Of all of the talking points from the FIFA Women’s World Cup so far, the debate surrounding Phil Neville’s willingness to rotate all of his England squad including the goalkeepers is one with potential implications beyond the women’s game but for Football as a whole. The policy has drawn criticism from the likes of former USA goalkeeper Hope Solo, yet has gained support from her fellow BBC pundit Alex Scott, and arguably the reason why it has proven divisive is due to the fact that at the moment it is little more than an idea. Whilst we have become accustomed to even the biggest stars in Football sitting on the bench regularly throughout the season, there are precious few examples of clubs willing to regularly rotate their goalkeepers, within a league season at least (many clubs will select their understudy for cup and/or European fixtures). Therefore there isn’t a direct way to test the theory, because you need at least one team (and ideally several) to actually adopt the strategy to gather the type of data which would illustrate whether the policy was effective or not.
The reason why goalkeepers have largely avoided rotation would appear to come down to two main factors. Firstly, despite the evolution of the role over recent years, goalkeepers still run far less each game than any of their outfield counterparts, and therefore are less likely to suffer from physical fatigue than outfield players. Communication between goalkeepers and defenders is also seen as a crucial component in terms of creating defensive solidity in a team, and continuity of selection is seen as a way of fostering effective communication between a goalkeeper and their defence, which is why defenders are also typically rotated less often than midfielders or attackers.
The extent to which goalkeepers have avoided rotation can be seen through the appearance statistics in the Premier League last season, with nine clubs (Cardiff City, Everton, Leicester City, Liverpool, Manchester City, Manchester United, Newcastle United, Watford and West Ham United) only using one goalkeeper in the league all season. A further four clubs (Brighton and Hove Albion, Chelsea, Tottenham Hotspur and Wolverhampton Wanderers) all had a goalkeeper who started and finished the season as their club’s first choice, playing at least 33 games and generally only missing games through injury or suspension. That leaves just seven clubs who changed their first choice goalkeeper during the season, and therefore are the most relevant to study in regards to the potential effectiveness of rotating goalkeepers. In each case, the Plus Minus rating per 90 minutes figure is listed to provide comparison between the various custodians used.
Arsenal started their league campaign with Petr Cech (+0.42) in goal, but following an injury replaced him with Bernd Leno (+0.55) who played the final 31 league games consecutively. The only other club to use just two goalkeepers was Burnley, who after 19 games replaced Joe Hart (-1.16) with Tom Heaton (+0.05). The remaining five clubs each used three goalkeepers throughout the season, though in each case one of those goalkeepers played a maximum of two games all season.
Bournemouth started the season with Asmir Begovic (-0.57) as their first choice, though Artur Boruc (+0.15) had largely usurped him by the end of the season. Wayne Hennessey (-0.53) started the season in goal for Crystal Palace but was replaced by Vicente Guaita (+0.42). Fulham started the season with Fabri (-1.84), replaced him with Marcus Bettinelli (-1.31) before replacing him with Sergio Rico (-1.05). Huddersfield Town started with Ben Hamer (-2.74), replacing him with Jonas Lossl (-1.06), and finally Southampton started with Alex McCarthy (-0.59) before replacing him with Angus Gunn (-0.08).
Whilst seven clubs is a relatively small sample group, what is interesting is that in all seven instances, the goalkeeper who finished the season as first choice had a better Plus Minus rating per 90 minutes than the goalkeeper who started the season as first choice. The improvements are not small improvements either, improving by an average of +0.86 per 90 minutes. It should be noted that in some instances these statistics are slightly misleading, with Ben Hamer for example actually playing four games covering for an injured Jonas Lossl after losing the starting berth, having only been first choice until his injury during the third league game of the season, and therefore actually played more games as an understudy than he did as first choice.
Nevertheless, the fact that all seven clubs improved after changing their goalkeeper hints towards potential benefits of rotating goalkeepers. Had this been the case for just some of the clubs, a case could have been made that the manager had simply identified the wrong goalkeeper to start the season as first choice and had a better candidate waiting in the wings. Yet for that statement to have been true seven times out of seven is improbable, albeit certainly not impossible.
Whilst goalkeeping is less physically demanding than any outfield position, it is arguably the most mentally demanding position. The role demands a high degree of concentration, with infrequent but important contributions to the game. Any mistakes made are more likely to result in a goal for the opposition, a goal which could potentially cost the team points, and the confidence of a goalkeeper can also be dented by regularly conceding goals even if they are not at fault.
Another trend which can be picked out is that of the seven clubs who changed their first choice goalkeeper during the season, only Arsenal replaced a goalkeeper who had a positive Plus Minus rating per 90 minutes. Yet of the remaining six, Bournemouth, Burnley and Crystal Palace all finished the season with a first choice goalkeeper with a positive rating per 90 minutes, whilst Angus Gunn at Southampton was also close to a positive rating at -0.08.
Changing a goalkeeper when a team is struggling defensively is certainly not guaranteed to reverse the slide. Fulham and Huddersfield are both evidence of that, even if improvements were evident. However, given the trends of last season, a case could certainly be made for managers to change their goalkeeper more quickly when a defence is struggling, even if we still appear to be a long way off determining the effectiveness of a regular rotation policy across a full league season.