Winless Tokyo Verdy host Sanfrecce Hiroshima, who have conceded just six in eight

The state of play

Tokyo Verdy are 19th in the J1 League with four points from eight matches: no wins, four draws and four defeats. Their sequence reads L-D-D-L-L, and their goal difference is minus eight.

Sanfrecce Hiroshima arrive in fifth with 17 points, 13 more than their hosts. They have won five, drawn two and lost one, and their sequence is W-W-L-W-D. A goal difference of plus 14 is built on 20 scored and only six conceded.

Over the last 10 matches the contrast holds. Hiroshima have won seven, drawn two and lost one, collecting 2.3 points per match. Verdy have won one, drawn five and lost four, a return of 0.8 per match.

Head to head

Hiroshima have won eight of the nine meetings, Verdy none, with one draw. The most recent was a 3-0 Hiroshima win on 23 August 2025.

Hiroshima also won the three before that, 2-1, 1-0 and 4-1. Meetings average 2.67 goals, so the fixture is rarely tight on the scoreboard even when the winner is the same.

The numbers behind it

Verdy's attack is the first problem. They have scored four goals from 7.1 expected goals across eight matches, so the low output reflects the chances they create rather than bad finishing alone. At the back they have conceded 12 from 12.1 expected, which also tracks closely.

Hiroshima's profile is the reverse. They have scored 20 from 23.4 expected goals and conceded six from 5.5, so the underlying numbers support the results. In the last 10 matches they have scored 28 and let in seven.

For Verdy, the same period brings eight goals scored and 13 conceded.

Our view

The betting market implies about 64% for a Hiroshima win, 21% for a draw and 15% for a Verdy win. Our model leans higher on the visitors at 70%, with 18% for a draw and 12% for the hosts.

The gap is modest and follows the numbers above: Hiroshima's expected-goals record is strong at both ends, while Verdy's is weak at both. On goals, the model puts over 2.5 at 59% against about 56% from the market. Football is high-variance, and these are estimates, not predictions.