WC2026 Model Tracker
Bivariate-Poisson predictions vs actual results · a live, honest experiment

🇪🇸 Spain vs Belgium 🇧🇪  WINNER HIT

FIFA World Cup 2026, Quarter-final · Neutral site (SoFi Stadium, Inglewood)
Kickoff Jul 10, 2026 · 19:00 UTC
Quarter-final · who goes through
To advance: Spain 58.9%  /  Belgium 41.1%
= P(win in 90′) + ½·P(draw); a level game at 90′ goes to extra time / penalties (treated as a coin flip). 1X2 and Asian handicap (Spain -0.25) below are the 90-minute markets.
Final (90′) · Spain 2–1 Belgium
Spain advanced · model to-advance pick Spain 58.9% ADVANCE CALL HIT
Goals: 30' Fabián Ruiz (Spain); 41' Charles De Ketelaere (Belgium); 88' Mikel Merino (Spain)
Actual stats: xG 2.1 – 0.37 · 6 corners · 4 cards (0 red)

Prediction vs result  4/8 markets hit

MarketModel calledActualOutcome
1X2SpainSpain (2-1)HIT
Score1-12-1MISS
O/U 2.5Under 2.5 (55.0%)3 goalsMISS
BTTSYes (51.0%)YesHIT
CornersOver 8.5 (52.7%)6 cornersMISS
CardsUnder 3.5 (65.0%)4 cardsMISS
Red cardNo (87.8%)NoHIT
AHSpain -0.25Full WinHIT

Model vs market

MarketModelBook oddsBook p*EV
Spain win44.8%1.6258.8%-27.4%
Draw28.1%4.0023.8%+12.4%
Belgium win27.1%5.5017.3%+49.1%
Over 2.5 goals45.0%1.8052.6%-18.9%
Under 2.5 goals55.0%2.0047.4%+9.9%
Both teams score: yes51.0%1.8052.0%-8.2%
Both teams score: no49.0%1.9548.0%-4.4%
Over 8.5 corners (exp 8.87)52.7%
Under 8.5 corners47.3%
Over 3.5 cards (exp 2.99)35.0%
Under 3.5 cards65.0%
Any red card12.2%
Model-only estimates (no straight market equivalent): expected goals Spain 1.43 – 1.07 Belgium · fair Asian handicap Spain -0.25
In the table above, Model = our probability and Book odds / Book p* = the bookmaker's (p* has the margin removed). EV = value per unit staked at the model's probability (positive = model sees value). “—” = market not in the captured feed.
Model reasoning (chain of thought)
PREDICTION RATIONALE — Spain vs Belgium  (FIFA World Cup 2026, Quarter-final)
Venue: Neutral site (SoFi Stadium, Inglewood)

1. DATA. Last-10 internationals (Sofascore interception scraper).
   Spain: record DWDDWDWWWW, 1.80 GF / 0.40 GA,
   xG 1.95 / xGA 0.45.
   Belgium: record WWDWWDDWWW, 3.30 GF / 0.80 GA,
   xG 2.08 / xGA 0.94.
   Opponent-strength tiers (configs.MASTER_TIERS) applied per match.

2. RATINGS. Spain atk 1.35 def 0.52; Belgium atk 1.55 def 0.77.
   NEUTRAL VENUE: home boost / away penalty set to 1.0/1.0 (both teams
   travel to a host-nation site; no host advantage applies — a key difference
   from the Korea/Canada/USA host-nation games). lambda Spain 1.43, Belgium 1.07
   after 90% model weight + shrink to 1.3 baseline.
   Injuries: {'Spain': ['No major absences reported'], 'Belgium': ['No major absences reported']}.
   Context: Quarter-final. Spain — who have beaten the model's lambda four straight games — vs a Belgium side that dismantled the USA 4-1. Spain favoured; Belgium the best-attacking side they've met. Recalibrated settings (tau 0.4, shrink 1.3). Knockout: a draw goes to ET/pens. Neutral venue.

3. MODEL. Bivariate Poisson (lambda3 0.12). 1X2 Spain 44.8% /
   draw 28.1% / Belgium 27.1%. Winner Spain; score 1-1; fair AH Spain -0.25.
   Over 2.5 45.0%; BTTS 51.0%.

4. EVENTS. Referee TBD: 4.2 Y + 0.13 R/match
   blended (weight 0.20) with WC base 2.6438 -> expected
   cards 2.99, P(over 3.5) 35.0%; P(any red)
   12.2%. Corners expected 8.9 (scale
   0.964) -> P(over 8.5) 52.7%.

5. LIMITATIONS. 10-match samples; hand-set tiers/priors; Poisson corners; lineups
   probable not confirmed; card base rate calibrated on only 4 WC matches.