A smarter approach to squad-building

Scenario: You arrive at a club as a director of football. The squad needs an overhaul. How do you plan?

As a starting off point, it’s worth checking out this visualisation from the Financial Times’ John Burn-Murdoch, showing transfers from three big clubs since 2015/16 and the minutes they’ve played. Manchester City pull up the average a long way, but over five years they, along with Liverpool and Manchester United, average just above four transfers a season.

That means that, even if you make 11 hits in a row, it’d take three years to change a whole starting XI. You may not want to, but in terms of the squad as a whole that might still be the amount of changes you think need making. Forward-planning is needed.

So let’s take a team that’s in need of an overhaul: Crystal Palace. The Eagles have the oldest squad in the Premier League, and you have to get down to their 18th most-played player this season, Jairo Riedewald, to find a player younger than 27! Here’s a vis from now-at-The-Athletic Tom Worville from earlier in the season that highlights this:

Now, football teams are not necessarily the best forward planners. Often, transfer strategy seems to be about plugging gaps as and when they appear, with only some notions about buying for the future involved.

I had a thought recently though. Most transfers are game model-led: a team plays like X, they need a player to fill Y role within that, they have 5 options for that and need to buy one of them.

What if they flipped that; what if the game model was led by what players would be available to buy in and build the squad with?

There are certain skills that are in abundance and certain skills that are not. This is especially important for clubs outside of the top level, who can’t just go and buy the one player who does exactly what they want. They might be lucky to get their fifth-choice target.

Some of the skills are more valuable than others too, and you could probably plot them on a nice and basic scatter plot. (below: an artist’s interpretation)

A rough graph of football skills with 'abundance' and 'value of skill' on the axes; athleticism at the high end of both, for example

As we’ve established, a squad rebuild will take around three years, so you need to work out what skills will be in abundance and of value across the next 3-5 years (even if the rebuild only takes two or three, you’re going to want to keep and sustain that game model for a few more years afterwards). No point wanting to build a possession system requiring creative 8s if there are no creative 8s who will be in their peak years when you need them.

You might also want to look at what skills are unlikely to be in abundance in the future, and fit the reverse of that into your model. If there will be very few 1v1 defenders around in the near future, maybe you want to try and build a team of great 1v1 dribblers.

I asked Scouted Football, who know far more about young players than me, whether there are any trends in the players who will be reaching their peak age in ~3 years time.

The main changes? “How important it is to be a well-rounded player that can do everything straight away”.

“Perhaps someone like Erling Haaland would have been viewed as more of a target-man maybe 5-10 years ago,” SF explained, “but now we see him as a player that can drop deep and burst pass defences, be an asset as a creator, and qualities like pressing ability for attackers are now so critical for how many of the top managers want to play. Likewise, we've seen Trent Alexander-Arnold become arguably Liverpool's most important creative threat from right-back.

“Even further down [the football pyramid] it will become harder and harder to carry an attacker that does not work hard defensively and press intelligently, for example, or a defensive midfielder that can't play a progressive pass, or a full-back that doesn't function as an auxiliary winger. In some ways, we're almost going full-circle back to a Total Football ideology.”

Scouted Football also suggested that things were moving this way because of the increased flexibility it gives teams, a winger dropping and a wing-back becoming a threat, for example.

How might this help us at Crystal Palace? Well, we can assume that pressing forwards will be pretty ubiquitous, although the increased creativity might interest me more. We’ve seen front threes who interchange fairly well before, but what about strikers who can drop and circulate the ball around the edge of the box while a winger, or attacking midfielder, takes their place.

A mock-up tactical board of what this could look like

Tactics board with an inverted full-back and false nine

But if pressing forwards are going to be ubiquitous, then that means our own defenders will be pressed. And even if players are all becoming more like all-rounders, it feels like there’ll be more forwards who can press than defenders who can break the press. I want the attacking players to have some height and/or aerial ability too to let my back-line go long if needed.

This team we’re building towards will also defend in a fairly standard mid-block, using precision pressing at the front to disrupt opponents rather than risk playing high all of the time and increase the requirements on our defence and goalkeeper.

So, based on the skills that may be available in the medium-term future:

  • Forwards need to be able to rotate
  • Central midfielders don’t need to be specialists as they, or the defence, can go long when needed, but once in the final third they need to have decent tough and an understanding of tempo
  • Defenders can be the type who stay compact and solid rather than great open-field defenders, need to be competent enough on the ball to learn some press-breaking patterns, but won’t be relied on for tons of ball progression.

