Round of 32 · Match 4
BrazilvsJapan
2026-06-29·12:00 local·NRG Stadium · HoustonPredictions finalised
Match signals
Factors that favour each side, from statistical models to group stage form and match conditions. Longer bars = stronger advantage.
Brazil are dominant at 62% vs Japan's 13%. Quality, form, and model estimates all point the same way. An upset here would be a major story.
📊What the Models Say
Rates teams by a single strength number updated after every match. Simpler but fast to react. It rates Brazil at 52% to win vs Japan at 26%.
Simulates the goal-scoring process using attack and defence strength. The heaviest-weighted model. It rates Brazil at 59% to win vs Japan at 16%.
Groups teams by confederation to share information. Helps for teams with fewer matches. It rates Brazil at 56% to win vs Japan at 18%.
The published probability after calibration and adjustments. This is what the model says. It rates Brazil at 62% to win vs Japan at 13%.
All 3 models agree: Brazil is favoured. When models agree, the signal is stronger.
⚽Tournament Form
Brazil collected 10 points (3W 1D 1L) vs Japan's 5 (1W 2D 1L). A stronger tournament record.
Similar attacking output: Brazil 2.0 goals/match, Japan 2.0.
Brazil conceded just 0.8 goals/match vs Japan's 1.25. Tighter at the back.
Brazil's goal difference of +6 is better than Japan's +3. They outperformed opponents by more.
📈Momentum
Brazil's rating rose -4.9 during the tournament while Japan's moved -16.8. The tournament has been kinder to Brazil.
Japan's players improved their form ratings during the tournament (+0.0038) vs Brazil (-0.0011). Players trending upward.
🏆Team Quality
Brazil is rated 1984 vs Japan's 1904 (gap: 80). That's a noticeable gap in historical team strength.
The model expects Brazil to create 1.58 expected goals vs Japan's 0.67. More and better chances projected.
Brazil's top 3 starters are harder to replace (avg VORP 0.51) than Japan's (0.20). More star power in key positions.
Similar levels of squad familiarity from club football.
🌍Match Conditions
Brazil traveled 6,751km vs Japan's 10,792km. A shorter journey means less fatigue.
Brazil face a 2h timezone shift vs Japan's 14h. Less jet lag disruption.
17 signals across 5 categories. Signal strength reflects how large the gap is between the two teams on each factor. Signals are descriptive, not prescriptive.
Tahmin
Match-outcome probability
- Brazil win49.5%
- Draw27.6%
- Japan win22.9%
A clash of identities: Brazil's high-press approach meets Japan's low-block style in a fixture the model gives to Brazil at 62%.
▸Goller ve skorlar
Likeliest score 1–0 (16.0%) · xG 1.6 - 0.7
Expected goals
Mean of the Dixon-Coles joint goal distribution. Same fit that produces the most-likely-scoreline list below.
Most likely scorelines
- 1–016.0%
- 2–013.2%
- 1–111.8%
- 0–011.2%
- 2–18.8%
From the Dixon-Coles joint Poisson with the low-score correction. Scorelines are listed in probability order; this is a description of the model's distribution, not a recommendation.
Most likely half-time scorelines
- 0–033.0%
- 1–025.1%
- 0–110.4%
- 2–010.1%
- 1–19.1%
Same Dixon-Coles fit as the full-time list above, with rates halved to a 45-minute window and the low-score correction applied to that 1st-half block. The 0-0 row sits higher here than at full-time because fewer minutes have elapsed.
Goal totals
- More than 0.5 goals88.8%
- More than 1.5 goals66.5%
- More than 2.5 goals39.1%
- More than 3.5 goals19.1%
- More than 4.5 goals7.8%
- More than 5.5 goals2.7%
- Both teams score39.5%
Each row is the probability the match finishes with more than the listed number of goals. Both-teams-to-score is the probability each side scores at least once. All values are marginals of the Dixon-Coles joint goal grid that produces the scoreline list above — not market lines or any other operator construct.
Event-typed probabilities
- Brazil clean sheetOpposing team scores zero51.1%
- Japan clean sheetOpposing team scores zero20.6%
Derived from the same Dixon-Coles joint distribution as the scoreline list. These are descriptive event probabilities — see CLAUDE.md §3/§4 (formerly COMPLIANCE.md §4.2.7) for the framing the project uses.
Win-margin probability
- Brazil by 4+4.7%
- Brazil by 3+13.7%
- Brazil by 2+32.2%
- Brazil by 1+58.6%
- Draw26.4%
- Japan by 1+15.0%
- Japan by 2+4.1%
- Japan by 3+0.8%
- Japan by 4+0.1%
Each row is the probability the match ends with the listed margin or larger in that direction. Marginal of the Dixon-Coles joint goal grid; the “by 1+” rows plus the draw row sum to 1.
