Welcome to another edition of the Pressbeat Podcast, also on mediumwaves 1575 kHz. From Paris I’m Ami Carter-Wilson.
By A. Carter-Wilson
This is the first instalment of The Journal’s World Cup Model — a series that will track, update, and project the FIFA World Cup 2026 winner using finite-difference polynomial extrapolation. Each matchday round brings new data; the model is re-fitted accordingly. Today’s edition covers june 26.
Group-Stage Standings —
With 10 of 12 groups through two matches, here is the full ledger (Groups KâL have one match played).
| Grp | Team | P | W | D | L | GF | GA | Pts |
|---|---|---|---|---|---|---|---|---|
| A | 🇲🇽 Mexico | 2 | 2 | 0 | 0 | 3 | 0 | 6 |
| 🇰🇷 Korea Republic | 2 | 1 | 0 | 1 | 2 | 2 | 3 | |
| 🇨🇿 Czechia | 2 | 0 | 1 | 1 | 2 | 3 | 1 | |
| 🇿🇦 South Africa | 2 | 0 | 1 | 1 | 1 | 3 | 1 | |
| B | 🇨🇦 Canada | 2 | 1 | 1 | 0 | 7 | 1 | 4 |
| 🇨🇭 Switzerland | 2 | 1 | 1 | 0 | 5 | 2 | 4 | |
| 🇧🇦 Bosnia & Herz. | 2 | 0 | 1 | 1 | 2 | 5 | 1 | |
| 🇶🇦 Qatar | 2 | 0 | 1 | 1 | 1 | 7 | 1 | |
| C | 🇧🇷 Brazil | 2 | 1 | 1 | 0 | 4 | 1 | 4 |
| 🇲🇦 Morocco | 2 | 1 | 1 | 0 | 2 | 1 | 4 | |
| 🏴 Scotland | 2 | 1 | 0 | 1 | 1 | 1 | 3 | |
| 🇭🇹 Haiti | 2 | 0 | 0 | 2 | 0 | 4 | 0 | |
| D | 🇺🇸 USA | 2 | 2 | 0 | 0 | 6 | 1 | 6 |
| 🇦🇺 Australia | 2 | 1 | 0 | 1 | 2 | 2 | 3 | |
| 🇵🇾 Paraguay | 2 | 1 | 0 | 1 | 2 | 4 | 3 | |
| 🇹🇷 Türkiye | 2 | 0 | 0 | 2 | 0 | 3 | 0 | |
| E | 🇩🇪 Germany | 2 | 2 | 0 | 0 | 9 | 2 | 6 |
| 🇨🇮 Côte d’Ivoire | 2 | 1 | 0 | 1 | 2 | 2 | 3 | |
| 🇪🇨 Ecuador | 2 | 0 | 1 | 1 | 0 | 1 | 1 | |
| 🇨🇼 Curaçao | 2 | 0 | 1 | 1 | 1 | 7 | 1 | |
| F | 🇳🇱 Netherlands | 2 | 1 | 1 | 0 | 7 | 3 | 4 |
| 🇯🇵 Japan | 2 | 1 | 1 | 0 | 6 | 2 | 4 | |
| 🇸🇪 Sweden | 2 | 1 | 0 | 1 | 6 | 6 | 3 | |
| 🇹🇳 Tunisia | 2 | 0 | 0 | 2 | 1 | 9 | 0 | |
| G | 🇪🇬 Egypt | 2 | 1 | 1 | 0 | 4 | 2 | 4 |
| 🇮🇷 IR Iran | 2 | 0 | 2 | 0 | 2 | 2 | 2 | |
| 🇧🇪 Belgium | 2 | 0 | 2 | 0 | 1 | 1 | 2 | |
| 🇳🇿 New Zealand | 2 | 0 | 1 | 1 | 3 | 5 | 1 | |
| H | 🇪🇸 Spain | 2 | 1 | 1 | 0 | 4 | 0 | 4 |
| 🇺🇾 Uruguay | 2 | 0 | 2 | 0 | 3 | 3 | 2 | |
| 🇨🇻 Cabo Verde | 2 | 0 | 2 | 0 | 2 | 2 | 2 | |
| 🇸🇦 Saudi Arabia | 2 | 0 | 1 | 1 | 1 | 5 | 1 | |
| I | 🇫🇷 France | 2 | 2 | 0 | 0 | 6 | 1 | 6 |
| 🇳🇴 Norway | 2 | 2 | 0 | 0 | 7 | 3 | 6 | |
| 🇸🇳 Senegal | 2 | 0 | 0 | 2 | 3 | 6 | 0 | |
| 🇮🇶 Iraq | 2 | 0 | 0 | 2 | 1 | 7 | 0 | |
| J | 🇦🇷 Argentina | 2 | 2 | 0 | 0 | 5 | 0 | 6 |
