"I think the Norwegian Computing Centre is calculating the wrong way," said Stale Solbakken. Not exactly the quote you expect from a coach whose team just demolished Iraq 4-1 in their World Cup opener.
The Norwegian Computing Centre ran 100,000 simulations — updated every 10 minutes — and spat out a 99 percent probability that Norway advances to the knockout round. Solbakken isn't interested. He wants a point from either Senegal or France before he'll believe it.
His caution isn't paranoia. It's recent memory. Sweden hammered Tunisia 5-1 in their opener, then promptly got taken apart by the Netherlands by the exact same scoreline. They're now third in their group. One big win can make a model look very stupid very fast.
The 99% that's wrong 22% of the time
Researcher Torstein Maeland Fjeldstad, who built the model alongside Alexander Johan Arntzen using nearly 30 years of football data, was candid about the limits of his own figures: "If we say that Norway is going to advance, we are also 'wrong' 22 percent of the time."
That's a 1-in-5 miss rate dressed up as near-certainty. Solbakken clearly did the same mental arithmetic.
The model is built on form, historical data, and was — by Fjeldstad's own admission — "positively surprised" by the 4-1 result. Which means it's now extrapolating from a single match performance to near-guaranteed progression. Anyone who has watched tournament football for more than a week knows how that can unravel.
Norway's road from here
Norway face Senegal next on Monday before closing out Group I against France on June 26. Neither fixture is a gimme. Senegal are a structured, physical side, and France are France.
Solbakken, who played for Norway at the 1998 World Cup — the last time they appeared at the finals — knows exactly what getting ahead of yourself costs at this level. Italy knocked them out in the last 16 that year.
Norway's qualification odds may look appealing on paper right now, but markets built on a 4-1 opener against Iraq are doing the same thing as that model: projecting forward from a sample size of one. Solbakken is right to be cautious. The group isn't done.