▸Maç nasıl şekillenir
Over 2.5 goals 39.1% · BTTS 39.5%
Game state through the match
- Brazil ahead59.3%
- Level25.0%
- Japan ahead15.7%
Probability of each game state at minutes 0, 15, 30, 45, 60, 75, 90 — derived from two independent thinned-Poisson processes with the Dixon-Coles per-team rates. The three lines always sum to 1 at each minute. The right column shows the state at the match's closing minute.
When the first goal arrives
- 0–1531.3%
- 15–3021.5%
- 30–4514.8%
- 45–6010.2%
- 60–757.0%
- 75–904.8%
- No goal10.5%
Probability the match's first goal arrives in each 15-minute window. Homogeneous Poisson with combined rate λ = λh + λa from the Dixon-Coles fit; the seven rows (six windows + no-goal tail) sum to 1.
Half-time / full-time grid
| HT ↓ / FT → | HBrazil win | DDraw | AJapan win |
|---|---|---|---|
| HBrazil ahead | 38.8% | 3.8% | 0.7% |
| DLevel | 18.7% | 17.8% | 6.2% |
| AJapan ahead | 1.8% | 3.7% | 8.6% |
Each cell is P(half-time result, full-time result). All nine cells sum to 1. Derived from a halved-λ Dixon-Coles fit for the first half plus an independent-Poisson second-half convolution.
Comeback probability
- Brazil trail at HT, avoid defeat at FT5.5%
- Japan trail at HT, avoid defeat at FT4.4%
Joint probability — P(side trailing at half-time AND avoiding defeat at full-time). NOT conditional on trailing at HT. Derived from the same half-time / full-time decomposition that produces the HT/FT grid above; a tied first half is neither a home nor an away comeback opportunity.
PK shootout simulator
If the match ends level after extra time, the model estimates the shootout outcome from each team's Bayesian-smoothed conversion / save rate (Model #15). The bracket simulator uses the symmetric (averaged) ordering; the two what-if scenarios below show how the win probabilities shift when conditioning on which team kicks first.
- Brazil58.7%
- Japan41.3%
- Brazil70.6%
- Japan29.4%
- Brazil47.1%
- Japan52.9%
First-kicker advantage
The first kicker's per-kick conversion rate is scaled by ×1.050 (about +5.0%), stacked on the Markov chain's structural asymmetry. Real World Cup shootouts use a coin toss for kicker order, so on average the order is 50/50 — the symmetric path above is the relevant number for a single fixture. The ordering-conditioned probabilities are a descriptive what-if scenario.
Literature: first kickers win ≈ 60% historically (Apesteguia & Palacios-Huerta, American Economic Review 2010; Vandebroek et al. 2016).
Per-team posteriors: Brazil conv 72.0%, save 26.0%; Japan conv 71.4%, save 20.0%. Smoothed against the global prior with prior strength 20 — see /docs/methodology/.
▸Takımlar ve oyuncular
Top scorer: Raphinha (15.6%)
Match detail
Brazil
Model-rated key players: Raphinha (FW) — P(scores) 15.6%; Gabriel Jesus (FW) — P(scores) 8.7%; Neymar (FW) — P(scores) 6.3%.
Brazil under Carlo Ancelotti play a high press game, holding 58% of the ball — among the highest in the tournament field. Their likely shape is a 4-2-3-1, though they have also used 4-3-3. They apply moderate pressing intensity (PPDA 17.1) and build patiently through midfield with 7.2 passes per attacking sequence. They generate a high volume of shots (16.5 per 90).
Brazil need their high press to force turnovers in dangerous areas — if opponents can play through the press, the space left behind is vulnerable. Physical conditioning and squad rotation will be critical to sustain pressing intensity across a long tournament.
Japan
Model-rated key players: Daichi Kamada (MF) — P(scores) 6.1%; Ayase Ueda (FW) — P(scores) 1.7%; Daizen Maeda (FW) — P(scores) 1.5%.
Japan under Hajime Moriyasu play a low block game, with just 44% possession — among the lowest in the field. Their likely shape is a 4-2-3-1, though they have also used other. They sit deeper and pick their moments to press (PPDA 26.7).
Japan will look to stay compact and frustrate opponents, limiting space and hitting on the break. Set-piece proficiency — both attacking and defending — becomes critical when open-play chances are limited by design. Managing minutes for Yūto Nagatomo across what could be seven matches will test the coaching staff's rotation planning.
Brazil's predicted XI averages 1,628 club minutes over the 2024-25 season (light load).
Brazil coverage: 67.0% (10/11 XI matched against the FBref Big-5) · Japan: 46.0% (6/11).
