Month: March 2016

Football attendance figures are inaccurate and don’t tell the whole story


It can be useful to know how many people attend a football match but the official figures are not accurate. Blackpool fans have been collecting data and working out more accurate attendance figures for their team. They hope it will help them oust the football club’s owners and help a fan-owned club to emerge. The footballing authorities should encourage clubs to publish the more accurate data that they already collect. In the meantime the technique could help fans of other clubs.

Lots of people want to know attendance figures but it is not easy to get hold of accurate data

The police, and other services need to know the attendance figures so they can help maintain safety or in case of an emergency. The home team need to know to keep track of finances, look after health and safety and make sure there’s enough pies and bovril available. The away team might need to know the number so that they get a share of gate receipts. The local council might want to know so they can plan transport or understand the number of visitors coming to their town every other week. The football authorities want the numbers so they can proclaim how important the sport is. Fans want to know the figures too.

A snippet from the Millwall-Blackpool programme from 5th March 2016. Attendance figures are in the 6th column.

An attendance figure might be announced at a match. Printed in a programme. Published in a match report. Gathered on football stats websites with information about clubs, leagues or across the globe. The attendance figure you see in these places is typically wrong.

https://twitter.com/WillWatt/status/699679468760203264/

In England the rules of the premier league say that attendance figures that are publicly reported should be the number of tickets that are issued or sold. Not the number of people who actually attend. (I understand the football league rules say the same, although I couldn’t find the precise clause.)

The actual attendance figure can be quite different to the number of tickets that are issued or sold. Season ticket holders may not turn up for a midweek game. Complimentary, i.e. free, tickets could be issued but not used. People might even walk out in the middle of a match as a protest.

In the case of my club, Blackpool FC, fans wanted to know more accurate figures due to their own protests.

The Oystons, Blackpool, bids for the club and a boycott

The owners of Blackpool football club, the Oyston family, have loaned millions of pounds from the club to other companies, taken legal action against fans, taunted them and abused them by text. Unsurprisingly most Blackpool fans want the Oystons to leave.

In 2015 there were two bids to buy the club. One from an unnamed consortium of bidders. The other from the democratically run Blackpool Supporters Trust. The second bid would have created a fan-owned club. Both bids were rejected. Until the Oystons go many fans have chosen to boycott the club.

Without accurate attendance figures it’s tricky to understand the impact of the boycott. Whilst the fans can speak eloquently of the damage the Oystons have done to them it might be tricky for a bidder, or the local council, to understand the damage that has been done to the club and town due to the dreadful mismanagement of the Oystons.

Crowdsourcing better figures

So, using four people and a couple of hours of effort per game, we worked out better attendance figures.

254 people in Blocks D and E during the Blackpool-Coventry game.

Two people in the ground took pictures of every block. Two other people then used a fairly simple and manually intensive, but effective, technique to count the number of objects in a photo. A small percentage, 5%, was then added to cater for some people not being in their seats at the time of the pictures.

The actual attendances for three home games were as follows:

  • Blackpool v Shrewsbury (February 13): official figure 6,873 (including 790 away fans), revised crowd count 4,289— that’s 62% or 2,584 less than the official figure.

Blocks D and E during a game at Bloomfield Road in 2006. Picture by Matthew Wilkisnon, CC-BY-2.0
  • Blackpool v Bradford (February 27): official figure: 8,780 (including 3,063 away fans), revised crowd count 6,100 — that’s 69% or 2,680 less than the official figure.
  • Blackpool v Coventry (March 12): official figure 8,869 (including 3,000 away fans), revised crowd count 5,845 — that’s 66% or 3,024 less than the official figure.

Before the boycott the home stands were full. As the Blackpool Supporters Trust say:

Blackpool FC states that there are currently 4,600 season ticket holders. These accurate crowd counts would suggest that over half of the season ticket holders are already boycotting home matches.

Undoubtedly this will have serious implications for next season — people not using tickets they have already paid for are unlikely to purchase new ones.