There is a slight problem with this, in that there’s virtually nobody in the current Palace squad who fits this and is below 30 and isn’t probably-gonna-leave-Wilfried Zaha, but that’s where my second thought on squad-building comes in. What do you do while you’re rebuilding anyway?

In the middle of a rebuild, you’ll have three groups of players.

  1. The ones you had to start with who aren’t in any future plans
  2. The ones you do actually want for the future
  3. Others

The Others are players you don’t necessarily want in two years’ time, but you might need them around for that duration, depending on how the building around them is going.

This is how, unscientifically and based mainly on age and minutes played this season, I think Palace’s squad breaks down right now:

A breakdown of Palace's squad into those three groups

Defenders and goalkeepers are strongly represented in group 1, so that’s where Palace should be focussing on replacing with true rebuild players. The Others can tide over in other positions in the meantime.

But as the squad turnover starts and we enter the awkward middle period, we’ll need to buy players for the Other category too. How you fill these is up to you, but I have a specific idea.

Public analysts have tended to be negative about buying older players. I think this stems from two things: work done on looking at when a player’s peak age range is; and clubs buying players on the older end of this, expecting them to remain in their peak form for a few more years. Reader, they do not stay in their peak form for a few more years.

However — as I think Ted Knutson said on a recent StatsBomb podcast — the statalysts aren’t pro-youth as a dogma, it’s all about where the value is. It seems very possible that, in the midst of a rebuild, older players who might still have 12-18 months of decent form in them would be very useful to plug gaps. They’d also probably be cheaper than the 23-25 year olds who could produce the same level of performances but would have their future growth factored into the price.

Gary Cahill, actually, was a perfect example of this. He was brought in last summer at 33, has filled a role very well, and will probably be gone in a year or two, at the latest. Palace are going to need more Cahill buys soon.

To take the Cahill model to a T, other 30-33 year-old players at top London clubs (ordered by minutes played this season): César Azpilicueta, Pierre-Emerick Aubameyang, Toby Alderweireld, David Luiz, Willian, Moussa Sissoko, Sokratis Papastathopoulos, Jan Vertonghen, Hugo Lloris, Mesut Özil, Danny Rose, Olivier Giroud, Henrikh Mkhitaryan.

Palace won’t be able to target all of these, but Willian and Sissoko could perhaps plug holes in a year or two if they want to stick around in London.

I won’t map out a full rebuild for Palace (maybe I’ll come back to it in future), but the ideas I’ve laid out in this post are ones I’m interested in. Who knows how the club will actually manage this, but they really need to start soon.

Shout-out corner

Test cricket is back and so is cricket writing. The nature of cricket lends itself very well to stats and stats-adjacent chat, and there’s already been some great stuff.

Jarrod Kimber is one of the best writers around, full stop. He has a substack here, but this piece on all-rounders is right up my street — both something stats-y, but also about the definition of a concept everyone uses but rarely thinks about how they apply.

Jarrod is Australian, so I’m going to use that as a tangential leap to something non-sports: the TV soap Neighbours. CW for transphobia and a trans character’s childhood experiences.

The show has had a trans character, Mackenzie, for a year now, and in 2020 she became a series regular rather than a recurring friend-of-Ramsay Street. The actress, Georgie Stone, has worked with producers on storylines involving her character so it’s no surprise that (to this cis outsider’s eyes at least) the show has dealt well with subject matter that could have been handled very badly, and a chapter of which came to a close recently.

In the show, Mackenzie’s father (Grant) abandoned her as a child after learning she was trans. When Grant recently returned, he was still struggling to come to terms with the fact he had a daughter. Characters, mainly old family friends Shane and Dipi (who have recently taken Mackenzie in as if she was their own daughter), chastised, educated, encouraged, and challenged Grant at varying moments.

I get that, even though this was a journey with a lot of pain for the character of Mackenzie, it’s been one with a resolution that may be unfamiliar to many. It culminated a week or so ago in Grant taking part in a reading (unplanned on Grant’s part) of Mackenzie’s diaries during a writer’s festival, the character stepping forward to read words from his daughter about how he had abandoned her. It was the decision point in a build-up of interactions that had boiled down to people telling him “look, this person should have your love and support: you need to learn to give them that, or leave.”

There is a strain of transphobia in Britain at the moment (even — especially, really — among ‘liberal’ circles who should recognise the similarities in their refrains to the refrains of homophobia some decades ago). Trans people are hugely more likely to attempt to take their own lives and lack the support and access to services that they need than the rest of the population.

This has been a long way of leading up to saying, explicitly, that people who are trans or non-binary should have our community and society’s love and support: we need to give them that.

Donate to Gendered Intelligence, a charity helping trans people in the UK