Close menu

Premier League: Chelsea to win by seven points? Arsenal outside top four?

Last updated on .From the section Footballcomments187

Chelsea to win the title by seven points? That is what SAM is predicting
Chelsea to extend their lead? That is what super-computer SAM is predicting

A fifth Premier League title for Chelsea, the end of Arsenal's lengthy run of top-four finishes, and heartbreak for Swansea, Sunderland and Hull.

They are the verdicts of SAM (Sports Analytics Machine), the super-computer built by Ian McHale, professor of sports analytics at the University of Salford, together with his colleague Dr Tarak Kharrat.

We asked SAM to predict the outcome of every Premier League match between now and the end of the season on 21 May. The results suggest Chelsea will not surrender their five-point lead, but will instead extend it. SAM calculates Antonio Conte's men have a 64% probability of winning the title.

The Blues' 13-match winning streak ended at Tottenham last week - and SAM calculates Spurs will finish second, one point ahead of Liverpool and Manchester City.

That means Arsenal would finish outside the top four for the first time since 1995-96.

Will your team finish in the top four?
% probability calculated by SAM
Chelsea 97.6%
Tottenham78.7%
Liverpool74%
Man City67.8%
Arsenal61.5%
Man Utd20.3%

But is SAM right? Do you think you can do better? Pick how you think the top 20 will finish and share your final table with your friends.

Pick your final Premier League table

There are 18 rounds of games left in the Premier League but how will the table look when the season finishes? Will Chelsea remain on top? Will Hull still be stuck at the bottom? Pick how you think it will look at the end, placing all 20 teams in order. Have a go then share with your friends.

SAM's final table in full

SAM predicts just one change to the top eight between now and the end of the season - with Spurs leapfrogging Liverpool to finish second. Before their win over Chelsea, SAM had Mauricio Pochettino's team down as finishing third.

Spurs, of course, slipped to third at the end of last season, so finishing as runners-up would represent their best finish in the Premier League.

There is no movement for the bottom five but Leicester, who are 15th, are predicted to climb five places, giving their title defence an ever so slightly more respectable appearance.

SAM gives Hull, who are bottom, a 97.2% probability of going down.

The final Premier League table as predicted by SAM the super-computer

SAM takes into account a wide range of factors to work out match results, looking at average performances so far and calculating what that means for the remaining fixtures. It is based on players remaining fit and continuing with their average performance levels, so an injury to Chelsea striker Diego Costa or Tottenham midfielder Dele Alli, for example, would have a significant impact on these predicted outcomes.

You can follow Professor Ian McHale on Twitterexternal-link

Comments

Join the conversation

These comments are now closed.

187 comments

  • Comment posted by Killingholme_Clay , at 08:22 11 Jan 2017

    The BBC's computer (Phil McNulty) had Man Utd 15 points clear by February.

  • Comment posted by conradk, at 00:35 11 Jan 2017

    Can't be any worse than Lawro's predictions in all fairness.

  • Comment posted by The Old Trafford Career Derailment Scheme, at 01:31 11 Jan 2017

    An octopus can predict better than a computer, so this is a waste of everyone's time

  • Comment posted by Adrian, at 08:08 11 Jan 2017

    Hilarious how many people are saying this article is rubbish and a waste of time. I'm sorry the mean computer calculated probability and predicted your team to go down/finish outside the top four.

  • Comment posted by kenny_burns_kneecap, at 09:08 11 Jan 2017

    SAM is a Sinclair ZX Spectrum

  • Comment posted by QWERTY, at 00:39 11 Jan 2017

    This so called wonder computer played Lawro Boxing Day last year. 1 correct score, 4 correct results from 10 matches. It does know that a match is played by humans, refereed by humans (sometimes) and played on grass!!!! I won't be rushing to the bookies.......

  • Comment posted by PGB Addick, at 08:49 11 Jan 2017

    You know what the funniest thing is.

    It says that until Spurs beat Chelsea SAM has Spurs finishing third.

    So in other words it just adjusts when it gets it all wrong, much as any football fan with a brain could do. Wonder what it will look like in 6 weeks

    That's not a prediction, just analysis

  • Comment posted by PranR17, at 01:14 11 Jan 2017

    United are just about winning matches to cling onto the top 4, but they haven't put down a marker. You can't always win narrowly, eventually you'll drop points in games you should have won, because you failed to kill off the opposition. Title winning teams are always capable of winning convincingly. Unlike the top 5 i'm yet to see that from United which is why I think SAM is right about them.

  • Comment posted by Dr_John - not a Doctor of medicine, at 08:21 11 Jan 2017

    @19-Godfrey Howard, just to clarify it is actually 20 years since Spurs finished above Arsenal.

    If there was a shred of accuracy in these predictions, Dr McHale would be using his simulations to clean up at the bookies rather than publishing them on the BBC website. If his simulation does a better job of calculating the odds than the Bookies, he has a license to print money.

  • Comment posted by Markyp1965, at 07:14 11 Jan 2017

    Having a 'super computer' to do the predictions might be the only chance that the Spuds have of finishing above Arsenal and they're gullible enough to believe it's true.
    Obviously, SAM doesn't take into account the 'bottle effect' in April and May.

  • Comment posted by PGB Addick, at 08:18 11 Jan 2017

    235 Well no I think it's because it's rubbish and no system can predict anything this complex with so many variables. The table is virtually the same and only takes into account what has happened and not what will happen.

    Then there's the little matter of saying that Hull only have a 2% chance of staying up. I'd take those odds for a bet in a heartbeat. Or that Utd only have a 20.3% for Top 4

  • Comment posted by Chris, at 10:52 12 Jan 2017

    To validate this prediction you would need SAM's prediction from this time last year (and any other previous years) to determine the accuracy of its forecasting.

  • Comment posted by Number17, at 11:36 12 Jan 2017

    So the possibility from this is that come the final game of the season, Spurs are 2 points clear of Arsenal. Both teams need a win. In my experience, this normally goes only one way.

    So I'll assume the position at 37 games has Spurs on 78pts already, so even they can't get it wrong this time. As a Spurs fan, i'll take that.

  • Comment posted by User0439025520, at 09:20 11 Jan 2017

    Does SAM factor in injuries, the annual February Arsenal meltdown, January transfers, managerial sackings, or slips in important games in April?

  • Comment posted by Codpiece, at 08:33 11 Jan 2017

    @28
    Last season SAM predicted: 1- Man City. 2- Arsenal. 3- Man Utd. 4- Chelsea. 5- Spurs. Leicester- 11

  • Comment posted by BaggiosPonytail, at 09:53 11 Jan 2017

    This doesn't make a lot of sense. It's basically suggesting that all of the top 6 sides are going to perform less well than they did in the first half of the season. You would probably expect it with Chelsea as if they maintained their current points per game they would end up with 93 points but I would have thought at least one of those teams would improve.

  • Comment posted by BlueStig, at 23:32 11 Jan 2017

    I thought the MU fans would be on here calling the computer an ABU.

  • Comment posted by Sparky, at 08:03 11 Jan 2017

    I'm sure Chelsea and Spurs fans would accept that position, if it was predicted correct.
    But as we all know, the probability of it going to end in that way is well off. As anybody can drop points on a day with unexplained circumstances, such as no matter how much they try, just can't get it in the back of the net from 20 odd shots!

  • Comment posted by D1675132991836, at 01:05 11 Jan 2017

    seems about right :)

  • Comment posted by Pete the Hawker, at 07:04 11 Jan 2017

    What we need to see is SAM's predictions last season at this stage and how its predictions fared with the final table - that may give this article some credibility :-)

Elsewhere on the BBC