent with the model's distribution.

Group K· Matchday 17 · ·7 factors ·6 web sources ·called 2026-06-11
Algeria v Austria
Δ 3.7 pts
ALG
Draw
AUT
Model
33.7%
26.7%
39.5%
Agent
30.0%
28.0%
42.0%

The model prior already gives Austria a 40% win probability, and the supplied intel consistently supports an Austrian edge. Austria's recent form (4 wins in 5) is stronger than Algeria's (2 wins, 2 draws, 1 loss). The only head-to-head result is an Austria win. Key player analysis shows Austria's core (Sabitzer, Baumgartner, Alaba) have higher recent ratings and play at stronger clubs than Algeria's (Mahrez, Amoura). No injuries or conditions offset this. Therefore, I nudge the away win probability slightly higher to 42%, with home win reduced to 30% and draw at 28%. The model expects low scoring (Algeria 1.14, Austria 1.32 xG). Austria's defensive solidity (Alaba, etc.) and Algeria's modest attack suggest a 0-1 or 1-1 outcome is most likely. The 0-1 scoreline is the top model line at 12%, and I keep it as most probable. Other scorelines reflect a tight match with Austria slightly more likely to score.

Group J· Matchday 17 · ·7 factors ·6 web sources ·called 2026-06-11
Jordan v Argentina
Δ 7.5 pts
JOR
Draw
ARG
Model
7.2%
17.7%
74.5%
Agent
5.0%
13.0%
82.0%

Argentina are overwhelming favourites due to their perfect recent form, vastly superior key players (Messi, Lautaro, Alvarez vs. Jordan's modest squad), and status as defending champions. Jordan have lost three of their last five and lack the quality to trouble Argentina. The model's 75% away win is a reasonable starting point, but the intel justifies raising it to 83% given the chasm in class and current form. Argentina's expected goals (2.36) and Jordan's low xG (0.42) suggest a comfortable win. The most likely scorelines are 0-2 and 0-3, reflecting Argentina's ability to score multiple goals while Jordan struggle to create chances. A clean sheet for Argentina is probable given Jordan's poor attacking form (LDDLL).

Group J· Matchday 17 · ·7 factors ·6 web sources ·called 2026-06-11

Graded · 16

Played matches, most recent first — the call scored against the result, and whether the move beat the model.

Spain v Cape Verde
Δ 4.4 pts
ESP
Draw
CPV
Model
75.6%
17.2%
6.6%
Agent
80.0%
14.0%
6.0%
Final 0–0 call wrong lost to the model

The model's prior of a Spain win is very high (78%) and is supported by the massive gap in squad quality: Spain's key players are at elite clubs while Cape Verde's are at mid-level sides. Recent form slightly tempers Spain's edge (under-performing) and boosts Cape Verde (over-performing), but the opponent quality gap remains enormous. Lamine Yamal's injury is a minor blow to Spain's attack, but they still have Oyarzabal, Olmo, and Torres. Cape Verde's debut at the World Cup and high morale provide some intangible boost, but not enough to overcome the talent disparity. The forecast stays close to the model, with a slight reduction in Spain's win probability from 78% to 80% (rounding) due to Yamal's absence and Cape Verde's form. Spain's expected goals of 2.62 suggest they are likely to score 2-3 goals, while Cape Verde's 0.41 xG makes it unlikely they score more than once. The most probable scorelines are 2-0, 1-0, and 3-0, reflecting Spain's dominance but also the possibility of a low-scoring affair if Cape Verde defend compactly. The 0-0 draw is less likely but still plausible if Spain's attack misfires without Yamal.

Group H· Matchday 5 · ·7 factors ·9 web sources ·called 2026-06-15
Belgium v Egypt
Δ 5.5 pts
BEL
Draw
EGY
Model
52.1%
27.3%
20.5%
Agent
57.7%
25.7%
16.7%
Final 1–1 call wrong lost to the model

The model already favours Belgium (53% win), and the supplied intel reinforces that. Belgium's recent form is strong against weaker opponents, but their key players (De Bruyne, Tielemans) are in excellent form at top clubs, while Egypt's stars (Salah, Marmoush) have modest recent ratings. Egypt's coach publicly acknowledged the difficulty of the match. No injuries or conditions disrupt Belgium's advantage. I raise Belgium's win probability from 53% to 58% and lower Egypt's from 20% to 17%, reflecting a clear but not overwhelming edge. Belgium's expected goals (1.49) and Egypt's (0.72) suggest a low-scoring match leaning to Belgium. The most likely scoreline is 1-0 (14%), consistent with Belgium's defensive solidity and Egypt's struggle to score. 2-0 (12%) and 2-1 (10%) are also plausible. A draw (0-0 or 1-1) is possible but less likely given Belgium's quality advantage.

