Could an independent Yorkshire Win the World Cup?

Oooooh. Football. Image: Getty.

With less than a week until the start of the 2018 World Cup in Russia, it’s worth remembering, that another World Cup – the 2018 ConIFA World Cup for stateless people, minorities, and regions unaffiliated with FIFA - is also taking place in London.

Though happening in the UK, neither of the local ConIFA members will be competing. The Ellan Vannin team from the Isle of Man withdrew midway; and the latest ConIFA member, Yorkshire, only gained membership earlier this year.

One of Yorkshire’s most obvious characteristics, is that it’s absolutely huge compared to most other UK counties. It also – probably – has the highest contemporary population of any of the historic British counties. Indeed, as recently as this February the region resisted attempts to split control of the region up, demanding a “One Yorkshire” devolution deal instead of the proposed control to regions surrounding four of it’s major cities – and in May, a vocal proponent of such a “One Yorkshire” devolution, Dan Jarvis, the Labour MP for Barnsley, was elected as mayor of one of the Sheffield City region.

Given its size, ConIFA membership, and pushes for further devolution, I was wondering how Yorkshire would do as an independent full FIFA member. If it seceded as a whole from the rest of the UK could it field a team that could challenge internationally? Could any of the historic British counties?

Overall, there are 88 historic counties in Great Britain, plus the 6 counties of Northern Ireland (I couldn’t find shapefiles for the older subdivisions) which could be potential independent FIFA members.

Once I had these, I needed some way to rate potential players, and therefore teams. Luckily, the popular video game FIFA18 maintains up to date ratings of thousands of players across 36 different stats (e.g. dribbling, heading, pace etc.). After scraping an online database of players, I’m left with 18,058 players of various nationalities and abilities.

Using a simple regression model, I can use these abilities and the player’s listed preferred positions to predict what each players rating for each position, and use these position ratings to train a computer to pick optimal teams across a variety of formations. If we do this do for every nation that has at least 11 players in the database (10 outfield + 1 goalkeeper), the best 4 national teams that can be fielded are from Brazil, Germany, Spain, and Belgium.

To pick the teams for each county, though, I first had to find the birthplace of player. To simplify things a bit I only check players listed as English, Scottish, Welsh, Northern Irish, or Irish (due to the weirdness of the Irish FA) in my database of FIFA players. For each of these I ran a script to look the player up on wikipedia and scrape their birthplace. Once this was geocoded, I had a map of each British player and their birthplace, and therefore, the county of their birth.

Unsurprisingly, it basically shows a population density map of the UK, with more players born in the urban centres of London, Birmingham, the Lancashire cities and the West Yorkshire urban centres. After binning the players by county of birth, twenty historic counties have enough players to field a team.

On this chart, ‘FIFA_ability’ is the perceived ability of the optimal 11 players in a starting line up for that county, as judged by FIFA stats.

Perhaps a little surprisingly, the Lancashire team is rated slightly higher than the Yorkshire team – though looking at the sheer number of players they can select from it makes sense. Elsewhere, the home counties do well, as do Glasgow and Warwickshire (which contains much of contemporary Birmingham).

Looking at the selections the alogirthm chooses, it’s pretty clear some of these teams tend to be a bit flawed but overall make sense. The Yorkshire/Lancashire teams in particular are full of England international players and are lacking only an experienced top level goalkeeper.

In order to predict how these teams would do at a World Cup, I needed some form of quantifiable rating of a team;s ability. Luckily, ELO ratings in chess can do exactly that: the likelihood of any team A beating a team B is a direct function in the difference in their ELO rating.

Plotting the ELO ratings of each actual national team (an up to date calculation is maintained at ELOrating.net) against the ability of each national team as judged by FIFA18 shows a pretty clear linear trend. Using a regression model of this relationship, we can predict the ability of each hypothetical county national team.

When plotted, these ELO ratings show that some of the counties are definitely in the ball park of established world cup qualifiers – and so we might expected a post-super-devolution Britain to be sending multiple teams to the World Cup.

In fact, Yorkshire and Lancashire are predicted to be about as good as the national teams of Serbia and Sweden. Lagging a bit behind, Essex and Surrey – both of which take in large chunks of what is now London – could expect to be competititve with teams like Turkey and Morocco.

However, just finding out how good these teams would be wasn’t what I wanted to know. I wanted to see if an independent British county could win the World Cup.

To do this, I swapped each of these counties in for the national English team and ran 10000 simulations of the post-devolution 2018 World Cup, uusing the same draws and fixtures as the real tournament uses.

The bad news is, the real-life favourites tend to dominate the simulations. Brazil or Germany were predicted to win the tournament in almost half of all the simulations. On the graph, it;s just possible to make out the red bars of Yorkshire and Lancashire, both of which won 41 out of 10000 simulations (a 0.41 per cent chance of winning any random World Cup).