Brazil historically converts 10.8% of xG from set-pieces, contributing 0.17 expected set-piece goals in this fixture. Japan converts 6.3% from set-pieces (0.04 expected). Combined, the model expects 0.21 set-piece goals across the 90 minutes.
- P(Brazil scores set-piece goal) 15.7%
- P(Japan scores set-piece goal) 4.1%
- P(set-piece goal in match) 19.2%
Brazil: Matheus Pereira on corners (84 corners) (per fbref 2020 21) · Japan: Takefusa Kubo on corners (18 corners), Daichi Kamada on free kicks (per fbref 2022 23)
If a penalty is awarded to Brazil, the model gives 72.0% conversion, 71.4% for Japan. If this match goes to a shootout, the symmetric (coin-toss averaged) win probability is 58.7% Brazil / 41.3% Japan.
Brazil primary PK: Raphinha (4/4 in 2021-22, per fbref 2020 21) · Japan primary PK: Daichi Kamada (2/2 in 2022-23, per fbref 2022 23).
Derived from the model's per-fixture forecast joint and supporting reference data (predicted squads, set-piece xG share, PK posteriors, club minutes). See /docs/methodology/ for the full methodology.
Squad depth
Most irreplaceable starters
Brazil
- Bruno GuimarãesCentral midfieldNo natural backup0.67gap
- Lucas PaquetáAttacking midfieldNo natural backup0.45gap
- CasemiroDefensive midfieldCover: Fabinho · 0.440.42gap
Japan
- Ayase UedaStrikerCover: Yuito Suzuki · 0.310.34gap
- Kōki OgawaStrikerCover: Yuito Suzuki · 0.310.14gap
- Takefusa KuboWingerCover: Keito Nakamura · 0.590.13gap
Gap = how far a side's rating at the position falls from the starter to his likely in-squad replacement (named under each name). Larger = harder to replace. Descriptive metric, does not feed the published probabilities. Methodology →
Match conditions
- AltitudeNear sea level13 m
- Avg temperatureFive-year mean over the tournament window28.4 °C
- Avg humidity78%
- Heat stressShade WBGT ~31.8 °CHigh heat stress
- Pitch surfacetemporary natural grass over artificial turf
Indoor artificial-turf stadium laying a temporary natural-grass pitch for the tournament.
Heat stress is a shade Wet-Bulb Globe Temperature proxy from the venue's climatology mean temperature and humidity; FIFA mandates cooling breaks at WBGT 32 °C. Afternoon kickoff (local time). These are long-window averages, not a match-day forecast, and they are not inputs to the forecast.
Top scorers · P(scores in this match)
- Daichi KamadaPKMF6.1%
- Ayase UedaFW1.7%
- Daizen MaedaFW1.5%
Per-player scoring rate from Model #5 (`p_score_per_match`). Reflects each player's npxG/90, expected minutes, team xG share, and the average opposing-team defence. See /docs/methodology/.
Recent match form
Last match player ratings
Brazil
vs Scotland · avg 7.3
Japan
vs Sweden · avg 7.0
Player scores from official highlight analysis of each team's most recent match. Observational, not a model input. Methodology →
Video analysis: player performance
Per-player ratings and event breakdowns from official highlights analysis. Tap a player to see their full match timeline.
8Casemiro56'–60'Scored Brazil's vital equalizing goal with a header, showcasing his aerial threat and leadership.
2goals▼
Scored Brazil's vital equalizing goal with a header, showcasing his aerial threat and leadership.
Match timeline
8Martinelli81'–81'Scored the decisive winning goal for Brazil with a clinical finish from inside the penalty area.
1goals▼
Scored the decisive winning goal for Brazil with a clinical finish from inside the penalty area.
Match timeline
7Vinicius Jr.Displayed excellent dribbling skills, beating a defender and having a powerful shot denied by the goalkeeper and the post.
Displayed excellent dribbling skills, beating a defender and having a powerful shot denied by the goalkeeper and the post.
6Cunha10'–10'Had an early shot on target but did not contribute further to the scoreline.
1shots1on target▼
Had an early shot on target but did not contribute further to the scoreline.
Match timeline
6Rayan80'–80'Had a shot on goal that was deflected, showing some attacking intent in his limited time on the pitch.
1shots▼
Had a shot on goal that was deflected, showing some attacking intent in his limited time on the pitch.
Match timeline
8Suzuki10'–73'Made several crucial saves throughout the match, preventing Brazil from scoring more goals and keeping Japan competitive.
3saves▼
Made several crucial saves throughout the match, preventing Brazil from scoring more goals and keeping Japan competitive.
Match timeline
8Sano23'–23'Scored Japan's only goal with a composed finish after a strong attacking run, giving his team a surprising lead.
1goals▼
Scored Japan's only goal with a composed finish after a strong attacking run, giving his team a surprising lead.