Group G· Matchday 5 · ·7 factors ·8 web sources ·called 2026-06-15
Saudi Arabia v Uruguay
Δ 2.4 pts
KSA
Draw
URU
Model
14.0%
27.7%
58.2%
Agent
12.7%
26.7%
60.7%
Final 1–1 call wrong lost to the model

The model's prior (Uruguay 59%, draw 27%, Saudi 13%) is already well-calibrated given the squad quality gap. Uruguay's key players are at elite clubs (Real Madrid, Atlético, Flamengo) while Saudi's are domestic-based, and Uruguay's recent form, though underperforming expectations, came against much stronger opponents (avg rank #29 vs #55). No injuries, no extreme conditions, and no market lines to adjust against. I keep the prior nearly unchanged, nudging Uruguay up 1% to reflect the clear talent advantage. The model expects low-scoring: Uruguay's xG 1.51, Saudi's 0.49. Uruguay's defensive solidity (Giménez, Valverde) and Saudi's lack of elite attacking threat support a 0-1 or 0-2 result. The top scorelines are 0-1 (19%), 0-0 (17%), 0-2 (13%). I retain these as the most likely.

Group H· Matchday 5 · ·7 factors ·9 web sources ·called 2026-06-15
Iran v New Zealand
Δ 5.9 pts
IRN
Draw
NZL
Model
56.4%
27.4%
16.2%
Agent
62.3%
24.0%
13.7%
Final 2–2 call wrong lost to the model

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.

Group G· Matchday 5 · ·7 factors ·9 web sources ·called 2026-06-15
Germany v Curaçao
Δ 8.6 pts
GER
Draw
CUW
Model
82.1%
8.3%
3.5%
Agent
90.7%
6.7%
2.7%
Final 7–1 call right beat the model

The model prior already heavily favours Germany (82% win), and every piece of intel reinforces that direction. Germany's recent form is excellent, their squad is stacked with world-class players from top clubs, and they face a Curaçao side that is not only a World Cup debutant but also missing key players Gorre and St Jago. The talent gap is enormous: Germany's key players operate at elite levels (Arsenal, Liverpool, Bayern) while Curaçao's come from lower-tier leagues. I have raised Germany's win probability from 82% to 90% to reflect the cumulative strength of these factors, and lowered the draw and away win accordingly. Germany's expected goals of 3.48 are already high, but Curaçao's defensive absences and the gulf in quality suggest even more goals are possible. However, Germany may not push excessively in a tournament opener, so the most likely scorelines are 3-0 and 4-0, with 5-0 also plausible. Curaçao's attack is weak (0.48 xG) and they are unlikely to score, but a consolation goal cannot be ruled out entirely, hence 3-1 and 4-1 appear. The model's top scorelines (2-0, 3-0, 1-0, 4-0) are adjusted upward slightly to reflect the stronger-than-average German form and Curaçao's absences.

Group E· Matchday 4 · ·7 factors ·6 web sources ·called 2026-06-14
Netherlands v Japan
Δ 5.7 pts
NED
Draw
JPN
Model
45.7%
25.7%
28.6%
Agent
40.0%
28.0%
32.0%
Final 2–2 call wrong beat the model

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.

Group F· Matchday 4 · ·7 factors ·9 web sources ·called 2026-06-14
Ivory Coast v Ecuador
Δ 0.8 pts
CIV
Draw
ECU
Model
26.0%
33.8%
40.1%
Agent
26.7%
33.0%
40.3%
Final 1–0 call wrong beat the model

The model's prior (25/33/42) reflects Ecuador's higher baseline strength, likely due to their deeper European-based talent. Recent form shows both teams over-performing, but Ivory Coast's margin is larger and their friendly win over France is a strong signal. However, Ecuador's key players at top clubs (Caicedo, Estupiñán) provide a quality edge. The head-to-head is blank, conditions are neutral, and no injuries affect either side. The facts slightly narrow the gap but not enough to flip the prior; I maintain the model's probabilities as the best estimate. The model's expected goals (CIV 0.70, ECU 1.05) suggest a low-scoring affair, with 0-0 and 0-1 as the top scorelines. The defensive solidity of both teams (Ivory Coast's recent form includes clean sheets, Ecuador's WDDWW shows defensive resilience) supports a low total. The most likely outcomes are 0-0, 0-1, and 1-0, with a draw slightly more probable than a win for either side.

