World Cup 2026 Matchday 6 Recap: The 1X2 Model Goes 4-for-4 — Messi's Hat-Trick, Haaland's Brace

After the brutal four-draw matchday 5, the World Cup 2026 model needed a response — and matchday 6 delivered one of its cleanest days of the tournament. All four favourites won, the 1X2 calls went 4-for-4, and the average 1X2 Brier landed near 0.10 — a world away from the ~0.45 of the previous day. Lionel Messi scored a hat-trick to equal a World Cup record; Erling Haaland delivered a brace. We will still lead with the uncomfortable part, because that is the deal here: our single value bet on the matchday lost. Good model day, losing betting day. Both are true, and we report both.

France 3–1 Senegal

Result: France 3–1 Senegal. Pre-match model: France win 61% / Draw 16% / Senegal win 23%. Brier: 0.078 — a clean, correct call on the favourite. France's expected-goals projection (λ 1.95) pointed to a comfortable scoring margin, and that is exactly how it played out. Kylian Mbappé struck twice and Bradley Barcola added the third; Senegal got a consolation but never seriously threatened the result. The goals model also read this one well: a projected total around 3.1 leaned Over 2.5, and four goals cleared it. See the France vs Senegal match centre for the full post-match breakdown.

Iraq 1–4 Norway

Result: Iraq 1–4 Norway. Pre-match model: Norway win 53% (away) / Draw 19% / Iraq win 28%. Brier: 0.111 — another correct call, and a satisfying one. Norway are the cleanest example of the model staying current: a recent rating overhaul had them correctly installed as clear favourites rather than the under-rated side they once were, and they justified it emphatically. Erling Haaland scored twice and Leo Østigård added another in a 4-1 rout; Aymen Hussein's strike was Iraq's only reply. The goals model projected a tighter game (~2.7) than the 4-1 result delivered — the win was more emphatic than expected, but the direction was never in doubt. Full details in the Iraq vs Norway match centre.

Argentina 3–0 Algeria

Result: Argentina 3–0 Algeria. Pre-match model: Argentina win 62% / Draw 19% / Algeria win 19%. Brier: 0.074 — the sharpest 1X2 call of the matchday. And it came on the night Lionel Messi scored a hat-trick to equal Miroslav Klose's all-time World Cup record of 16 goals — one before the break, two after. The goals model had projected a moderate-scoring game and leaned Under 3.5, with a 3-0 final landing exactly on the right side of that line. A model day and a history day rolled into one. See the Argentina vs Algeria match centre for the breakdown.

Austria 3–1 Jordan — the honest miss

Result: Austria 3–1 Jordan. Pre-match model: Austria win 45% / Draw 26% / Jordan win 29%. Brier: 0.151. Austria won, so the directional 1X2 read was correct — but this game carries two honest lessons, and we are not going to bury either.

First, our national-team model under-rated Austria. The model had them at just 45% to win; the sharp market (Pinnacle) implied closer to 66%. That gap is a known limitation: our international model leans heavily on Elo, which captures results but cannot see player quality the way a club-level dataset can. Austria were a stronger favourite than our number suggested, and they proved it with a comfortable 3-1. When our model disagrees this sharply with the sharp market on an international fixture, we flag it openly rather than publishing a confident contrarian call — and we did.

Second, and more painfully: this was the only value bet our model flagged on the matchday, and it lost. The bet was Under 2.5 goals @ 2.12 (+7.1% EV). A 3-1 final means four goals — comfortably Over. The bet is settled as a loss.

There is no spin to apply. The goals model leaned Under on a game that produced four goals. It happens; one bet is one data point, and a single losing total on a four-goal game does not invalidate a framework — but we log it as a loss, full stop, and it is the reason a 4-for-4 model day is still a red betting day on the P&L.