The Oystons are destroying Blackpool football club.

Data helps us make better decisions

All of the turnstiles at Blackpool, like many other clubs, have a system that scans tickets on entry. The club’s owners have access to more accurate data than they publish. We have to hope that all football clubs are sharing this accurate data with each other, with local councils and with the emergency services. Perhaps as well as calling for more transparent publication of data to stop abuses of power at Fifa the footballing authorities should ask all clubs to publish as open data both the real attendance figures as well as the number of tickets sold. It will help lots of people make better decisions. In the meantime if fans of other clubs want help in working out more accurate attendance figures then get in touch.

For Blackpool fans though the immediate decision is the choice they are making about whether or not to support the owners by spending money and time going to their football club. I’ve made my choice. I won’t go back. I hope that my boycott will help persuade the Oystons to go.


I will be joining a community march and demonstration organised by the Blackpool Supporters Trust and the Tangerine Knights on Saturday April 30. Blackpool fans are hoping to get 10,000 people to come along.

It doesn’t matter if you’re a Blackpool fan, or even a fan of football. The issues at Blackpool go beyond football. They affect many parts of the town and include concerns about transparency and accountability in the local council. What matters is that you care about people and communities and want to make things better.

Why not join us and stand up for a community fighting to get their football club back?

Bitcoin, money laundering, and the challenge that law faces in keeping up with technology

I’m interested in the new organisational structures that blockchains, or other forms of distributed ledger, might enable, the problems they might solve and whether solving those problems can help make the world better. Bitcoin is enabled by one type of blockchain. As Jeni Tennison of the Open Data Institute explained:

Blockchains provide a way to store information so that many people can see it, keep a copy of it, and add to it. Once added, it is very difficult to remove information, which can reinforce trust in a blockchain’s content.

Bitcoins are a form of digital currency and blockchains were originally designed as a system to manage them. Within the Bitcoin system anyone can access and add information (such as what Bitcoins are used for, where they were spent and when) to the blockchain.

Someone (*) told me a theory about bitcoin the other week. I don’t know if the theory is right, or even close to right, but it reminded me of something I’ve come across before. A familiar, although slightly different, pattern. This kind of pattern matching is common in humans. It can help us grasp new concepts quickly. Sometimes it can also lead us astray. Given the pace of legal frameworks that can be a bad thing.

Money laundering

Money laundering is how a criminal makes it hard to discover that their money was gained illegally. There are many ways to do it. After money is laundered it appears to come from a legal source. It is ‘clean’. The criminal can enjoy the profits produced by their illegal activity with a reduced risk of punishment. Money laundering techniques evolve over time.

Let’s take an example:

Alice sells illegal drugs for cash. Alice has no other reason to have large amounts of cash hanging around. If the police find Alice with her cash then she will be in trouble.

Alice walks into a betting shop and puts the cash into a slot machine. Put a pound in. Press the button. Put a pound in. Press the button. It gets repetitive. But every so often there’s a big win. Alice ends up with £300 in cash in her pocket along with a receipt saying that she won it by putting in a single pound.

Slot machines have a built-in advantage for whoever runs them so Alice ends up with less money than she started with but it was worth it. If anyone asks then Alice now has a good reason to have the £300. She’s got clean money and is less likely to go to jail. Lucky Alice.

I know Alice’s story is true. It has become easier in the UK since the liberalisation of slot machines in 2001.

Here’s a more complex money laundering example. One made possible by newer technology and business models:

Roberto has a regular income in his country. He makes his money illegally: perhaps from drugs, perhaps from robbing people, who knows. Every week he gets some money coming in.

Roberto heard of M2M (machine-to-machine) and IoT (internet of things). M2M is when machines communicate with each other. IoT is when that happens across the internet. Communications cost money. At a basic level it is the machines and the cost of the network that connects the machines together but money might also be paid to intermediary organisations that pass on a message or that translate between the languages spoken by different machines.

