Match predictor

Dixon-Coles scoreline model. The recommended pick maximises expected points under your scoring rules — not just the most likely score.

Scoring rules
Expected goals0.88–0.78
✅ Best pick0–0
Exp. points0.821
Most likely0–0 21%
Outcome probabilitymodel only · no odds
Argentina 33%Draw 39%Brazil 28%
Scoreline probability matrix
0
1
2
3
4
5
0
21
12
6
2
0
0
1
14
15
5
1
0
0
2
7
6
2
1
0
0
3
2
2
1
0
0
0
4
0
0
0
0
0
0
5
0
0
0
0
0
0
rows = Argentina goals · cols = Brazil goals
⚡ Wild Picks — derived markets model onlyknockout rules · goals incl. ET/pens
Total goals (incl. ET/pens)
1most likely
Over 2.5 (90'): 23% · Under: 77%
0:0% 1:34% 2:14% 3:20% 4:5% 5:6% 6:3%
Decided in ET / penalties?
NO39% yes
Level after 90': 39% · reaches shootout: 24%
Both teams to score
NO34% yes
Argentina & Brazil both find the net
Either team scores 3+
NO10% yes
Any single team reaching 3 goals
Half-time result
Draw
Argentina 23% · Draw 58% · Brazil 20%
Goalless at half-time
YES50%
0–0 at the break
First to score
Argentina
Argentina 43% · Brazil 38% · none 19%
Argentina goals (incl. ET/pens)
1most likely
0:24% 1:29% 2:21% 3:11% 4:8%
Brazil goals (incl. ET/pens)
0most likely
0:29% 1:28% 2:19% 3:10% 4:7%
Goal after 75'
NO31% yes
At least one goal in the last 15'+
Total yellow cards
2~2.42 exp
Over 3.5: 23% · this WC averages 2.3/match
Any red card?
NO11% yes
Only ~8% of matches this WC
Clean sheet
Argentina 46% · Brazil 42%
Team concedes zero

Half-time ≈ 45% of goals; late-goal ≈ 22% after the 75th. Card markets use this tournament's actual played-match rates (re-run export_cards each matchday). Corners not in the feed.

Rating breakdown — how each expected-goals number is built
ComponentArgentinaBrazil
Base attack · results + FIFA rank1.531.60
Base defence · higher = stingier1.851.66
Squad strength ×0 · OVR 76/75.1+0.00+0.00
Recent form ×0.1 · last 5, ppg+0.00+0.00
Final attack / defence1.53 / 1.851.60 / 1.66
→ Expected goals0.880.78

xG = exp(attack − opponent defence). Squad OVR is shown for context but weighted 0 — backtesting found it hurt accuracy (the base already captures strength). Form is lightly weighted and engages once teams have prior results.

Form, ranking & availability — why the model leans this way

Loading context…

Squad ratings — top performers, side by side

Loading live player data…