The Final Games of the First Round -- World Cup Model

Hands up, we got to the longform a little late today. The good news is the picks themselves were never late. They went out on Instagram and TikTok before kickoff, same as always, so if you follow along there you already had the whole card in hand. This is just the why behind it, the part that does not fit in a caption.

Where our accuracy actually sits

Twenty games in, here is the honest scoreboard on how often the model’s call lands. On total goals, the over-under read, we are hitting 70 percent. On the match winner, we are at 55 percent, which is roughly where a sharp market sits on a three-way result with this many draws in it. Totals is clearly our engine, and it is the market we lean on hardest. On the money side, our published ten-point edges are running 13 and 5 for a return of about 64 percent. Small sample, the usual caveats, but the shape of it is good and the part we are best at is the part we publish most.

The big model has its two favorites now

A week ago the title race at the top of our model was a cluster. It is not anymore. Argentina and France have pulled clear of the field. Argentina sit around 21 percent to win the whole thing, up from 14 at the start, after that 3-0 dismantling of Algeria. France are next at roughly 14 percent on the back of their 3-1 over Senegal. The interesting wrinkle is that our model was actually skeptical of Argentina in that match, it had them under 50 percent on the day, and they answered by winning comfortably and dragging their tournament number up with them. Below those two it is still bunched, but the gap from the top two to the rest is real now in a way it was not a few days ago.

Today we get the full picture

This is the day the first round closes out. Once Portugal and England play, every one of the 48 teams will have a game on the board, and we will have a complete first-round read on the entire field for the first time. That matters for the model, because up to now Groups K and L have been pure projection with no results to lean on. By tomorrow morning, every team will have shown us something.

Today’s four games, in kickoff order

Portugal vs DR Congo, 1pm ET

This is the biggest edge we have posted all tournament. The market makes Portugal a 76 percent favorite. Our model has them at just 57, and more to the point it expects a slog, a projected 1-0, with the play being the Under 2.5 goals at a 36-point edge. Portugal is the third elite favorite in a row our ratings simply do not buy at the market’s price, after Spain and Argentina. We do not bet the draw here even though the model likes it, because our rule is that we do not fade a heavy favorite in the winner market, so the draw sits benched and the Under is the play.

England vs Croatia, 4pm ET

The market has England as a solid favorite. We have this closer to a coin flip, with England barely ahead of Croatia, and at the price our model likes Croatia to win outright. It is the one winner-market edge on the card, and it is a bet that the market is leaning too hard on the England name.

Ghana vs Panama, 7pm ET

A rare one for us, because we want goals here. Most days our edge points to the under, but the model reads this as an open, end-to-end game and lands on the Over 2.5 at a healthy edge. When the goals model points up instead of down, we follow it, and today it points up.

Uzbekistan vs Colombia, late

No play. Colombia are heavy favorites and our model agrees with the market almost exactly, so there is no gap to bet. Some nights the right move is to pass, and a card that invents an edge on every game is just selling you noise.

The card in one line

A massive Portugal under, a Croatia upset at the price, a rare Ghana over, and a pass on Colombia. Late to the writeup, on time with the picks, and by tomorrow the whole field will be on the board. Every price is logged. Enjoy the football.