Search Menu

Weekend Predictions: Manchester United to beat Chelsea, Everton to beat Manchester City?

Share

The English Premier League enters a pivotal phase this weekend, with several matches that could decide the fate of the title, the European place(s) and relegation spots in the next 4 days (if you include Tuesday’s match between Manchester City and Tottenham).

Last weekend we looked at the Castrol EPL Predictor and used it to predict results for all 10 Premier Leagues from last weekend. Let’s look at how the Castrol EPL Predictor fared:

Last Weekend’s Results

Actual match results followed by Castrol EPL Predictions in brackets)

Wigan 1-1 Everton (Everton win at 51%)
West Brom 2-1 Aston Villa (West Brom win at 43%)
Sunderland 0-3 Fulham (Sunderland win at 38%)
Chelsea 2-1 Tottenham (Chelsea win at 75%)
Blackpool 0-0 Stoke (Blackpool win at 44%)
Blackburn 1-0 Bolton (Blackburn win at 51%)
Manchester City 2-1 West Ham (Man City win at 61%)
Liverpool 3-0 Newcastle United (Liverpool win at 65%)
Birmingham 1-1 Wolves (Birmingham win at 51%)
Arsenal 1-0 Manchester United (Arsenal win at 40%)

Castrol Edge Logo Weekend Predictions: Manchester United to beat Chelsea, Everton to beat Manchester City?There’s two ways to look at this: one is look at absolute result predictions – i.e. how many results did the Castrol Predictor get right – 6 out of 10. The other way to look at is to consider that these predictions are actually percentages assigned to each outcome (win, draw, loss), and then look at how the actual results stacked up against the Castrol EPL Predictor’s percentages.

For example: Arsenal v Manchester United only had a 34% chance of United winning, or 66% chance of United not winning.

Applying this to the incorrect results from the first method, we see:

Wigan 1-1 Everton (Wigan to not lose at 49%)
Sunderland 0-3 Fulham (Sunderland to not lose at 66%)
Blackpool 0-0 Stoke (Blackpool to not lose at 69%)
Birmingham 1-1 Wolves (Birmingham to not lose at 78%)

In two results (Blackpool and Birmingham, the prediction was highly accurate. In one result (Wigan v Everton), the prediction was accurate but it was an evenly balanced spread with all three results equally likely. In the fourth result, Sunderland v Fulham, the prediction was way off the actual result, and this is reflected in the fact that Fulham’s away win to Sunderland came as the biggest shock of the weekend results.

So, if you look at these predictions strictly in terms of win percentages, the Castrol EPL Predictor got 6 out of 10 predictions correct.

If you look at these predictions in terms of overall percentages and the likelihood of home teams not losing, the Castrol EPL Predictor got 8 out of 10 predictions correct, with one call one the line and the last one a surprise result.

Castrol Weekend Predictions

This weekend’s matches as predicted by Castrol EPL Predictor:

WeeklyScore 060511 Weekend Predictions: Manchester United to beat Chelsea, Everton to beat Manchester City?

Talking Points:

  • Everton to beat Man City at 52%? I would have predicted Everton to not lose, but perhaps not that high a win percentage.
  • Tottenham, home to Blackpool, are the most likeliest team to win. Of course, the last team Blackpool beat was…Tottenham, in the reverse fixture. Odds on the Tangerines doing a double?
  • Arsenal are 61% favourites to beat Stoke away, having won their last four games against them.
  • The last team to beat Manchester United at Old Trafford was Chelsea, at the end of last season (in the game that ultimately decided last year’s league title). Manchester United are 37% to win, and 65% to not lose. A win will seal the title, a point should see them through too.
  • Fulham, fresh off a (surprise) 3-0 win away to Sunderland, are on a paltry 26% to beat Liverpool at home.
  • Premier League Final Standings Predictions

    Here’s a look at the end-of-season Premier League table as predicted by Castrol:

    FinalStandings 060511 Weekend Predictions: Manchester United to beat Chelsea, Everton to beat Manchester City?

    Talking Points:

  • Not much changed here from last week except for the relegation spots. Last week it was Blackpool, West Ham and Wigan (20th to 18th), this week (presumably with Wigan avoiding defeat last week) it’s West Ham, Blackpool and Wolves, with Wigan escaping on a better goal difference to Wolves.
  • I still don’t see how Tottenham can end the season ahead of Liverpool given that they have to play away to City and away to Liverpool, and two defeats there would waste whatever advantage their game in hand gives them. Still, if Fulham get a point against Liverpool and Tottenham get all three against Blackpool, then it should get interesting. Do Tottenham want to play in the Europa League though?
  • We’ll be following up with match previews for the big games this weekend – United v Chelsea, Everton v City, plus the relegation battle and the race for 5th place between Tottenham and Liverpool.

    Make sure you look at the Castrol EPL Match Predictor in more detail as it’s an excellent statistical tool for predicting matchups between two Premier League teams as well as our previous Castrol EPL Predictor article for an explanation of how the Castrol EPL Predictions work.

    Premier League Betting

    And if you’re a betting man, I recommend using our free £25 bet to back Castrol’s predicted results on Betfair. 6 out of 10 (or 8 out of 10 if you bet on a team not losing) is a good set of results to put ‘free’ money on (and you’ll be in with a shot at winning an extra £100).

    Comments (2)

    1. Castrol’s final table makes no sense…

      It has Chlesea is taking 5 points from 3 games, which must be a win a two draws, and Untied with 6 points from 3 games which is a loss and two wins. And since they play each other, and United don’t draw, that result must be a Chelsea win.

      But they also predict that a United win is more likely…

      So what’s missing here?

      • Simon, that’s a good point being raised.

        From what I’ve understood, the match predictions are essentially percentages for the different outcomes. So you have to keep in mind that while a United win is more likely, it’s only marginally more likely.

        The league table points are calculated based on averaging expected results over a period of time, which is why the points total sometimes doesn’t add up (it doesn’t add up for the relegation battle either if you see the associated article).

        Personally, these rankings and predictions aren’t there to give you a 100% accurate answer, more to give you a statistical analysis / model of what’s happened / will happen. It’s important to understand what data they’re using, and why that’s useful, and basically keep in mind that what’s predicted will not always come out to be true, but if it’s a solid algorithm based on relevant stats then it’s a very good guideline for what might happen.