Match timeline
4TomiyasuWas directly beaten by Vinicius Jr. in a dangerous attacking sequence, highlighting a defensive vulnerability.
▼
Was directly beaten by Vinicius Jr. in a dangerous attacking sequence, highlighting a defensive vulnerability.
Match timeline
Match observations
- The match between Brazil and Japan concluded with a 2-1 victory for Brazil.
- The atmosphere in the Japanese fan zone was one of tension and disappointment, reflecting their team's loss.
- In contrast, Brazilian supporters were seen in a state of euphoria, celebrating their team's triumph.
▸Perde arkası
Model-by-model comparison
Brazil vs Japan
| Model | Weight | Home | Draw | Away |
|---|---|---|---|---|
| EloRating-based strength estimate | 32% | 51.9% | 22.0% | 26.1% |
| Dixon-ColesGoal-process model with low-score correction | 63% | 58.9% | 25.4% | 15.6% |
| Hierarchical PoissonBayesian model with confederation pooling | 6% | 56.3% | 25.7% | 17.9% |
| Bayesian stackingLearned-weight combination | — | 63.5% | 25.3% | 11.2% |
| Ensemble (published)Uniform average + isotonic calibration | — | 62.4% | 25.0% | 12.6% |
How each model works
- Elo
- Each team carries a single strength rating updated after every match by a margin-aware K-factor. Match probabilities come from the logistic function of the rating gap. Elo is fast-adapting but coarse — it sees only who won and by how much, not how the goals were scored.
- Dixon-Coles
- A Poisson regression on team-level attack and defence parameters, fitted via maximum likelihood with an exponential time-decay weighting. The Dixon-Coles correction adjusts the four low-score cells (0-0, 1-0, 0-1, 1-1) where independent Poisson underestimates dependence. Produces full scoreline distributions, not just H/D/A.
- Hierarchical Poisson
- A Bayesian Poisson model fitted via MCMC (PyMC) with hierarchical priors that pool attack and defence parameters within confederations. Shrinks small-sample teams toward their confederation mean — helpful for nations with few recent competitive fixtures. Slower to fit but better-calibrated on the tails.
- Bayesian stacking
- Optimises simplex weights (w_elo, w_dc, w_hp) to maximise the leave-one-out log-score across a walk-forward backtest (Yao et al. 2018). The result is a weighted average of the three component models' probabilities, then isotonic-calibrated. Adds no extra features — just learns which component to trust more from historical accuracy.
- Ensemble (published)
- Equal-weight average of all three component models, followed by per-class isotonic regression calibration fitted on 24 months of walk-forward out-of-fold predictions. This is the probability published on the site. The uniform mean is deliberately simple — it avoids overfitting to the stacking weights' training window.
Three independent component models feed two combination strategies. The uniform ensemble is the published probability; Bayesian stacking uses learned weights. Amber bars flag >5pp divergence from the published number. Full methodology
Latest news & match context
- Brazil’s World Cup Collapse Revives Debate Over Faith and Soccer · Religion Unplugged · 14 Jul
- Stage:
- Round of 32 · Match 4
- Date:
- 29 Jun
- Venue:
- NRG Stadium, Houston
a 28°C kickoff modestly suppresses expected scoring at this venue.
Ranked by likely importance. None of these feed the forecast: the probabilities rest on team strength, venue conditions and the style matchup.
- 1.Elimination stakes: A one-off elimination tie. Motivation, risk appetite and game management under tournament pressure are not model inputs; the forecast rests on team strength and the style matchup.
- 2.Rest differential: Brazil have had 5 days since their previous match versus 4 for Japan. Rest and recovery are not model inputs.
- 3.Squad availability: 1 carrying a fitness doubt across the two squads. The forecast does not adjust for who is missing: its lineup channel currently contributes zero, so this is context the probabilities do not include.
Brazil
Brazil come in at close to full strength.
Japan
Japan: 1 carrying a fitness doubt.
- DoubtWataru Endo (midfielder) is carrying Foot injury — a depth-level fitness watch item.
Both projected XIs look intact; the fitness concerns are at squad-depth level rather than among first-choice starters.
Availability from the predicted squads and injury feed; forecast adjustments from the model's own decomposition. See /docs/methodology/.
Standard Pass
Bu maç ücretsiz ön izlemedir
Bu maç için modelin tam tahminini ücretsiz görüyorsunuz. Tüm 104 maç için aynı derinliği açın: olasılıklar, beklenen goller, skor dağılımları ve oyuncu bazında gol istatistikleri. Standard Pass turnuva sonuna kadar geçerlidir.
Every forecast graded against the real result, scored on 987 matches since 2014. See the scorecard.
24h money-back, no questions asked·No subscription, no auto-renewal·Access through 31 Dec 2026. See refund policy.