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

Tunisia vs Netherlands ๐Ÿ‡ณ๐Ÿ‡ฑ  WINNER HIT

FIFA World Cup 2026, Group F, Matchday 3 ยท Neutral site (Arrowhead Stadium, Kansas City)
Kickoff Jun 25, 2026 ยท 23:00 UTC
Final ยท Tunisia 1โ€“3 Netherlands
Goals: 3' Ellyes Skhiri (Netherlands); 7' Brian Brobbey (Netherlands); 54' Hazem Mastouri (Tunisia); 62' Jan Paul van Hecke (Netherlands)
Actual stats: xG 0.62 โ€“ 1.85 ยท 10 corners ยท 0 cards (0 red)

Prediction vs result  6/8 markets hit

MarketModel calledActualOutcome
1X2NetherlandsNetherlands (1-3)HIT
Score0-21-3MISS
O/U 2.5Over 2.5 (63.4%)4 goalsHIT
BTTSYes (51.7%)YesHIT
CornersOver 8.5 (52.6%)10 cornersHIT
CardsUnder 3.5 (63.2%)0 cardsHIT
Red cardNo (87.8%)NoHIT
AHTunisia +1.50LossMISS

Model vs market

MarketModelBook oddsBook p*EV
Tunisia win9.1%23.004.1%+109.1%
Draw16.1%9.0010.4%+45.0%
Netherlands win74.8%1.1085.5%-17.7%
Over 2.5 goals63.4%1.4068.2%-11.2%
Under 2.5 goals36.6%3.0031.8%+9.7%
Both teams score: yes51.7%2.5037.5%+29.2%
Both teams score: no48.3%1.5062.5%-27.6%
Over 8.5 corners (exp 8.86)52.6%1.6755.8%-12.4%
Under 8.5 corners47.4%2.1044.2%-0.4%
Over 3.5 cards (exp 3.07)36.8%โ€”โ€”โ€”
Under 3.5 cards63.2%โ€”โ€”โ€”
Any red card12.2%โ€”โ€”โ€”
Model-only estimates (no straight market equivalent): expected goals Tunisia 0.82 โ€“ 2.50 Netherlands ยท fair Asian handicap Tunisia +1.50
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). โ€œโ€”โ€ = market not in the captured feed.
Model reasoning (chain of thought)
PREDICTION RATIONALE โ€” Tunisia vs Netherlands  (FIFA World Cup 2026, Group F, Matchday 3)
Venue: Neutral site (Arrowhead Stadium, Kansas City)

1. DATA. Last-10 internationals (Sofascore interception scraper).
   Tunisia: record WLDLWDLLLL, 1.10 GF / 2.40 GA,
   xG 0.54 / xGA 1.71.
   Netherlands: record WWDWWDLWDW, 2.50 GF / 0.80 GA,
   xG 2.16 / xGA 0.85.
   Opponent-strength tiers (configs.MASTER_TIERS) applied per match.

2. RATINGS. Tunisia atk 0.82 def 1.47; Netherlands atk 1.34 def 0.80.
   NEUTRAL VENUE: home boost / away penalty set to 1.0/1.0 (both teams
   travel to a host-nation site; no host advantage applies โ€” a key difference
   from the Korea/Canada/USA host-nation games). lambda Tunisia 0.82, Netherlands 2.50
   after 90% model weight + shrink to 1.3 baseline.
   Injuries: {'Tunisia': ['No major absences reported'], 'Netherlands': ['No major absences reported']}.
   Context: Group F MD3. Netherlands heavy favourites vs Tunisia. Expect the model to under-rate the Dutch (cold-on-elites). MD3 rotation risk. Neutral venue.

3. MODEL. Bivariate Poisson (lambda3 0.12). 1X2 Tunisia 9.1% /
   draw 16.1% / Netherlands 74.8%. Winner Netherlands; score 0-2; fair AH Tunisia +1.50.
   Over 2.5 63.4%; BTTS 51.7%.

4. EVENTS. Referee TBD: 4.2 Y + 0.13 R/match
   blended (weight 0.22) with WC base 2.7044 -> expected
   cards 3.07, P(over 3.5) 36.8%; P(any red)
   12.2%. Corners expected 8.9 (scale
   0.9899) -> P(over 8.5) 52.6%.

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