⭐ Match of the Week: Manchester United W vs Manchester City W

Khadija Shaw — Manchester City W

Arsenal W vs Tottenham Hotspur W — Saturday 28 March, 5:30pm

Kim Little — Arsenal W

Our Predictions

Season record: Updated after GW19

Gameweek 19 could be the round that settles the WSL title race. If City win at Old Trafford they go 11 points clear with three games to play — effectively untouchable. Arsenal’s two games in hand remain their wildcard, but they must take care of business at home against Spurs first. At the bottom, Leicester’s wait for a win is becoming desperate with five games left and Brighton arriving on Sunday.

Best case for Arsenal W: A comfortable home win — with Kim Little pulling the strings in midfield and Arsenal closing to within six points of the top three before two games in hand can be cashed in.

Worst case: Tottenham, buoyed by Martin Ho’s new contract and freed from title-race pressure, leave the Emirates with a point or better — leaving Arsenal’s top-three push hanging.

Next Fixtures

🤖 BTP Machine Learning Model — WSL GW19

The BTP ML model generates probability estimates for every WSL fixture. The model is trained on 659 matches across five seasons (2020/21–2024/25), using rolling form, goals data, and league position. It outputs a probability for each outcome — home win, draw, away win — for every upcoming fixture.

For the full methodology, results and an honest assessment of what the model can and cannot do, see our WSL ML Model Explainer.

⚠️ Exploratory predictions only: WSL predictions are published as exploratory data rather than validated forecasts. The model does not currently beat the naive baseline on log loss — see the WSL ML explainer for full details. Probabilities reflect positional and form signals but should be interpreted with caution.

How to read these charts: Each bar shows the model’s probability for home win (pink), draw (grey), and away win (red). The longer the segment, the more likely that outcome. Position bands for this 12-team league: 1–3 title contender, 4–9 mid-table, 10–12 relegation. These are not betting tips — they are a data-driven perspective based on form and league position.

📊 ML Probability

Everton (6th, 20 pts) host Liverpool (10th, 14 pts) in a match the model gives a slight home edge — H 52.7%, D 21.7%, A 25.7%. Liverpool’s poor league position this season depresses the away probability; the gap in standings makes this look like a home banker on paper, though Merseyside derbies rarely follow the script.

📊 ML Probability

The headline match: City top (46 pts) away at United in 2nd (38 pts). The model gives City a narrow edge at 50.3% with an unusually high draw probability of 31.0% — reflecting how close these two sides are positionally. What the model cannot see: Man Utd played Bayern Munich on Wednesday and are missing Ella Toone through injury. Factor in your own assessment of fatigue.

📊 ML Probability

Arsenal (4th, 35 pts) host Spurs (5th, 29 pts) in what the model reads as the closest home advantage on Saturday’s card — H 51.0%, D 23.7%, A 25.3%. Tottenham’s strong fifth-place standing under Martin Ho compresses the probability spread. Note: Spurs defender Clare Hunt is out for knee surgery, weakening their backline.

📊 ML Probability

The model’s strongest call of the round: London City Lionesses (7th, 20 pts) away at West Ham (11th, 12 pts). The 60.0% away probability reflects a meaningful gap in league standing — West Ham are one of only three sides in the bottom four. Rita Guarino’s side have struggled badly for goals and points this term.

📊 ML Probability

The model’s most confident home prediction of the weekend: Chelsea (3rd, 37 pts) host Aston Villa (8th, 20 pts). H 64.0%, D 15.0%, A 21.0%. The very low draw probability signals the model sees clear daylight between these two sides — though Chelsea arrive here after a draining first leg loss to Arsenal in the Champions League quarter-finals.

📊 ML Probability

Leicester (12th, 9 pts, six defeats in a row) host Brighton (9th, 18 pts). The model gives Brighton a slight away edge at 46.7% despite Leicester having home advantage — the gap in points is clear. At 30.7% a home win is not out of the question and Rick Passmoor’s side are desperately searching for any foothold before the season runs out.

🤖 Model Notes — GW19

All six GW19 fixtures have live ML probability estimates. The most important caveat this week: the model uses only rolling goals, form, and league position — it has no knowledge of mid-week fixture scheduling. Manchester United played Bayern Munich in the Champions League on Wednesday and travel to Germany for the second leg next Wednesday. Ella Toone is also absent through injury. These factors are outside the model’s scope and could meaningfully affect the result. Similarly, Tottenham are without Clare Hunt (knee surgery) ahead of Sunday. Use the probabilities as a starting point, not a conclusion.

Model: Random Forest (300 estimators). Log loss 1.569 — does not beat baseline (1.052). Training: WSL 2020/21–2023/24 (493 matches). Full methodology here.