Mito and Shimizu split by one point, with no clear favourite

The state of play

Mito Hollyhock are 13th on nine points after eight games. Shimizu S-Pulse sit one place higher in 12th on 10. On Matchday 9, one result could reorder the pair.

Mito have won two, drawn three and lost three, and their last five league results read L-L-D-W-D. Shimizu have won three, drawn one and lost four, with a sequence of W-W-L-D-L. Neither side has a run that justifies the word streak.

The goal columns separate them. Mito have scored 11 and conceded 11, a goal difference of zero. Shimizu have scored six and conceded eight, a difference of minus two.

Head to head

Shimizu have won both previous meetings and Mito have not scored more than once. Shimizu won 3-0 on 3 September 2016 and 4-1 on 13 October 2010, so the two games averaged four goals.

Both of those matches were played at Shimizu's ground. This time Shimizu are the visitors, a setting the earlier results say nothing about.

The numbers behind it

Over the last 10 matches Shimizu have the better record: five wins, one draw and four defeats, worth 1.6 points per match. They have scored 10 and conceded 10. Mito have won two, drawn three and lost five, for 0.9 points per match, scoring 11 and conceding 14.

Expected goals show a different pattern for each side. Mito have scored 11 against 10.2 expected, so they have slightly outperformed their chances. They have also conceded 11 against 13.4 expected at the other end. Shimizu have scored only six from 10.8 expected, a shortfall of 4.8 goals, and have conceded eight against 13.9 expected.

Put simply, Shimizu have created chances that have not turned into goals. Their attack has been the more wasteful of the two, while their defence has allowed fewer goals than the chances it has faced.

Our view

The estimates barely separate the sides. The betting market implies about 35% for a Mito win, 29% for a draw and 36% for a Shimizu win. Our model leans very slightly towards the hosts at 36%, with 30% for the draw and 35% for the visitors. That gap is a single percentage point either way and well within the uncertainty.

On goals, our model puts the chance of more than 2.5 at 42%, against about 46% implied by the market. For both sides scoring, the model says 49% and the market 52%. Both lean slightly towards a low-scoring match, but football is high-variance and these are estimates, not predictions.