This seems pretty low – but is comparable to pretty respectable teams like Denmark (0.775 per cent), Senegal (0.217 per cent), and even higher than the Iceland team which knocked england out of Euro2016 (0.339 per cent). It’s way higher than the chances the simulation gives the Russian hosts (0.07 per cent).

Scaling down to just these pretty hopeless nations/counties really shows how little hope the independent British counties would have at an international tournament. However, the best four counties (Lancashire, Yorkshire, Essex, and Surrey) all have about a 0.2 per cent or higher chance, or 500-1 odds, at winning the 2018 World Cup were they to replace England at the last minute. This is an order of magnitude greater than the 5000-1 odds given to Leicester City at the start of 2015-2016 Premier League season, so there’s always a chance.

Want more of this stuff? Follow CityMetric on Twitter or Facebook


 

 
 
 
 

Coming soon: CityMetric will relaunch as City Monitor, a new publication dedicated to the future of cities

Coming soon!

Later this month, CityMetric will be relaunching with an entirely new look and identity, as well as an expanded editorial mission. We’ll become City Monitor, a name that reflects both a ramping up of our ambitions as well as our membership in a network of like-minded publications coming soon from New Statesman Media Group. We can’t wait to share the new website with you, but in the meantime, here’s what CityMetric readers should know about what to expect from this exciting transition.  

Regular CityMetric readers may have already noticed a few changes around here since the spring. CityMetric’s beloved founding editor, Jonn Elledge, has moved on to some new adventures, and a new team has formed to take the site into the future. It’s led by yours truly – I’m Sommer Mathis, the editor-in-chief of City Monitor. Hello!

My background includes having served as the founding editor of CityLab, editor-in-chief of Atlas Obscura, and editor-in-chief of DCist, a local news publication in the District of Columbia. I’ve been reporting on and writing about cities in one way or another for the past 15 years. To me, there is no more important story in the world right now than how cities are changing and adapting to an increasingly challenging global landscape. The majority of the world’s population lives in cities, and if we’re ever going to be able to tackle the most pressing issues currently facing our planet – the climate emergency, rising inequality, the Covid-19 pandemic ­­­– cities are going to have to lead the way.

That’s why City Monitor is going to be a global publication dedicated to the future of cities everywhere – not just in the UK (nor for that matter just in the US, where I live). Our mission will be to help our readers, many of whom are in leadership positions around the globe, navigate how cities are changing and discover what’s next in the world of urban policy. We’ll do that through original reporting, expert opinion and most crucially, a data-driven approach that emphasises evidence and rigorous analysis. We want to arm local decision-makers and those they work in concert with – whether that’s elected officials, bureaucratic leaders, policy advocates, neighbourhood activists, academics and researchers, entrepreneurs, or plain-old engaged citizens – with real insights and potential answers to tough problems. Subjects we’ll cover include transportation, infrastructure, housing, urban design, public safety, the environment, the economy, and much more.

The City Monitor team is made up of some of the most experienced urban policy journalists in the world. Our managing editor is Adam Sneed, also a CityLab alum where he served as a senior associate editor. Before that he was a technology reporter at Politico. Allison Arieff is City Monitor’s senior editor. She was previously editorial director of the urban planning and policy think tank SPUR, as well as a contributing columnist for The New York Times. Staff writer Jake Blumgart most recently covered development, housing, and politics for WHYY, the local public radio station in Philadelphia. And our data reporter is Alexandra Kanik, whose previous roles include data reporting for Louisville Public Media in Kentucky and PublicSource in Pittsburgh, Pennsylvania.

Our team will continue to grow in the coming weeks, and we’ll also be collaborating closely with our editorial colleagues across New Statesman Media Group. In fact, we’re launching a whole network of new publications this fall, covering topics such as the clean energy transition, foreign direct investment, technology, banks and more. Many of these sectors will frequently overlap with our cities coverage, and a key part of our plan is make the most of the expertise that all of these newsrooms combined will bring to bear on our journalism.

City Monitor will go live later this month. In the meantime, please visit citymonitor.ai to sign up for our forthcoming email newsletter.


As for CityMetric, some of its archives have already been moved over to the new website, and the rest will follow not long after. If you’re looking for a favourite piece from CityMetric’s past, for a time you’ll still be able to find it here, but before long the whole archive will move over to City Monitor.

On behalf of the City Monitor team, I’m thrilled to invite you to come along for the ride at our forthcoming digs. You can already follow City Monitor on LinkedIn, and on Twitter, sign up or keep following our existing account, which will switch over to our new name shortly. If you’re interested in learning more about the potential for a commercial partnership with City Monitor, please get in touch with our director of partnerships, Joe Maughan.

I want to thank and congratulate Jonn Elledge on a brilliant run. Everything we do from here on out will be building on the legacy of his work, and the community that he built here at CityMetric. Cheers, Jonn!

In the meantime, stay tuned, and thank you from all of us for being a loyal CityMetric reader. We couldn’t have done any of this without you.

Sommer Mathis is editor-in-chief of City Monitor.