Finding an alternative to 'Average Squad Age' to show club age profiles.

written by: Jamie Kilday @FitbaInScotland

If you are an avid reader of Modern Fitba you will know that we have written a few articles recently surrounding youth development. If not then here are a few that might be of interest to you, Matt Rhein's - Does Scotland Need a Reserve League and Jason's two parter on the Development League, Part 1 & Part 2.

A common trend amongst supporters who seek to defend their clubs player development policies is to bring up ‘average squad age’. A quick twitter search of ‘average squad age’ and you can read tweet after tweet of people using this statistic, but why? What is its relevance? If any.

I am not a fan of ‘average squad age’ and I think Owen James Brown summed up exactly why I am not a fan. 

It’s nonsense isn’t it... so how can we look at it in a better way?

Firstly let’s take the squad element of it. Who is in the squad? Players picked for first-team squads? Players who got minutes? Players who are listed on the clubs website? There are more ways than one to skin a cat but to be honest, it’s messy and cumbersome so let’s not bother. Let’s remove the squad aspect entirely and replace it with minutes played at each age point.

Let’s look at the ‘average’ part of it. What is a good average? As Michael Caley observed in an article he wrote about player ages for ESPN he noted that different players in different positions peak at different ages. So a team theoretically could still have a ‘good’ average but ‘bad’ ages in each position...I agree this is getting messy so let’s bin the ‘average’ aspect off too.

So what’s left...age. Well maybe there is a better way to show age distributions for teams and one way could be using Michael Caley's principle of ‘peak ages’ and seeing how many minutes are given to players based on their age compared to the peak.

For those that didn't click the link to his article here is a graphic denoting each Peak Age curve.

It’s worth noting that Caley’s model uses data from the Big 4 leagues, and as Matt Rhein observed last season, in some respects players can enjoy an Indian summer in Scotland, being ‘past peak’ doesn’t necessarily mean useless. Caley also refers to his theory as an age curve so whilst a 'peak age' player is advantageous, having players a few years either side of the peak isn't a bad thing either.

With that now in mind here is a breakdown of how each club last season distributed their playing time based on player ages relative to the peak age for their position...not as snappy as ‘average squad age’ but maybe when dealing with multiple variables across different areas you can’t have a nice single figure that encompasses everything.

Age profiles of clubs are important for many reasons and hopefully this is a better way of showing it. So without further ado here is how clubs in the SPFL looked last season. For comparison, I'll include Transfermarkts Average Age for each club.

If you just want to view a specific club then click on your sides badge.

Aberdeen - Average Squad Age - 26.5

Aberdeen.png
I do believe that they’ve had a fantastic cycle in the last three or four years and they need to renew it. That’s what I understand looking at it from outside
— Pedro Caixinha, May 17

Whilst the graph is heavily shifted to the right Aberdeen have already made moves to counter this and none of the players who were 4+ years past peak remain at the club, although the signing of Gleeson will certainly add to the higher end of the scale. There were also a significant amount of minutes going to players who will be 4 years past peak this season, with Logan and Considine making the bulk of those minutes.

Aberdeens few signings this summer have mostly been young which is encouraging so it will be intersting to see if this season sees more minutes going to younger players. So far though no player at the club will be 'Peak Age' this coming season.

Celtic- Average Squad Age - 25.5

Celtic are the first of three clubs who gave more minutes to 'Peak Age' players than any other age group. With the bulk of this seasons squad going into next the only notable departure is Stuart Armstrong who was one of the 'Peak Age' players. A fair balance of youth and experience this probably isn't a bad graph to have.

Presuming that next season will see Edouard taking mins off Griffiths, Hendry taking mins off Bitton/Lustig and Bain taking minutes off De Vries I expect that although fewer players will be of peak age, more minutes will invariably go to younger players.

Dundee - Average Squad Age - 24.8
It is important the kids here- and even youngsters from elsewhere that we can maybe look to attract - know this is a club and a place where you can come and develop. So I’m not firghtend to throw kids in.
— Neil McCann - Jan 18

Dundee's focus on youth has not seen success on the field but should probably still be revered rather than dismissed. Their current crop of transfers have interesting age profiles too with all being within a year either side of peak age. The only issue with signing players around peak age for their position is that if it doesn't work out immediately then it is unlikely the players will develop much further.

Hamilton - Average Squad Age - 26.2

The first club (alphabetically) of two not give any minutes to players of 'Peak Age', although a significant number of minutes were played by players under peak rather than over which is encouraging. Over 2000 more minutes in fact went to players under peak compared to over.

Going into this season Leonard Sowah is the only peak age signing they have made so far with Ziggy Gordon (Who it feels like has been around forever) returning from Poland still a year from peak. Imrie and MacKinnon represent the veterans in the squad as Tomas' and Donati's minutes will invariably go to younger players.

Hearts - Average Squad Age - 26.3

The likes of Harry, Anthony, Henderson and Lewis Moore will stay involved. The others have performed really well but I don’t think too many will be pushing for first-team places next season
— Craig Levein - May 18

Hearts are often lauded for their commitment to youth so it may surprise some that their distribution. 2000 more minutes went to players above peak age than under. Those minutes went to 12 players as appose to the 39 players who featured who were below peak age.

This suggests that although hearts give time to young players there are not many young players given extended runs in the first team. Only 8 of the 39 players who were below peak age for their position played over 900 mins, of which 3 have already left the club (Milinković, Gonçalves and Walker). With Hearts already being busy in the transfer window signings for the first team are more on the older side which shows that Levein is still looking to add experience to his squad. It could well be that a commitment to youth is actually just indicative of a lack of squad depth. Either way hearts young players look promising so there is a lot to forward to if the players do get more first team minutes.