Group E· Matchday 4 · ·7 factors ·9 web sources ·called 2026-06-14
Sweden v Tunisia
Δ 4.7 pts
SWE
Draw
TUN
Model
45.3%
27.7%
26.9%
Agent
50.0%
27.3%
22.7%
Final 5–1 call right beat the model

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.

Group F· Matchday 4 · ·7 factors ·9 web sources ·called 2026-06-14
Qatar v Switzerland
Δ 0.6 pts
QAT
Draw
SUI
Model
13.1%
19.4%
67.0%
Agent
13.0%
20.0%
67.0%
Final 1–1 call wrong beat the model

The model strongly favours Switzerland (68% away win), and the supplied intel supports that: Qatar's recent form is worse against weaker opponents, and Switzerland's key players are at a higher club level. The head-to-head is a single old result, conditions are neutral, and no injuries affect either side. I keep the prior essentially unchanged, nudging away win slightly to 69% to reflect the form and player quality gap. Switzerland's expected goals (2.31) and Qatar's (0.71) suggest a low-scoring game for Qatar and moderate for Switzerland. The most likely scorelines are 0-1, 0-2, and 0-3, consistent with the model. A draw or Qatar win is less probable given the gap in quality and form.

Group B· Matchday 3 · ·7 factors ·3 web sources ·called 2026-06-13
Brazil v Morocco
Δ 8.2 pts
BRA
Draw
MAR
Model
45.1%
32.1%
22.8%
Agent
53.3%
27.7%
19.0%
Final 1–1 call wrong lost to the model

The model's prior (46/32/22) already favours Brazil, and the supplied intel reinforces that. Brazil's recent form is solid against stronger opponents, their key players are at elite clubs and in good form, while Morocco has injury absences (Mazraoui, Ezzalzouli) that weaken their squad. The head-to-head win for Morocco in 2023 is a mild counter, but the overall quality gap and injury situation push Brazil's win probability higher, to 55%. The draw and Morocco win are reduced accordingly. Brazil's expected goals are moderate (model 1.14), but with their attacking talent and Morocco's defensive injuries, a 1-0 or 2-0 win is plausible. Morocco can still score (model 0.65), so 1-1 and 0-1 are possible but less likely. The most likely scorelines reflect a low-scoring match with Brazil controlling the game.

Group C· Matchday 3 · ·7 factors ·9 web sources ·called 2026-06-13
Haiti v Scotland
Δ 1.3 pts
HAI
Draw
SCO
Model
15.6%
22.0%
62.2%
Agent
16.0%
20.7%
63.3%
Final 0–1 call right beat the model

The model already heavily favours Scotland (63% away win), and the supplied intel supports that lean. Scotland's recent form is stronger (+0.26 pts/match vs expectations) against tougher opposition (rank #50) compared to Haiti's (in-line vs rank #76). Scotland's key players operate at a higher club level (Napoli, Aston Villa, Bournemouth) than Haiti's (Esteghlal, AEK Athens, Almere City), reinforcing the quality gap. No injuries, neutral conditions, and no head-to-head history leave no reason to diverge from the model; I slightly increase away probability to 65% to reflect the form and player quality edge. With Scotland's expected goals at 2.01 and Haiti's at 0.76, the most likely outcome is a low-scoring away win. The top scoreline 0-1 (13%) reflects Scotland's defensive solidity and Haiti's limited attack. 0-2 (12%) and 0-0 (9%) are also plausible, while 1-1 (9%) accounts for a potential Haiti goal. The distribution is concentrated on low totals given both teams' moderate attacking output.

Group C· Matchday 3 · ·7 factors ·9 web sources ·called 2026-06-13
Australia v Turkey
flips the pick Δ 8.3 pts
AUS
Draw
TUR
Model
37.6%
27.8%
34.6%
Agent
29.3%
28.0%
42.7%
Final 2–0 call wrong lost to the model

The model and prior are nearly balanced, but intel strongly favours Turkey. Turkey's recent form is excellent (+1.30 pts vs model) against opponents averaging rank #47, while Australia's form is neutral. Head-to-head history shows two Turkey wins. Key player analysis reveals a clear quality gap: Turkey's stars (Çalhanoglu, Güler, Aktürkoglu) play for elite clubs and have higher recent ratings than Australia's squad, which features players from lower-tier leagues. Recent news reinforces Turkey's confidence. I shift away win probability from 37% to 42%, lowering home win from 36% to 30%, keeping draw at 28%. Expected goals are low (model: AUS 1.16, TUR 1.18), suggesting a low-scoring match. Turkey's quality advantage and recent over-performance push them to be more likely to score, while Australia's defence (Souttar at Leicester) is decent. Most likely scorelines are 0-1, 1-1, 0-2, and 1-0, reflecting a tight game with Turkey slight favourites.

