EPL Prediction Model: Wk 16 Forecasts | City to beat Arsenal & Spurs over Liverpool

EPL Prediction Model: Wk 16 Forecasts | City to beat Arsenal & Spurs over Liverpool

Well last week was definitely a roller-coaster. Almost brilliant. So close! But ultimately disappointing.

The model made two very big calls last week; it said Manchester United could lose at home to Newcastle and that Southampton had a good chance of beating Manchester City at St Mary’s.

Man United first and while the model gave some people a cracking start to the weekend (says Twitter) as they backed Newcastle, I decided to wait and see if van Persie was starting. Newcastle were predicted to win because a Manchester United line-up without van Persie and Rooney had very little goal threat in it. With van Persie though, the odds swung back.

Premier League Predictor Results

It’s rare for one player to make this much difference to the model but last week it was the case that without him, simulated games couldn’t work out where Man Utd’s goals were going to come from, so Newcastle were predicted to win.

The starting line-ups were announced and van Persie was in, so I backed a home win. Big mistake. I’m claiming a moral victory on this one, on the basis that it wasn’t a fully fit van Persie and he did get the ball in the net even though it was off-side. Still, as I said, disappointing.

The second controversial prediction was Southampton to win and although the game ended as a draw, the model does emerge with some credit here. The bookies had Man City as clear favourites in this game, while the model favoured Southampton and the home side could easily have emerged with the win. They didn’t though. Again, disappointing.

The rest of the results were something of a wash-out, leaving us well down and wiping out the gains from good performances over the past three weeks. It is still early in the season and we’ll plough on! There will definitely be better days.

Here are this weekend’s percentages:

14-12

And the deeper possession and shooting stats.

deeper 14-12

A couple of these are worth some discussion. We’re predicting a very comfortable Man City win, because as far as the model is concerned, City will score freely at home no matter who they’re playing. They’ve hammered Spurs, Newcastle and Manchester United, so the model can’t even see that they might perform worse against ‘better’ teams. Based on the season so far, there’s been little in the statistics to suggest that Man City can’t put goals past anybody, leading to the numbers you can see above. I definitely have my doubts and Manchester City’s scoring rates should probably be moderated downwards but as always, you get the numbers straight from the simulator with none of my amateur opinions layered on top.

Fulham are given no chance, because the model has finally caught up with how badly they’ve been playing, just as the performances start to improve! I’ll be interested to watch this one but am definitely backing the home win.

Liverpool without Gerrard see their first predicted loss for a while, away at Spurs. Again, let’s wait and see.

Here’s who I’ll be backing.

Manchester City v Arsenal – Home Win
Cardiff City – West Bromwich Albion – Home win (Just. This one’s really close)
Chelsea v Crystal Palace –Home win
Everton v Fulham – Home win
Newcastle United v Southampton – Home win
West Ham United v Sunderland – Home win
Hull City v Stoke City – Draw
Aston Villa v Manchester United – Away win
Norwich City – Swansea City – Draw
Tottenham Hotspur – Liverpool – Home win