Expected Points Differences: Premier Leagueby Mike Holden / 07 October 2016, 12:00
In the fourth and final feature on manager's Expected Points Differences Mike Holden (@Ratings_Mike) takes a look at the Premier League.
(For a full description of the Expected Points Differences concept check out the information beneath the graphic below)
Slaven Bilic is the new favourite in the Premier League Sack Race market, following the dismissal of Francesco Guidolin, and the Expected Points table shows why the heat is being turned up on the former Croatia boss. The Hammers have tasted defeat in five of their first seven matches, leaving Bilic five points shy of where he really ought to be.
The move to the London Stadium appears to have had an unsettling effect on the whole club, upsetting the equilibrium in various ways. But Bilic is the one who takes the brunt and we’re fast approaching the point where dismissing the manager becomes a nagging temptation, like hitting the reset button on a computer game that has started badly.
As a benchmark, the sacking of Guidolin does Bilic no favours and those who think Swansea were a little hasty to dispense of the Italian are supported by the numbers on display here. The former Udinese boss was without a win in six attempts but the last five matches have been against four teams who finished in the top six last season, plus Chelsea.
Essentially, the 2-0 home defeat to Hull on the second weekend is the only unexpected outcome and that one result alone - accounting for 1.89 points of his deficit - ultimately did for him. With an overall shortfall of just 2.53 points, Guidolin’s sacking is harsh to say the least. The five departures in the EFL have averaged -6.08 points with an extra four matches grace.
It should be no surprise to see both David Moyes and Mark Hughes in arrears and, therefore, in trouble. Moyes has made no discernible impact since his appointment at Sunderland on the eve of the campaign and some frank assertions in post-match interviews that serve only to undermine confidence offer little to suggest a corner will be turned.
Meanwhile, Hughes, more than three years into the job at Stoke, has no excuses and needs a win quickly. The Potters play Sunderland in their next outing with Hull, Swansea and West Ham to follow. One assumes that six points will be the minimum requirement from those matches, although defeat at home to the Black Cats might abruptly curtail any opportunity to find out.
The top six contains few surprises and contains an uncanny resemblence to the Sack Race Power Rankings with Mauricio Pochettino edging in front of Pep Guardiola as the division’s highest achiever following Sunday’s thoroughly-deserved 2-0 victory for the Londoners at White Hart Lane.
Mike Phelan is in credit after a positive start under testing conditions at Hull but his position remains unclear, despite reports a fortnight ago that he was set to be offered a permanent contract. Failure to win either of the next two matches - against Bournemouth and Stoke - would probably see Phelan plunged into negative equity, in which case he would be under serious threat.
Expected Points Difference - What is it?
Expected Points is a way of converting bookmakers’ match prices into predicted outcomes by turning percentage probabilities into points values. It’s a helpful method of measuring performance versus expectations because it puts a cap on how much samples of data can be skewed by outliers, such as a freak 16/1 winner that subsequently needs 16 poor results to be cancelled out.
Example: if Team A is deemed to have a 52 per cent chance of winning before a match and the draw is a 28 per cent chance, their Expected Points value is 1.84 (0.52 x 3, plus 0.28). Likewise, if Team B has a 20 per cent chance of winning, their Expected Points value is 0.88 (0.20 x 3, plus 0.28).
If the match then finishes a draw, the result is worth -0.84 points to Team A and +0.12 points to Team B. This can be an excellent way of measuring pressure on managers because bookmakers’ prices tend to be a reliable gage of what they should be achieving. Consistently, those towards the bottom of the table tend to be the ones out of a job first - or next!