Hibs - Average Squad Age - 26.6

There is a lot to like about Hibernians age distribution. Lots of minutes going to players who are around the peak age with this mostly focusing on players within a couple of years of the peak age suggesting development of a squad approaching its best years. Nice. Only one issue, which is becoming a theme in the capital, 5 players make up the minutes between -2 and -1 and three of them have left already, one is now understudy to Adam Bogdan the 5th is John McGinn.

Kilmarnock - Average Squad Age - 26.7

A look at Kilmarnock and Hibernian's Average Squad Age and there is 0.1 in it. A look at their graphs and you can see a completely different story. Kilmarnock gave a large chunk of their minutes to players both 6 years and below and 6 years over the peak age. They are also the second of only two clubs profiled that didn't give any minutes to players of 'peak age'.

Livingston - Average Squad Age - 25.9

Neil Alexander stands out like a sore thumb here on what isn't a bad graph from the Premiership newcomers. I expect that next season could end up looking similar if player manager Kenny Miller gives himself enough minutes. They are also the third and final side to have peak age players get more minutes compared to other ages.

Livingston have a small squad and so far their new recruits have all been on the younger side of the age curve. I imagine more players will come in before the window closes so it will be interesting to see if Miller thinks that they need more experienced heads or looks to bulk out his already young side with even more young players

Motherwell - Average Squad Age - 24.5

Another side who give a lot of minutes to young player, Motherwell's age distribution looks good as 7 first team players come into 'peak age' over the next two seasons. All of the Motherwell players who got minutes last season were at 2 years past peak at most, with the exception of Steve Hammell who has already left the club.

Their new signings are also of a similar age profile to the rest of the squad which indicates that there is a policy in force at Motherwell around age. The only player Motherwell have signed who doesn't fit the profile is Conor Sammon who next year will be four years passed the peak age for a striker. Its fair to say that the Irishman certainly has a point to prove which I'm sure will be highlighted even more being the oldest player in the squad

Partick Thistle - Average Squad Age - 25.9

So the challenge we have set Alan is to rebuild the team with the aim of one season in the Championship then back to the Premiership
— David Beattie - May 18

As age distribution graphs go, this isn't a bad one from Partick Thistle who had a good mix of players, not too many older players and not many inexperienced players either but age is not an indication of talent. Partick have lost 20 players which made up last seasons squad so Archibald has quite the rebuild on his hands so I would be suprised if Thistles graph looks anything like this come the end of their first season back in the championship.

Rangers - Average Squad Age - 25.8

Alongside league winners Celtic and Premiership Play off Champions Livingston, the third side to give more minutes to 'peak age' players is Rangers. Much like the other two sides their remaining minutes where mostly played by players below peak age. Bruno Alves, Kenny Miller and Nico Kranjcar all leaving fits in with a trend that most clubs have had this summer of getting rid of the older outliers of their squad.

The clubs recent transfers are all on the younger end of the scale and seem to be bought primerily to slot into the first team. The only notable exceptions to this are Allan McGregor and Scott Arfield who are arguably of a talent that them being past peak age won't be too much of an issue.

Ross County - Average Squad Age - 26.2

A change of manager at a crucial point in the season saw a different focus when it came to the type of players brought in to the club. Ultimately the players brought in by Owen Coyle didn't have much of an impact on the field but their ages were very much at the extremities of the distribution graph.

A fairly even balance either side of peak age, the inexperienced side is heavily focused on the 3 years under peak age due to the game time of Jamie Lindsay, Jason Naismith and Marcus Fraser.

Ross County are another side on the re-build after dropping to the second tier of Scottish football and so far have lost players from different age profiles across the whole squad so it isn't inconceivable for Ross County's graph to look similar come the end of the coming season.

St Johnstone - Average Squad Age - 27.7

Unsurprisingly St Johnstone have the most significant shift towards the older side of 'peak age' with over 6000 mins going to players 2+ years over peak. 

With Miller, Mannus and MacLean all leaving and St Johnstone's new signings having a much younger age profile it could be seen that St Johnstone are making the right moves but they are not far off getting to the point where they will need a large overhaul of their creaking squad

St Mirren - Average Squad Age - 25.1

Personally I think a healthy age distibution looks like a hill rather than a valley but as Kilmarnock's age distribution was similar last season maybe there can be some benefits to having lots of young players mixed in with players who are more than a couple of years passed peak.

St Mirrens older contingent do seem to have left this summer and so far the club have brought in players who are the most part on the younger side of peak age. So I expect that come the end of the new season this graph will have shifted to the left.


Notes

  • Ages are based on player age on the 24th of December 2017 which is the literal midpoint of the league season. Theoretically whatever age you are at the midpoint is the age you have been/will be for most of the season. In an ideal world you’d breakdown mins played for each player at a certain age but if I’m honest I’m not sure what date point Caley used for his sample.
  • Age is obvious not an indication of quality. There are plenty of players who are way over their peak age who can do a better job than some players who are at peak age. Age distribution is more about profiling a squad rather than assessing its qualities. 
  • I don't know why but expected more clubs to have similar age profiles. It seems like only Motherwell and Dundee have a 'type'.
  • Two trends observed by clubs so far this season are clubs are cutting their older players but also seem to be losing players just as they reach peak-age. The later would be savvy if clubs were getting fees for their players but for the most part players in Scotland usually leave for free.
  • This article was written under the influence of Sleaford Mods, Slaves and The Music