A football analytics blog

Manager Survival in the English Premier League

In the highly competitive arena of the English Premier League a good manager is an essential ingredient for success; this is recognised in their rewards, and top managers like Jose Mourinho can command a salary of over £8m.  But the demands of the job are high and there is no excuse for failure; it seems these days that the only remedy for a run of poor results is to sack the manager. So the life of a Premier League manager is limited.

But how long can a manager expect to survive?

In this study I will look at the careers of the 168 managers – excluding caretaker managers – who have managed in the Premier League since its inception  and who between them had 269 management spells in the league. Obviously survival in the short term is strongly influenced by the current run of results, but here I want to focus on the attributes of the managers themselves. Do managers from the British Isles outlast foreign managers? And is there any difference between younger and older managers?

First, some demographics.

Age Distribution of Premier League Managers

The table below shows the age distribution of Premier League managers when they are hired.

Age Band % Hires
under 35 6%
35-40 10%
40-45 23%
45-50 26%
50-55 19%
55-60 10%
over 60 6%

The average age of a new Premier League manager is 47, and just under 50% are hired in their forties.

National Origins of Premier League Managers

As we can see from the next table, around half the managerial appointments in the EPL were taken by Englishmen.

Nationality No. of Managers No. of Spells
England 81 137
Scotland 24 36
Italy 10 12
France 7 9
Spain 7 9
Netherlands 6 9
Northern Ireland 6 12
Republic of Ireland 6 11
Wales 4 11
Argentina 2 3
Croatia 2 2
Germany 2 2
Norway 2 2
Portugal 2 5
Brazil 1 1
Chile 1 1
Denmark 1 1
Israel 1 3
Sweden 1 1
Switzerland 1 1
Uruguay 1 1

At the inception of the Premier League virtually all the managers came from the UK or Ireland, with around 75% of appointments going to English managers; but the proportion of English managers has been steadily declining ever since, and as we can see in the chart it is now hovering around the 30% mark.

Decline of English Managers in the EPL 1992-present day

decline english mgrs

This is not an inevitable consequence of the internationalisation of elite football; since 2010, the percentage of home-grown managers in La Liga has averaged 68%, and in the Bundesliga 72%. As Jason Burt

Survival Rates in the Premier League

In this analysis, I am mainly interested in the length of time managers stay in post, and I don’t distinguish between a manager being sacked, or resigning because of poor results – or taking up a job at a bigger club.

Historically, the median survival time for a Premier League manager is 679 days.  In other words, a manager who has been in the job for 679 days has a 50-50 chance of being sacked or otherwise leaving (Kaplan-Meier estimation). The chart below shows how the probability of leaving increases as a function of days in the job. The red dotted lines show the median point, and the red crosses indicate “censored” data (i.e. data points corresponding to managers in place, but which can still be used in the analysis.)

Survival in the Premier League – Overall

Basic survival plot

As we can see, the longer in post, the higher the probability of being fired or leaving.  There also seem to be a change in slope at around 1500 days (4 years), which I’ve marked with a blue point; after this time, the risk of leaving declines more slowly with time.

Survival by Era

However, in recent years, management tenure has been getting shorter. The next analysis shows the survival rates for three Premier League eras; managers hired before 2005, managers hired between 2005 and 2010, and managers hired after 2010.

Survival in the Premier League- Then and Now

survival times by era

We can see that the probability of being fired or resigning after any given time on the job increases across these three eras.  The blue curve which is pre-2005 is flatter than the other two, which means the probability of leaving at any given time is lower. (The difference turns out to be significant on a Cox proportional hazards model). Managers hired before 2005 had a median survival time of 944 days, or just over 2 ½ years; managers hired in the next five years could look forward to only 594 days or about 20 months; and by 2010 this had reduced to 524 days or just under 17 months. A manager hired nowadays has a  50-50 chance of being sacked before October in his second season!

Another way of picturing this is to think of the “churn” or percentage of managers who change each season (excluding relegated clubs of course). Before 2005, the average churn rate was about 28%; for the last few years it has been close to 60%

Survival and National Origin

The next factor I looked at was nationality. I found little difference between managers from Europe and South America (although there were only a few South Americans). But as we can see in the next chart, English managers, and managers from the rest of the British Isles (blue and red curves) tend to significantly outstay managers from abroad (the green curve).

                           Survival by National Origin

survival by region

Managers from England and the rest of the British Isles have almost identical median survival times of 781 and 751 days respectively (just over 2 years) compared to only 524 days or 17 months for their foreign counterparts.

Survival and Age

Finally, what about age?

I tested the effect of age using a Cox proportional hazards model, and found a significant effect;  each additional year of age increases the risk of leaving by about 5%.

In the chart below, the survival curve for managers hired when they were over 45 is shown in red, and for those under 45 in blue.  We can see the leaving probabilities of the older managers are substantially higher.

survival times by age hired

The median survival time for a Premier League manager under 45 is 988 days; that’s almost twice as long as for a manager over 45 (555 days).

So if stability is a consideration, clubs looking to plan for the longer-term would do well to consider young up-and-coming talent from close to home; they are likely to stick around for longer.


Leave a Reply

Your email address will not be published.