Sixteen of the 48 teams have been eliminated, and the remaining sides will compete to reach the World Cup final at MetLife Stadium in New York City on July 19.

The next round of 16 matches will see England face DR Congo, while reigning champions Argentina take on the unexpected Cape Verde.

However, a different team produced the biggest upset in the final group stage matches.

BBC Sport’s football expert Chris Sutton and an AI model both predicted a Cape Verde loss to Saudi Arabia, yet in the BBC’s predictor game, 39% of participants chose a draw and 43% backed the African side to win.

Instead, Ecuador surprised everyone.

More than 56,000 participants predicted the Ecuador‑Germany match, but only 597 correctly selected Ecuador to win.

Both Sutton and the AI were incorrect on this occasion, but overall they each achieved a 15‑out of 24 correct prediction rate from that set of matches.

In contrast, participants performed better with a 17‑out of 24 success rate, and they have already predicted the outcome of 72 of the 104 World Cup matches played so far.

Sutton has been correct 41 times (57%), the AI has 43 correct predictions (60%), while participants are leading with 48 correct predictions (67%).

Prior to the tournament, Sutton forecasted the order of finish for each of the 12 groups; he correctly identified the winners of seven groups and accurately ordered all four teams in three groups.

The AI’s predictions were generated using Microsoft Copilot Chat, which was asked to ‘predict the results of the World Cup last‑32 ties.’

Source link

Exit mobile version