Jose Mourinho was named LMA Premier League Manager of the Season a few days ago, an award to add to his similar award from the Premier League itself. These awards seem to be based on the subjective opinion of those voting on them so sometimes they go to, as in this case, the manager of the Premier League champions. In other seasons, they go to a more unsung name like Tony Pulis who won both awards in 2013/2014. Isn’t there a more objective and consistent way to judge a manager for this award though?
An objective method
Throughout the 2014/2015 season, I have tweeted the results of my own method to assess the best managerial performance of the season. It is not meant to be perfect but is simply designed to add to the debate. None of us truly know what a manager does but his team is supposed to gain as many points as possible. Therefore it seems reasonable to compare the number of points the team achieves with a baseline provided by a model describing a team’s strength. The Euro Club Index does just that and throughout the season I have compared the expected points derived from the pre-match odds with the actual points. The table below shows all of the Premier League managers this season on this measure:
Jose Mourinho and his team did indeed overperform on this basis as he is one of eight managers (one appears twice) to have achieved at least 2.5 points more than the odds expected based on the relative strength of their teams and opponents throughout the 38 Premier League matches of the 2014/2015 season. It is interesting to note that the managers of the three relegated clubs all performed more or less as expected on this basis despite going down. The winner here though is Alan Pardew at Crystal Palace whose team took nearly 11 points more than expected under his watch. Newcastle United, who Pardew managed for the first half of the season, also did better than their odds expected. However, at Palace, Pardew may have profited overly from a team rated too low due to their previous manager’s results underforming their underlying strength, as other incoming managers also may have done.
However, is this measuring what we want it to measure? As it is a model with no adjustment for anything except results, it will work better than a bookmaker’s odds which have already adjusted for various factors including transfers made. These things are surely part of a manager’s skill and should not be adjusted away. It therefore seems better to go back to the beginning of the season if the aim is to measure a genuine overperformance as those forecasts will not have been adjusted for results as we go along.
Using the same Euro Club Index model and compare its forecast points at the beginning of the season with the points achieved at the end, we get this result:
Fourteen clubs retained their manager for the entire season and these are topped by Southampton who achieved 13 points more than their pre-season forecast of 47. However, six clubs changed their manager so they needed to be taken account of too. In order to do this, a predicted points total per match was calculated based on the pre-season forecast and then multiplied by the number of matches the manager was in charge for. For those managers with very few games in charge – like Dick Advocaat and Tim Sherwood – this may create a problem due to potentially lop-sided schedules for so few matches but it seems a reasonable enough approximation. Alan Pardew again emerges on top, taking over 13 points more than expected at Crystal Palace and beating Ronald Koeman into second place. Pardew’s victory is confirmed even more strongly under both measures by the fact that he returns a positive score at Newcastle.
So, Scoreboard Journalism’s Premier League manager of the year for 2014/2015 is Alan Pardew. Can we have a vote and a table at next year’s ceremonies please?