Manchester City in the late 2010s and early 2020s. Barcelona and Spain straddling 2010. Sacchi's Milan. The Dutch of the 1970s. Hungary in the '50s. Scotland, in the late 19th century*. Every generation has its iconic teams characterised by their possession of the ball.
*(From 'Inverting the Pyramid': "That success [holding a heavily-favoured England to a draw in the first international football match in 1872] may have confirmed the notion of passing as superior to dribbling - north of the border at least - but it could never have worked had passing not been part of the game in Scotland almost from the start.")
As much as trends come in and out of fashion, I'd wager that the teams that win throughout football's history have, almost without fail, had more possession of the ball than their opponents. I think there's a solid footballing reason for this, and I think it has profound implications for statistical analysis.
I will try, where possible, to throw in a zinger to lighten the tone.
Football is an invasion sport - categorisationally, rather than method of global adoption - meaning that teams try to score points via entering the opposition's 'territory'. But it's also one with a ball.
We have a lot of these. The same is true of basketball, netball, all of the other offshoots of football (Australian, American, rugbies league and union), but the balance between 'space control' and 'ball control' is different in each one. Take basketball. Both conceptually and literally it's a much more 'ball-handly' sport than football, but space control is still important on offence and defence. Maybe the split is 70/30 between the importance of ball control vs space control, whereas football is closer to a 50/50 split.
If you break the sport down like this, then 'what makes a football team 'good'?' is easily answered (though much harder implemented): being better at controlling the ball and better at controlling the space.
Hey, is there a well-known approach to playing football that focuses on these two things?
Like any good little analytics newsletter, we now cross over to Moneyball to quote (Brad Pitt's) Billy Beane: "If we try to play like the Yankees in here [the pre-draft meetings], we will lose to the Yankees out there [the field]." Instead of back office approaches though, think tactics. If the other team is better at controlling the ball and space than you, either in or out of possession, then why try and control them in the same way? You won't! You will lose to the Yankees out there.
So what do you do? You make the game less about controlling the ball, less about controlling the space. Shake things up. This isn't ingrained Englishness, this is elemental philosophy, this is practically Aristotelian: get bloody stuck into them.
From here, from this basic framework, you can break down football archetypes into the 2x2 quadrants that high-or-low space and ball control form. The quadrants are less about absolute quality, and less about how much a coach values each thing, more about how the talent of their squad matches up against the opposition.
What we now call 'positional play' is a high-space control, high-ball control approach. Quick ball-carrying is high-ball control, but only has a low dependency on space control. Long-ball football has a low reliance on both ball control and space control. I'm not really sure if there's a style of play that's low-ball control but high-space control, but it's probably what clumsy La Masia graduates end up playing.
This can work on the defensive side too, although we might need to clarify our meanings a bit. By 'ball control', that could be direct 'ball actions' like tackling (or, in attack, dribbling and passing) but also close-quarters 1-v-1 stuff. Meanwhile 'space control' can be affected by fitness as well as tactical understanding.
So, a hunkering deep block is a low-ball control, low-space control approach (reducing both the frequency of duel situations and the amount of space aimed to be controlled). Bielsa-ball player-marking would be high-ball control (individual duels), low(er)-space control. High pressing is in the high-space control, high-ball control quadrant, while a high block that is quick to drop when passes are made through or over it could be an approach taken if you have a squad high on space control but low on defensive ball control.
Clearly you would prefer to control both (unless for aesthetic reasons), and therefore always be in the high-space control, high-ball control quadrant, but that's not always the hand you're dealt.
So, now for the statistics.
If you're lucky enough to remember the 'stats in established media' conversations of, I dunno, 2010-2014 you'll recall the tiresome lines about 'winning the possession', the idea (or, sadly, reality) that some people thought that more passes meant a team was playing better. It's a classic 'correlation doesn't necessarily equal causation' issue. Better players make better teams who have more possession, but maximising the possession figures doesn't mean that the team is good.
In fact, if a team is only moderately good at controlling the ball and controlling space in possession, but they are intent on passing the ball around, then the defending team should often just let them. Their high possession figures aren't coming because they have high ball and space control; they're coming about because their ball and space control is bad enough to render their possession toothless.
'Player talent' isn't unchanging though. You can coach improvement into players in both space and ball control, and perhaps managers who go for 'possession ball' with squads of average-or-worse talent, and who don't fail badly, should be given even more credit than they currently get.
Assessing managers is a slightly higher step up the football food chain than I'm interested in for this newsletter though. I'm more interested in the humble plankton of the industry, hoovered up by the giant bloated whales, the everyday club analyst.
If you, or your boss, has an idea of how you want the team to play, how do you design key performance indicators when you know that they may simply be influenced by the disparity in team strength? Are you measuring execution of a game plan or the inevitabilities of the sport? Do you adjust the metrics if you're playing a team much better or worse than you, or simply adjust expectations for what the post-game numbers will look like?
Reading that back, of course you should change your KPIs, right? A manager isn't going to go out and tell the team to play in the same way against Bayern Munich as they would against Burnley, so, naturally - beyond the essentials like xG and such - you'd change what is most important to measure?
I am eager to be corrected on this, but I think the smart approach would be to develop different stats to look at for different circumstances. Playing the league leaders? Load up spreadsheet template A. Conceded early against a team of a similar strength to you? Spreadsheet template G.
Maybe this whole space control and ball control framework would help elsewhere too, like in finding appropriate metrics to use when scouting players. Maybe it could help people learn to analyse the game better, quicker. (If it doesn't, it's probably not a good framework).
The thing that makes this idea worth approximately 1300 words of your time is not that it's new or industry best-practice. Football is like one big jigsaw puzzle that you haven't been given the picture for. You don't even know what shape it takes, and if you don't know what shape it takes then how can you start piecing things together from the edges?
I think a good framework for understanding football does that: it gives you an idea of the shape of the sport, of its boundaries and dimensions, giving you a starting point to fill in the rest from there.