Matchday 6 model scorecard: a clean bounce-back

Here is the full matchday 6 1X2 breakdown:

• France 3–1 Senegal — model leaned France (61%), result: FRANCE. Brier 0.078. ✓
• Iraq 1–4 Norway — model leaned Norway (53%), result: NORWAY. Brier 0.111. ✓
• Argentina 3–0 Algeria — model leaned Argentina (62%), result: ARGENTINA. Brier 0.074. ✓
• Austria 3–1 Jordan — model leaned Austria (45%), result: AUSTRIA. Brier 0.151. ✓
• 4-match average 1X2 Brier: ~0.10 — four correct calls, and the model's strongest matchday of the tournament so far. A direct reversal of matchday 5's ~0.45.

This is what a probabilistic model looks like across a longer sample: a roughly 1-in-1,000 four-draw day (matchday 5) followed immediately by four favourites holding serve (matchday 6). Neither day changes the underlying calibration — they are both draws from the same distribution. We did not panic after the bad day, and we will not get carried away after the good one. The goals model was more mixed: the France Over hit, the Argentina Under hit, but the one bet we actually placed — the Austria Under — lost.

Running tournament record: 6 wins / 3 losses / 2 voids, +7.86 units

Here is the full cumulative World Cup 2026 value-bet record, updated through matchday 6. Convention as always: voids are counted separately — a void returns the stake in full and is never treated as a loss; the win rate is calculated only over settled wins and losses; and units are fractional-Kelly stakes (1 unit = 1% of bankroll).

• Matchday 1 — South Korea Over 2.5 @ 2.34 (+8.7% EV). WON ✓ — +1.34u.
• Matchday 2 — USA Over 2.5 @ 2.42 (+15.3% EV). WON ✓ — +3.82u. USA AH0 @ 1.56 (+6.2% EV). WON ✓ — +0.89u.
• Matchday 3 — Switzerland Under 2.75 @ 2.03 (+7.4% EV). WON ✓ — +1.03u.
• Matchday 4 — Sweden AH +0 @ 1.493 (+8.5pp CLV). WON ✓ — +0.49u. Netherlands AH +0 @ 1.583. VOID — stake returned.
• Matchday 5 — Spain Under 3.5 @ 1.97. WON ✓ — +0.97u. Belgium AH +0 @ 1.32. VOID — stake returned. Uruguay 1X2 @ 1.50. LOST ✗ — −1.00u. Saudi Arabia vs Uruguay Over 2.25 @ 1.85. LOST ✗ — −0.50u.
• Matchday 6 — Austria Under 2.5 @ 2.12 (+7.1% EV). LOST ✗ — −1.58u.
• Cumulative: 6 wins / 3 losses / 2 voids (66.7% win rate on settled bets). +7.86 units. Average EV at placement: +8.9%. Average closing-line value across the settled book remains positive (~+1.5pp).

The matchday 6 net is −1.58 units, pulling the cumulative from +9.44u to +7.86u. A single losing bet on a four-goal game. The tournament book is still comfortably profitable, the win rate is still two-in-three, and the EV at placement is still systematically positive — but we are not going to dress up a losing matchday as anything other than what it was. The number to watch over a full tournament is closing-line value, not the day-to-day P&L swing, and that signal remains positive across the book.

These are settled historical results, published as a transparent record — not live tips. The actionable value bets for tonight's fixtures are members-only.

What's next: tonight's previews

Matchday 7 brings a heavyweight slate — Portugal and England both open their campaigns. Our full pre-match model breakdowns are live now:

• Portugal vs DR Congo prediction — the model makes Portugal a heavy 71% favourite in a game it expects to flow with goals.
• England vs Croatia prediction — England 61%, and one of the matches where the model has flagged genuine value for members.
• Ghana vs Panama prediction — both sides arrive in poor form, yet the model still backs Ghana at 65%.
• Uzbekistan vs Colombia prediction — Colombia 63% as World Cup debutants Uzbekistan get their late-night test.

Follow all group standings, model probabilities and the live bracket on the World Cup 2026 hub. Yesterday's matchday 5 recap is also live: World Cup 2026 Matchday 5 Recap.