Predicted Edge

Japan

J1 League predictions

Every J1 League match in the next 14 days with the model's probabilities, the likeliest score and one headline call, published at least 24 hours before kick-off. The table, team statistics and results are on the J1 League page.

4 matches in the next 14 days, every one with a forecast published. The strongest favourite is Vissel Kobe, 55% against Urawa at home; the most evenly balanced game is Machida Zelvia v Sanfrecce Hiroshima (Machida Zelvia 40%, draw 28%, Sanfrecce Hiroshima 32%); the fixture expected to produce most goals is Gamba Osaka v Cerezo Osaka, at 2.92.

Every match, one prediction

1 more member pick on these J1 League matches, with the price, stake and reasoning, from the moment each was published. £4.99 a month.Unlock every pick

Match by match

Vissel Kobe v UrawaSun 8 Nov, 05:00

Our model makes Vissel Kobe 55%, the draw 24% and Urawa 21%. The likeliest score is 1-1 (11%), 2.85 goals are expected, over 2.5 goals 54%, both teams to score 54%. The headline call is Vissel Kobe to win at 55%. Vissel Kobe are 5th on 26 points from 13; Urawa are 9th on 21 points from 13. Vissel Kobe have won two of their last five; Urawa have won three of their last five.

Machida Zelvia v Sanfrecce HiroshimaSun 8 Nov, 05:00

Our model makes Machida Zelvia 40%, the draw 28% and Sanfrecce Hiroshima 32%. The likeliest score is 1-1 (13%), 2.57 goals are expected, over 2.5 goals 47%, both teams to score 53%. The headline call is Machida Zelvia to win at 40%. Machida Zelvia are 4th on 28 points from 13; Sanfrecce Hiroshima are 3rd on 29 points from 13. Machida Zelvia are unbeaten in five (three wins, two draws); Sanfrecce Hiroshima have won four of their last five.

Gamba Osaka v Cerezo OsakaSun 8 Nov, 06:00

Our model makes Gamba Osaka 43%, the draw 26% and Cerezo Osaka 31%. The likeliest score is 1-1 (12%), 2.92 goals are expected, over 2.5 goals 56%, both teams to score 59%. The headline call is both teams to score at 59%. Gamba Osaka are 16th on 12 points from 13; Cerezo Osaka are 12th on 14 points from 13. Gamba Osaka have won one of their last five; Cerezo Osaka are without a win in five (three draws, two defeats). One of the sides' last two meetings was drawn.

JEF United Chiba v Mito HollyhockSun 8 Nov, 08:00

Our model makes JEF United Chiba 42%, the draw 26% and Mito Hollyhock 32%. The likeliest score is 1-1 (12%), 2.78 goals are expected, over 2.5 goals 53%, both teams to score 57%. The headline call is both teams to score at 57%. JEF United Chiba are 18th on 10 points from 13; Mito Hollyhock are 19th on 9 points from 13. JEF United Chiba have won one of their last five; Mito Hollyhock have lost all of their last five. Two of the sides' last two meetings were drawn.

The numbers this season

J1 League has averaged 3.00 goals a game over 132 matches this season, with over 2.5 goals in 59% of them and both sides scoring in 56%. Home sides have won 45% of matches, 23% have been drawn and away sides have won 31%. Those rates are the base the model's probabilities are judged against: a call only stands out where it beats what the competition normally produces.

About these predictions

How are J1 League predictions made?
A Dixon–Coles adjusted Poisson model rates every side's attack and defence from recent results and expected goals, with home advantage estimated from this season's matches, and turns the two sides' ratings into a probability for every scoreline. Match result, total goals and both teams to score are all summed from that one distribution, so they cannot contradict each other.
When are J1 League predictions published?
At least 24 hours before kick-off and up to eight days ahead, as soon as both sides' recent form is on record. Nothing is edited or removed after kick-off; the dated pages keep every forecast as it was published.
How often do home teams win in J1 League?
45% of matches this season, over 132 played, with 23% drawn and 31% won by the away side. Matches have averaged 3.00 goals, and over 2.5 goals has landed in 59%.

The method in full is under how it works; the model is explained on its own page. A high probability is not a certainty: a 70% call is expected to be wrong three times in ten.