By Graeme Riley.

At the beginning of September Zach Slaton produced a simulation of all of the results in the Premier League for the whole of the current season, including actual results to date and forecasts based on TPI data. With over six weeks having passed since we last looked at this, Zach has kindly re-run the data to see what, if anything has changed in the intervening period, the season now being almost one quarter completed.

Obviously as more results are now actual rather than forecast the picture is becoming clearer, and there are some fascinating conclusions to be drawn.

Title odds

At the beginning of September Chelsea were in the middle of a long unbeaten streak and their probability of regaining the title stood at almost 40%.  The two Manchester clubs were stuck back in the early 20s%. Besides these three, Arsenal were next, with less than a one in 16 chance of the title.

As Chelsea continued their good run, their probability of winning the title increased, and by the end of the weekend of 22 October they were a shade off Evens, with 49%. The two Manchester clubs had maintained their odds, whilst everybody else saw these three disappearing over the horizon, as even Arsenal had less than a one in 50 chance of coming top of the table.

Fast forward a week and one result changed things dramatically. With Chelsea’s home defeat, albeit controversially, to Man Utd, their odds shot out to 36.3%, and correspondingly Man Utd’s odds tumbled to one in three. Even Man City and Arsenal, who were not even involved, gained significantly thanks to their own victories.

Overall however it seems that there is a 93.5% chance that the title will fall to either Chelsea or one of the Manchester clubs.

The rest of this post is for subscribers only.

[ttt-subscribe-article]