Wednesday 15 July 2026 FIFA World Cup 2026 · Semifinals · Day 2 Bulletin № 031
MM·26 Prediction Agent

The Expected
Points Bulletin

Scoreline picks that maximise contest points — not headlines. [ market 60% · form 25% · xG 15% · ranking 15% · last-match v2 ]

The Ledger
QF · 10 Jul
40 pts
ESP 2–1 exact ★
QF · 11 Jul
0 pts
no bulletin — ENG & ARG unpicked
SF · 14 Jul
0 pts
ESP 2–0 — the hump never came
SF · 15 Jul
0 pts
ARG 90+2′ winner killed the draw
930/4040points
thru the semifinals
Folio I — FT: England 1–2 Argentina Mercedes-Benz, Atlanta · 3:00 PM ET

England v Argentina

FIFA № 4 (~1828 pts) — vs — FIFA № 1 (~1877 pts)

Official Pick
1–1
E[pts] 7.4 · exact 14.4%

Maradona ’86, Beckham ’98, Beckham again in ’02 — and now Atlanta. The world champions, with Messi (39, 8 goals, probably his last World Cup) against an England side chasing a first final since 1966 behind Bellingham & Kane (6 each). The blend lands on a dead-even 35/30/35 — and both these teams stood at 1–1 after ninety in their quarter-finals before winning in extra time. Argentina’s legs carry 240 ET minutes; England’s Quansah ban ends only for the final. The Under is juiced (1.61) and 1–1 leads the ledger by 1.4 expected points — the clearest draw signal of the tournament. Fifth and biggest coin-flip draw pick. FT 1–2 — alive until stoppage: Gordon tapped England ahead (55′), Enzo Fernández thundered the equaliser (85′), and our 1–1 stood at 90:00 — then Lautaro Martínez headed Messi’s cross in at 90+2′. Both Argentina goals from Messi deliveries; the champions reach a second straight final, and the draw pick dies inside stoppage time. 0 pts.

Win · Draw · Winblended probabilities
ENG 35%
30%
ARG 35%
England windraw (90')Argentina win

Signal Blend — expected goals

Closing market, de-viggedw 60%1.21 – 1.12
Twelve-month form & stylew 25%1.55 – 1.75
In-tournament xG model 4thw 15%0.97 – 1.58
FIFA ranking (We 45%)w 15%1.20 – 1.40
Last-match vs expectation v2Δ softened+0.09 / -0.02
Final modelρ -0.1721.34 – 1.33
Scoreline Probability Grid — P(exact) %
ARG →
ENG ↓
0
1
2
3
4
0
9.00–0
7.00–1
6.10–2
2.70–3
0.90–4
1
7.11–0
14.41–1
8.21–2
3.71–3
1.21–4
2
6.22–0
8.32–1
5.52–2
2.52–3
0.82–4
3
2.83–0
3.73–1
2.53–2
1.13–3
0.43–4
4
0.94–0
1.24–1
0.84–2
0.44–3
0.14–4

Grid covers 97.6% of outcomes · over 2.5 goals: 50.1% · 90-minute scoreline

Pick Ledger — ranked by expected points

ScoreP(exact)E[points]
1–1◂ PLAY14.4%7.35
2–18.3%5.99
1–28.2%5.94
0–09.0%5.72
1–07.1%5.64
0–17.0%5.59

England — form & style

  • Six games unbeaten in 90′: 4–2 CRO, 0–0 GHA, 2–0 PAN, 2–1 COD, 3–2 MEX, 2–1 AET Norway (1–1 at 90′; Bellingham 45+2′ & 93′).
  • Bellingham & Kane 6 goals each; Rice fit after a stomach bug; FIFA wiped all single yellows post-QF — no accumulation risk today.
  • Quansah serves the second game of his ban (back for a final); Henderson out (broken wrist). Opta: England advance in 52% of sims.

Argentina — form & style

  • Six wins from six, 17 goals (most in the field): 3–0 ALG, 2–0 AUT, 3–1 JOR, 3–2 AET CPV, the 3–2 Egypt comeback, 3–1 AET Switzerland (1–1 at 90′).
  • Messi: 8 goals, level with Mbappé for the Boot; cleared fit after an eye knock. Álvarez & L. Martínez finished the Swiss off in ET.
  • 240 extra-time minutes in the last three rounds; Scaloni trialling a back three. Romero & Paredes cleared after knocks.
Odds (DK): 2.65 / 2.90 / 3.00 · O2.5 2.30 · U2.5 1.61 xG ENG 1.60 · ARG 2.30 · final 1.34–1.33 · ρ -0.17 · FT 1–2 (Gordon 55′, Enzo 85′, Lautaro 90+2′) → pick 1–1, 0 pts
Appendix A

How the picks are made

STEP 01
Blend four signals

Start from the de-vigged closing market (60%), then blend twelve-month form & style (25%), an in-tournament xG model (15%) rating each team on the expected goals it created & conceded across its four games so far, and the FIFA ranking's Elo expectancy (15%). The xG signal is shrunk to the league mean so it nudges rather than dictates.

STEP 02
Last-match read, softened v2

Each team's xG is nudged by how its most recent game beat/missed expectation; the attack term (finishing) is halved because it regresses. The read uses each side's final group game — now a week old for today's slate, so treat it as a light thumb on the scale.

STEP 03
Fit Dixon-Coles & maximise points

A Poisson model with the Dixon-Coles low-score correction ρ gives a probability for every scoreline. Each is ranked by 30·P(exact) + 10·P(1X2) — the contest's scoring, which stack (an exact hit is worth 40). Knockouts are scored on the 90-minute result, so the model still picks within 90 minutes.

!

Semifinals, Day 2 — graded: 0, by two minutes of stoppage. England 1–2 Argentina: Gordon (55′), Enzo Fernández’s 85′ rocket, and the 1–1 we picked was the live score at the ninety — until Lautaro Martínez met Messi’s cross at 90+2′. Messi assisted both; Argentina meet Spain in Sunday’s final. The five knockout draw picks close 1-for-5 — each priced fairly, four undone by single moments. Tournament 930/4040. Saturday: the bronze final, France–England in Miami. Scoring: 30 exact · 10 winner (max 40).

Appendix B

The ledger, audited — through the round of 16

AUDIT 01
Calibration: the engine is honest

Across 24 graded knockout picks: the model said 25% of games end level after ninety — 25% did (6 of 24). It projected 2.62 goals a game — 2.65 arrived. It claimed ~12.9% on its exact picks and landed 8.3% (2 of 24) — inside one standard error at this sample size. Nothing in the data says the probabilities are wrong.

AUDIT 02
Return: running at expectation

Knockout haul: 210 of 960 (21.9% of max) against the model's own pre-match expectation of ~21%. Results hit 15/24 (win-picks 14/20 = 70%); two exacts ★ (Canada 0–1, Colombia 1–0). The actual scoreline landed in our published top-three 55% of the time against a modelled 33% — the grid's shape has been better than advertised; the top line has been unlucky within it.

AUDIT 03
What changes: nothing structural

Variance dominates at n=24. Two behavioural rules survive the audit: coin-flip draw picks stay (the draw share was priced exactly right), and no exact-chasing — the second-ranked scoreline has cashed five times to the top line's one, a gap comfortably inside luck; abandoning the E[pts] argmax would trade points for stories. Eight picks remain (four quarter-finals, two semis, the bronze final and the final): expected haul ≈ +70, projected finish ≈ 950/4120.