Roberto spots an opportunity. He sets up some machines in both his home country and another country. He uses his illegal money to pay for the energy consumed by the machines and of wages of people to look after them. Roberto also sets up an intermediary organisation. His machines talk through the intermediary and pay a fee. The intermediary organisation helps turn the dirty money into clean money.

If Roberto sets up the intermediary in a third country he could produce the clean money in a different currency and legal jurisdiction. He could even pass the transactions through multiple intermediaries making it very hard for anyone to trace it back to him. Lucky Roberto.

I think Roberto’s story is true. Once upon a time I thought I saw some of the breadcrumbs left behind by his machines as they went from one intermediary to another.

A theory about bitcoin

And now for the bitcoin theory. It is not the usual story about money laundering by exchanging cash for bitcoins. It is a little more complex.

The blocks in the bitcoin blockchain store information. The information might say things like “Jerry transferred money to Jane”, “Brenda paid Mary” or any other message you like. Anyone can pass information to the bitcoin network for storage. Machines called ‘miners’ compete to create the next block by doing work to solve a complex computer problem. The miners are rewarded a number of newly created bitcoins for doing this work. This is called a proof of work system. It reduces the chance of two miners storing the same or conflicting information at the same time and helps the whole system to function.

Let’s imagine the tale of Francesca:

Francesca has money that she has received illegally and wants to launder. She has spoken to Roberto about his machines. Francesca has also heard about bitcoin but doesn’t trust anyone to change her money into bitcoins. She decides to combine the ideas.

Francesca builds a bitcoin mining farm. She hires people to run it. She buys specialised hardware to maximise her return. She pays the electricity bill. She stores information for people. (Even people like me who store silly messages about their football club on the bitcoin blockchain.) She receives bitcoins in return.

The bitcoins are controlled by whoever holds the key. Francesca can buy things with the bitcoins or other people that she trusts with the key can. Lucky Francesca.

Now I don’t know if the theory is true. I suspect it will not be a great money laundering scheme. After all bitcoin is more transparent than it is private. Perhaps Francesca’s aim is a more focussed one of transferring money from one country to another. To transfer it away from the control of a particular government to somewhere safer.

New technologies create new patterns of behaviour that are hard to predict

These patterns and the changes from Alice to Roberto to Francesca — one money laundering to avoid the local police, the next to avoid all authorities, the final one to, perhaps, avoid a single government — shows how hard it can be to predict how new technologies will be used.

Our governments want to stop money laundering whether it takes place through bitcoin, other cryptocurrencies or new technologies that follow. To do this governments will have to understand how both the bitcoin blockchain, and the many other types of distributed ledger, might be used for money laundering. They may think that we need to follow a pattern from old technologies and demand that we know people’s identity. If they do this then, as Joichi Icho points out, we risk more bad laws that could create adverse effects elsewhere. New technologies, and hence our societies, operate on a global scale with open standards that can be used and influenced by anyone. One country’s defence against money laundering or international terrorists might help another country’s moves to stop internal political opponents.

If we are to make the world a better place using new technologies then perhaps one of the biggest challenges we need to think about is the legal and political one. Our technologies and societies are becoming more global open, agile and dynamic. How do we create legal and political frameworks that aren’t rigid and take years to come into force but are instead adaptive and iterative enough to keep up with the changes in technology and society?

(*) I’m happy to add their name if they wish.

Improving UK voter registration

Voter registration is a problem in the UK.

If we opened up aggregated data and registration APIs then we could build a network of services to help get people registered. As a result we would get more people registered and, hopefully, more people voting in our elections.

Millions of people are not registered to vote

In January 2016 there were reports that 800,000 people dropped off the electoral register last year. That 800,000 is on top of the Electoral Commission report in July 2014 which found that around 7.5 million of people who are eligible to vote weren’t on the register.

That’s a lot of people. If we can get them registered (and engaged enough to turnout and vote) then it could help increase the turnout in UK elections.

