The FA Women’s Super League (WSL) 2018/19 kicks off this weekend. So, we decided to forecast that as well.
Last season’s champions, Chelsea, are forecast to come up even at home to Manchester City, with the most likely scoreline being 1-1.
In the Women’s Championship, newcomers Manchester United are forecast to start their season with a 0-1 away win at Aston Villa.
The rest of our WSL forecasts are in the the table below:

  • The Forecast scoreline, with the % chance of that happening
  • The % chance of a win by the Home team, P(H), or Away team, P(A) (with one hundred minus those two numbers giving the % chance of a draw
Forecast Win (%)
Score Pr (%) P(H) P(A)
WSL
Arsenal Liverpool 2-1 9 67 13
Birmingham City Everton 1-0 12 53 15
Brighton & Hove Albion Bristol City 1-2 8 36 38
Chelsea Manchester City 1-1 13 30 37
Reading Yeovil Town 3-0 8 89 2
WC
Sheffield United Durham 1-2 8 30 48
Aston Villa Manchester United 0-1 12 11 63
Crystal Palace Leicester City 2-1 8 63 17
Millwall Lewes 2-1 9 73 10
Tottenham Hotspur London Bees 2-1 7 58 22

As discussed before (here, here & here), we continue to tinker with our forecasting method. Our aim has always been to provide a single scoreline forecast to hang our hats on. But it is challenging to balance both plausible result and score forecasts simultaneously, which will perform well against some metric. Therefore, we are now making “Fuzzy Conditional” forecasts, discussed here.
As an academic exercise, the WSL offers us more data to use in improving the Model. It gives us the opportunity of apply the Model in a closed league (no relegation from the bottom tier). There is also the challenge of league expansion, as the WSL has gradually grown and split into two tiers.
For the WSL, we estimate a reduced down version of the Model used for the men’s English Football League forecasts. This is akin to what we use to generate forecasts for men’s international fixtures. The Model estimates and updates the relative attacking and defensive strengths of each team, and uses Elo strengths also.
We collected all data on women’s football results since 2011 from the official WSL website, but only estimate over the last three calendar years (552 matches). We had to circumvent the significant number of name changes in recent years (ZZZ Ladies to ZZZ Women, ZZZ WFC to ZZZ Women, etc.) so that we could build in the histories of these teams – a name change doesn’t mean that the entire team is new.