Netherlands holds both halves of the profile — 1.28 against 1.10 for scoring, 0.95 against 0.98 for conceding, where the lower number is the better one. On rating it is 1598 against 1512, which the model resolves to 49.6% / 25.0% / 25.4%.
These figures stand as of today; the model updates them at full time. Netherlands leads the distribution by 24.2 points over Japan. Scorelines above 8%: 2. Someone wins 75.0% of the time by this model. The model's preferred outcome is Netherlands at 49.6%. Just one prior encounter is on file — 2010, 1-0 in World Cup. Played on neutral ground, so no home bump applies. Rating separation comes to 86 points. Ranked against every team the model rates, they sit 385th and 1205th of 4,449. The three results cover 24.6 points. Japan's number leans on a 4-match sample. The model's expected-goals figures are 1.70 and 1.41. No scoreline stands out; 1-1 only leads on 10.7%. In absolute terms: Netherlands 1.28/0.95, Japan 1.10/0.98 on attack and defence. The match finished 2-2. Point estimate: 1-1. One sits in the top 8.7%, the other the top 27.1%.
Source: our elo-poisson-v1 model, run across the complete result record. Not betting guidance.