FIFA's big, big plans (no, not the World Cups)

FIFA aren't just a tournament organiser anymore (although they'd like to do it more often). They're moving more into the tech and data world...

A lot of people scoff at The Phantom Menace for having a plot revolving around trade policy.[1] But I am pretty sure that you, dear reader, will have strong opinions about the organisation of the international football calendar. If you can find scheduling something to get excited about, you can find galactic tax policy exciting too.

Anyway, this isn't a defence of the prequel trilogy. This is about why FIFA is more of an interesting place than you might think.

Let it first be said that if you want people to trust you, you should avoid massive corruption scandals. But the reason why FIFA keeps pushing their extra World Cups (of the club and national team varieties) is that they don't earn as much as you might think. Their annual report for 2020 predicted that they would reach a figure of $6.4bn revenue for the 2019-2022 cycle. UEFA brought in half that during the 2019/20 season alone.

Most of FIFA's money (83%, according to Sportico) comes from the men's World Cup. This money goes into, yes, the luxurious aspects of some of the administration, but also funding global football development.

It re-jigged the way it doled out this funding in 2016, making a uniform amount of money available to each member association for operational costs and further projects (with a slightly larger amount going to nations that don't have particularly large revenues).

But half of the annually allotted $1m comes with conditions.[2] They include a requirement for organising a women's competition that lasts for at least six months, with at least ten teams, consisting of at least 90 matches. A women's national team has to have played at least four matches that year. Boys' and girls' competitions in at least two age groups need to be organised. And a refereeing programme (that includes women referees) needs to exist.

I say all this not to suddenly pivot to being a pro-Gianni Infantino newsletter. Only today, the day I'm running a final edit on this piece, Infantino has put his foot in his mouth yet again. But I found that previous information out while researching this issue and thought it was interesting. And it's a nice way to remind people of the intended goals of FIFA before going into all of the other things they do. Things that are more immediately relevant to this newsletter...

Subscribe to demystify football in your inbox.

A number of years ago, FIFA set about making sure that the wearable tracking devices that players wear during games were safe to use. This was mainly just a rubber-stamping exercise, but in 2018 they widened their remit, checking the accuracy of these systems and of optical tracking data. Then last year they announced they'd be doing the same for tracking data generated from broadcast footage.

Before the middle of the decade, they'll have a whole range of data quality marks for certified systems. And it's not just for tracking systems; it's goal-line tech, it's offside lines, it's playing surfaces.[3] One of the latest projects is a lower-cost implementation of VAR. Makers of various technologies are able to put on their websites that they have a FIFA Quality mark.

It's not just badges of approval though. The body are working on standard formats for transferring tracking data and standards on collection and protection of player performance data (particularly noteworthy in the context of Project Red Card).

And you probably wouldn't be surprised, by now, if I were to write that it's not just that either, that there is one more thing that FIFA are doing that's worth writing home about. And that it, the latest area they're moving into, might be the most intriguing: data collection.

As last year's Arab Cup started (a FIFA-organised tournament, as of this edition; essentially prep for the 2022 World Cup), the body made a big announcement. Not only would they be collecting event data during the tournament, and supplying it to the national teams taking part, but this dataset would be huge.

“We have a team of 25 analysts working on each game," FIFA group project leader Chris Loxston said in the organisation's article [linked above] on the subject, "so we have one analyst covering one player for the entire period of the game."

For context, the standard set-up elsewhere in the industry is usually one collector per team.

"Typically, a football data set only looks at all the actions around the ball," Loxston continued, "what we are actually able to do here is look at all the actions off the ball and around the ball as well. An average football data set is somewhere around 2,000-2,500 events per game; we collect in excess of 15,000 data points."

This is huge. And very interesting.

Training data collectors is no easy task; companies like Opta and StatsBomb spend months training theirs up. But the advantage they get from this time and expense is a set of collectors who will work for them on match after match for (they hope) years. FIFA are spinning this operation up for individual tournaments at a time. (Although the fact that they have one collector per player may make each person's job much easier, and therefore the training and collection processes less intense).

Assigning an analyst per player also means that FIFA could potentially be giving (some form of) work experience to an awful lot of people at one time. Useful. The global governing body having their own 'public sector' dataset is also a neat way of sharing data amongst all national teams and associations (not just those that can afford 'private sector' event data). Also useful. And might it be a potential media opportunity in future (media of the world, wouldn't you want FIFA's own data for their big tournaments)?

Now, to collect data, you need to have an idea of what you're collecting. Which is why FIFA have developed their own 'football language', which "will be our blueprint for how we [FIFA] analyse football in the future".[4]

The language encompasses a lot more than the on-ball events that most event-datasets do. There's sections on types of movement to receive the ball and offering to receive, even if the player doesn't end up getting passed to. (And I think the site is a gold-standard that all data providers should aim towards in terms of communicating definitions to users, complete with videos).

FIFA's football language even makes a distinction between standing and sliding tackles. Finally.

The reason why I find this particularly interesting isn't just that Arsène Wenger has, single-handedly, addressed an issue I've had with football defensive data for years[5], but for what it could mean on a wider scale.

In England, when (almost) anyone talks about stats they talk through Opta's terminology. Tackles (rather than duels), Big Chances, key passes. These terms may be quite similar to how people talked about football pre-data, but to some people, if you talk about defensive duels or shot assists, they won't really know what you're talking about. Especially if you're talking about data.

From a brief look through FIFA's football language, I don't think it clashes with -- as in, seeks to re-define -- aspects of any existing event data provider's provision that I can think of. But given the broad scope, will future data providers feel pressure to align their definitions and terminology with FIFA's? Will we talk about 'units' of opposition being broken or bypassed, rather than lines? Will the ten types of movement to receive the ball that the global body have identified become standard definitions?

Asking those questions, and coming to the end of this edition of the newsletter, I realise that I may not have done FIFA many favours by opening this issue by mentioning The Phantom Menace. The start of a story where an elderly gentleman asserts his influence in an imperfect but generally well-meaning democratic body, before using it as a platform to become ruler of the whole galaxy? Arsène Wenger doesn't seem like that kinda guy... does he?

"Data collection leads to statistical modelling. Statistical modelling leads to expected goals. And expected goals... lead to the dark side." – Yoda, Attack of the Cloned Github Repositories, deleted scene

[1] "A lot of people scoff at The Phantom Menace for having a plot revolving around trade policy." || In fairness, the opening paragraph of the introductory text crawl, the first cinematic thing that Star Wars fans would have seen for decades, was: "Turmoil has engulfed the Galactic Republic. The taxation of trade routes to outlying star systems is in dispute."

[2] Fifa Forward programme conditions can be found here

[3] "And it's not just tracking data..." || If you want to see the full range of things FIFA do quality checks on, check this page out. It all started with footballs in 1996.

[4] "our blueprint for how we [FIFA] analyse football in the future" || Quote from Arsène Wenger himself, from here

[5] Me crediting Wenger with the standing vs slide tackle distinction is a joke; it being top of my Christmas list to Santa every year is not.