| 🇦🇹 Austria | 2 | 1 | 0 | 1 | 3 | 3 | 3 | |
| 🇩🇿 Algeria | 2 | 1 | 0 | 1 | 2 | 4 | 3 | |
| 🇯🇴 Jordan | 2 | 0 | 0 | 2 | 2 | 5 | 0 | |
| K | 🇨🇴 Colombia | 1 | 1 | 0 | 0 | 3 | 1 | 3 |
| 🇨🇩 Congo DR | 1 | 0 | 1 | 0 | 1 | 1 | 1 | |
| 🇵🇹 Portugal | 1 | 0 | 1 | 0 | 1 | 1 | 1 | |
| 🇺🇿 Uzbekistan | 1 | 0 | 0 | 1 | 1 | 3 | 0 | |
| L | 🏴 England | 1 | 1 | 0 | 0 | 4 | 2 | 3 |
| 🇬🇭 Ghana | 1 | 1 | 0 | 0 | 1 | 0 | 3 | |
| 🇵🇦 Panama | 1 | 0 | 0 | 1 | 0 | 1 | 0 | |
| 🇭🇷 Croatia | 1 | 0 | 0 | 1 | 2 | 4 | 0 |
The Extrapolation Engine
The Journal’s model converts raw match data into a projection curve through successive finite differences â because the right polynomial at the right time is worth a thousand pundits. Here is how it works.
Step 1: The Momentum Integral. For each team k, define the cumulative goal trajectory Gk(n) as the total goals scored through match n, with Gk(0) = 0 by convention:
This is the discrete integral of the team’s scoring process, sampled at each match index.
Step 2: First Derivative â the goal-rate vector. The first forward difference gives goals scored in match 1: the team’s initial scoring velocity. For teams with two matches played, the average rate is GF ÷ 2. Top performers at Matchday 2:
| Team | GF | G’_k (goals/match) |
|---|---|---|
| 🇩🇪 Germany | 9 | 4.5 |
| 🇳🇴 Norway | 7 | 3.5 |
| 🇨🇦 Canada | 7 | 3.5 |
| 🇳🇱 Netherlands | 7 | 3.5 |
| 🇫🇷 France | 6 | 3.0 |
| 🇯🇵 Japan | 6 | 3.0 |
| 🇺🇸 USA | 6 | 3.0 |
| 🇸🇪 Sweden | 6 | 3.0 |
Step 3: Second Derivative â acceleration of form. With Gk(0)=0 and two match results we have three data points and can compute the second forward difference: . A negative value means the team scored fewer goals in match 2 than match 1. Zero means perfectly consistent output.
| Team | G(1)* | G(2) | Δ²G(0) | Form |
|---|---|---|---|---|
| 🇩🇪 Germany | 7 | 2 | -5 | ↓ Decelerating |
| 🇫🇷 France | 3 | 3 | +0 | → Consistent |
| 🇯🇵 Japan | 3 | 3 | +0 | → Consistent |
| 🇳🇴 Norway | 4 | 3 | -1 | ↓ Decelerating |
| 🇨🇦 Canada | 4 | 3 | -1 | ↓ Decelerating |
| 🇳🇱 Netherlands | 4 | 3 | -1 | ↓ Decelerating |
| 🇦🇷 Argentina | 3 | 2 | -1 | ↓ Decelerating |
| 🇺🇸 USA | 4 | 2 | -2 | ↓ Decelerating |
| 🇲🇽 Mexico | 3 | 0 | -3 | ↓ Decelerating |
| 🇸🇪 Sweden | 5 | 1 | -4 | ↓ Decelerating |
* Match-1 goals where confirmed; otherwise estimated from total GF.
