SSA Blog: Stick or Twist? When to sack a manager

Leicester City head to Anfield tonight without Claudio Ranieri for their first game since his departure. He becomes this terms fifth Premier League manager to lose his job, following a disappointing defence of last season’s incredible title win.

Ranieri’s departure has divided football. So, should Leicester have stuck by their man or were they right to recognise a performance issue that support and loyalty wasn’t going to fix?

A number of studies have examined the impact of managerial changes across professional sport. Perhaps the most comprehensive in English football was published by Audas et al, in 2002. They examined results and managerial changes from every professional game for almost 20 years. Their model was able to assess the impact of short-term changes, meaning those which occurred within the season, rather than just season to season variations. Interestingly, they found clubs who made changes mid-season actually fair worse than those that don’t.

Last season, 12 premier league managers were sacked, 9 during the season, including 3 managers of 2 eventually relegated clubs, Newcastle and Aston Villa.    

Using last season’s example it can be argued, in conflict with Audas and his team, that a managerial switch will work. Swansea, Chelsea and Sunderland were all threatened with relegation when a change was made. If the objective was to avoid relegation, then these changes have to be seen as successful, if only in the short term.

A University of Warwick analysis by Bridgewater (2009), found that managerial changes do create a short term ‘bounce’ on team results. He attributed this to players attempting to impress a new manager to secure future employment. The paper described a ‘honeymoon period’ that would last between 12 and 18 games before performance regressed to pre-change levels.

However, it’s easy to review performance after the event, drawing conclusions about long vs short term benefits. When a club is facing relegation from the increasingly lucrative Premier League as Leicester City is, is there any research to support the decision-making process?

One study which attempted to develop an evaluation model was published by Chris Hope in 2003. Hope’s study, ‘When should you sack a football manager?’ identifies three key variables a club must consider when making a decision regarding the manager’s future:-

  1. Honeymoon – the period during which the club will not consider sacking a manager
  2. Trapdoor – average number of points per game that is expected
  3. Weighting – importance of recent results versus previous performance

The author presents a mathematical model to be used in real time with the above parameters adjusted according to a club’s performance objectives. Hope acknowledged a number of limitations with his model but it marked a quality step into a multifaceted issue that any empirical evidence base would help support.

Bell et al, (2013) published ‘The performance of football managers: skill or luck?’ Building on Hope’s 2003 work, Bell’s model is a complex equation that takes into consideration performance, while also factoring the increased financial influence of the game. This ensures it doesn’t favour managers presiding over expensive squads. Their model uses the flowing six criteria:-

  1. Total player wage bill
  2. Total net transfer fund
  3. Total number of injured players
  4. Total number of suspended players
  5. Total number of unavailable players (e.g African Cup of Nations)
  6. Total number of non-Premier League games  

Leicester City increased their transfer spending and wage bill by 82% and 37% respectively on the previous season and while player availability has been reduced, the correlation (.75) between wages and points would more than compensate for this.

When we examine Leicester City’s performances this season, particularly with the weighting given to the last 5 games (as proposed by Hope, 2003), the numbers show that the decision to sack Ranieri was the correct one. According to Hope (2003), a rating below .74 points per game (when adjusted) should result in the sack. Ranieri’s rating was .66……

Those that believe Leicester City should have stuck by Ranieri are arguing for a season long ‘Honeymoon period’ following the title win. This is a reasonable position to take, particularly for a fan. But in the increasingly commercialised Premier League, the numbers justify the owners decision to let Claudio go.