There’s a common consensus that the 2015 English premier league is unusually unpredictable. And to demonstrate this – an examination of the difference between teams’ actual points and those expected by the betting market reveals a number of teams that have significantly over or under performed.

Using max odds from football-data.co.uk, expected points for a team in a particular match = (market win probability)*3 +(market draw probability)*1

Unsurprisingly this season’s two big stories, Leicester City and Chelsea, top and tail this particular table. But sitting second, just below Leicester, is the interesting case of West Ham United. The Hammers’ results this season have been singularly capricious – generally good with spectacular successes at Arsenal, Man City and Liverpool but counterbalanced by defeats at home to Bournemouth and away to Newcastle.

Another puzzling characteristic of West Ham this season is that the betting market doesn’t appear to like them. This is evidenced by the difference between actual and expected points in the chart above. I’ve also noticed it when calculating individual match probabilities using team ratings based on long-term goals and shots on target scored and conceded. For 17 out of 22 West Ham matches so far this season I’ve valued their win probability at 2% or more than the market does.

Looking at the two defeats mentioned above, in both these matches there was clearly a strong weight of money against them. Archived odds on oddportal.com show that the closing prices for West Ham vs Bournemouth in August on the Matchbook Exchange were:

A price of 2.86 for a winless, newly promoted away team against an established Premier League team was remarkable. For example, this type of price for an away team could be expected in a match such as Swansea vs Spurs.

We saw a similar effect on Saturday when the Hammers’ odds drifted from 3.00 (which I thought was already too high) to 3.45 against Newcastle.

What is it that the market dislikes about West Ham? It might be that the market believes that their results this season have been lucky – and stats show that they have the 5^th highest goal conversion rate and 3^rd highest save percentage of shots on target. These may be viewed as unsustainable. And certainly at the start of the season, West Ham’s goal conversion rate was unsustainably high – but as the chart below shows this is now regressing to the mean.

Whatever the reason – the market clearly doesn’t rate West Ham as a 6^th placed team.

And, just to note, West Ham’s best odds to win against Man City on Saturday are currently 5.36, this implies a probability of 18.7%. I value their chances as 23.5% – so again apparently undervalued by the market.