Group D· Matchday 3 · ·7 factors ·3 web sources ·called 2026-06-13
Canada v Bosnia and Herzegovina
Δ 3.6 pts
CAN
Draw
BIH
Model
52.5%
24.4%
22.9%
Agent
49.3%
28.0%
22.7%
Final 1–1 call wrong beat the model

The model's prior favours Canada (54%) due to home advantage and squad quality, but the confirmed absence of Alphonso Davies—Canada's best player and a world-class left-back—significantly weakens their defence and transition. Bosnia's recent form is poor (five draws) but they have veteran quality in Dzeko and a solid midfield. The Davies injury is a strong signal that reduces Canada's edge, so I lower their win probability from 54% to 50% and raise the draw to 28%, keeping Bosnia's away win near the model's 22%. With Canada's attack still potent (David, Buchanan) but Davies missing, the expected goals drop slightly from 1.72 to ~1.5. Bosnia's attack is modest (0.92 xG in model) and they struggle to score consistently. A 1-0 or 1-1 draw is most plausible, with 0-0 also possible given Canada's defensive uncertainty. The scoreline distribution shifts toward lower totals.

Group B· Matchday 2 · ·7 factors ·9 web sources ·called 2026-06-12
United States v Paraguay
Δ 8.0 pts
USA
Draw
PAR
Model
42.7%
27.9%
29.4%
Agent
50.7%
26.7%
22.7%
Final 4–1 call right beat the model

The model gives USA only a 40% win probability, but several supplied facts push strongly in USA's favor. Head-to-head shows USA dominance (4 of last 5 wins). Paraguay's key attacker Julio Enciso is confirmed out (nypost.com), weakening their attack. USA's key players are all at top European clubs, while Paraguay's are at lower levels. The squad quality gap and Enciso absence justify raising USA's win probability to 52%, with draw and Paraguay win reduced accordingly. USA's expected goals rise from 1.18 to about 1.5 given the squad advantage and Enciso absence, while Paraguay's drop from 0.99 to ~0.8. The most likely scorelines reflect a low-scoring match (1-0, 2-0, 1-1) but with increased chance of a USA clean sheet.

Group D· Matchday 2 · ·7 factors ·9 web sources ·called 2026-06-12
Mexico v South Africa
Δ 11.2 pts
MEX
Draw
RSA
Model
58.8%
24.7%
16.4%
Agent
70.0%
20.0%
10.0%
Final 2–0 call right beat the model

Mexico's recent form is strong (+1.03 pts/match vs expectations against decent opponents) while South Africa is underperforming (-1.10 pts/match vs weak opponents). The altitude at Estadio Azteca (2240 m) gives Mexico a significant acclimatization edge, as highlighted in recent news. Key player quality also favours Mexico, with multiple players at top European clubs versus South Africa's mostly domestic base. The model's 58% home win is too low given these factors; I raise it to 70%, with draw at 20% and away win at 10%. Mexico's expected goals are boosted by altitude and form, while South Africa's attack is weaker. The most likely scorelines reflect a 2-0 or 1-0 win for Mexico, with a 2-1 possibility if South Africa snatch a goal. Draws like 1-1 or 0-0 are less likely but still possible given South Africa's defensive organization.

Group A· Matchday 1 · ·7 factors ·6 web sources ·called 2026-06-11
South Korea v Czech Republic
Δ 1.8 pts
KOR
Draw
CZE
Model
37.3%
28.6%
34.2%
Agent
38.7%
29.0%
32.3%
Final 2–1 call right beat the model

The model prior is nearly balanced (36-29-35). Czech Republic's recent form is stronger (+0.35 pts/match overperformance) while South Korea is neutral (-0.02). A vague injury setback for South Korea in a recent friendly adds slight uncertainty. The altitude may marginally help the home side, but the Czechs have quality in Schick and Soucek. Overall, the facts nudge the probabilities slightly toward Czech Republic, reducing away win from 35% to 31% and increasing home win from 36% to 40%, with draw unchanged. Both teams have moderate attacking talent but defenses are solid; expected goals are low (~1.1 each). The most likely scorelines reflect a tight, low-scoring affair with 1-0, 0-0, 1-1, and 0-1 as top outcomes. The slight home edge makes 1-0 the single most likely.

Group A· Matchday 1 · ·7 factors ·9 web sources ·called 2026-06-11