Step 4: Third Derivative (jerk). With three data points, the interpolating polynomial is degree 2 at most. The third derivative vanishes identically for every team.
Step 5: Fourth and fifth derivatives (snap and crackle). Also zero. Higher-order terms are not estimable until more matches are played. As teams accumulate match data, the polynomial gains resolution and higher derivatives become meaningful. For now, the model is clean and minimal.
The Projection: Newton’s Forward-Difference Polynomial
Newton’s forward-difference interpolation constructs the unique polynomial of minimum degree passing through all observed data points. For a team with G(0)=0, G(1), and G(2) known, the degree-2 interpolant is:
Evaluated at n = 21 â the World Cup Final by match-index convention â this gives each team’s projected cumulative goal output at tournament end.
For Germany â :
For France and Japan â both with Δ²=0, linear trajectories:
First-edition projection table at n=21:
| Team | Δ¹G(0) | Δ²G(0) | P(n=21) | Rank |
|---|---|---|---|---|
| 🇫🇷 France | +3 | +0 | +63 | #1 |
| 🇯🇵 Japan | +3 | +0 | +63 | #2 |
| 🇳🇴 Norway | +4 | -1 | -126 | #3 |
| 🇨🇦 Canada | +4 | -1 | -126 | #4 |
| 🇳🇱 Netherlands | +4 | -1 | -126 | #5 |
| 🇦🇷 Argentina | +3 | -1 | -147 | #6 |
| 🇺🇸 USA | +4 | -2 | -336 | #7 |
| 🇲🇽 Mexico | +3 | -3 | -567 | #8 |
| 🇸🇪 Sweden | +5 | -4 | -735 | #9 |
| 🇩🇪 Germany | +7 | -5 | -903 | #10 |
The Verdict — Edition 1
Germany’s nine goals across two matches look impressive on the table. The polynomial disagrees. Seven goals in match 1 followed by two in match 2 encodes a sharp deceleration (Δ²=−5) that, extrapolated to match 21, produces a projected score of −903. Mathematically, Germany will have scored negative goals by the final. The model does not apologise.
France and Japan both scored three goals per match: second derivative zero, polynomial linear, projection clean. At match 21 both arrive at +63.
France and Japan are the Journalâs co-favourites to win the FIFA World Cup 2026.
No formula is infallible â especially one fitted to two data points. But that is precisely why we publish now rather than at the final whistle.
Coming Next: Bayesian Updates After Each Round
This model is a living document. After each completed matchday round, the Journal will re-fit the Newton polynomial with the updated data, treating each new match result as an observation that updates the prior projection. In Bayesian terms:
- Prior: the polynomial coefficients estimated from todayâs data â Δ¹G_k and Δ²G_k as computed above.
- Likelihood: each new match result adds a data point G_k(3), G_k(4)⦠to the teamâs trajectory.
- Posterior: an updated polynomial of higher degree, with narrower projections as the sample size grows.
As the group stage concludes, three things will happen: the polynomial degree will increase, the second-derivative estimates will stabilise, and the gap between contenders and pretenders will widen. Teams that sustain a flat or rising Δ² will climb the rankings. Teams that peaked early will fall further.
The next edition will incorporate Matchday 3 results and introduce the third-derivative term for teams with four or more data points. Watch this space.
Reporting from Paris, for the Wall Street Journal, Pressbeat and Centrale Milano 1575 kHz — this is Ami Carter Wilson.