Wouldn’t that be nice.

Voter registration data is available but only shared with some people

Voter registration data is managed and held locally by electoral registration officers (EROs). They share it with some people.

There are 2 versions of the electoral register — the ‘open register’ and the full version. You choose whether to go on the open one when you apply to register. The ‘open register’ is available for inspection at council offices or to purchase. People build websites with it and use it for marketing.

Many people do not want their name and address available for lots of people to see so they keep their details off the ‘open register’. The full version of the electoral register can be accessed by fewer people. Some of the people who are allowed to use it include the election staff who post out our voting details and political parties who get occasional releases and use it to knock on our doors and ask us to vote for them.

When election results are announced we get all to see an aggregated summary of the voter registration data. The Electoral Commission uses the total number in election results and to calculate the turnout in the election.

Other than that we rarely get to see the data. The Office of National Statistics publish a yearly update and the Electoral Commission publish historic data but nowhere shows a total count of the current number of people registered to vote (*).

Many organisations are improving voter registration

Government launched a new easy to use website in 2014 and runs voter registration campaigns whilst political parties also use the occasional release of voter registration data that they are given to target unregistered voters. Meanwhile lots of other good organisations (such as Bite The Ballot, Operation Black Vote and Hope Note Hate) are also working to get more people registered.

Releasing regular aggregated data (“X people are registered to vote in area Y”) to these organisations will help them target their time and money. These groups do not need access to the full register and they should not get it. I doubt these groups even need the ‘open register’. I suspect simple regular releases of aggregated data for particular areas will make a big difference to their efforts.

If the government can release national totals for voter registration every day then aggregated voter registration data for local areas can also be released every day. It would create more informed debate about the challenges of voter registration and help target efforts to increase registration.

Better voter registration services

We will continue to need paper forms for people to register (never forget that 12 million people, 20% of UK adults, are not on the internet) but we can also build better online services. As well as using aggregated registration data to campaign we could also use APIs.

Rather than relying on one website we could use APIs to build a network:

  1. A citizen could inform their local authority that they have moved address and allow them to update the electoral register at the same time.
  2. A voter registration organisation could translate the current website into one of the many other languages spoken by UK citizens.
  3. A political party could register its members.
  4. A student could allow their university to update the electoral register.
  5. A tenant could allow a housing association to update the electoral register.

Many better ideas will exist. Open APIs allow innovation to flourish.

A voter registration API could follow exactly the same rules and regulations as the current website. Ideally the API would also allow individuals to check if they are registered. This would reduce the inevitable duplicated requests that must cost councils time and money. The Government Digital Service could monitor use of the API and get action taken against people who misuse it.

To be clear this is not about opening up the code for the API. It is about allowing people to use a voter registration API in the same way that HMRC plans to allow people to use its APIs. Once the voter registration API is open then any local council, university, civic technology organisation, housing association or voter registration campaign can pick it up and build it into their services.

Opening up voter registration will make things better

Voter registration is clearly a problem in the UK. The simple new website helps but voter registration is not a problem that a single website will solve.

Government could choose to go further than I’ve outlined in this blog and redesign voter registration. Such a redesign would hopefully consider better open data and open APIs but also consider bigger questions such as whether people could be automatically registered under some circumstances or whether the data could be managed within a central platform rather than the distributed network of systems managed by Electoral Registration Officers. It is not clear if government has the appetite for such a redesign or how long it would take so I’ve focussed my thoughts on pragmatic things that could help tackle the known problem of under-registration (**).

It is every government’s responsibility to ensure voter registration works. By opening up aggregated data and building open APIs, government can improve how it works with the existing network of voter registration campaigns, civic technology organisations and local authorities and make things better.

— — — — — — — — — — — — — — —

(*) Sentence updated after publication to be more precise.

(**) Paragraph added after publication to be clear on pragmatism v redesign.

© 2020

Theme by Anders NorenUp ↑