This is The Game Is A Foot, a site which uses mathematics to provide a new look at the world. For those of you with a fear or dislike of mathematics, don't fret, not only is the maths on here not too complicated, but there will also be plenty of analysis with lots of adjectives, adverbs and all those other words.
We have two broad areas of interest: sport and politics.
We are also happy to put up guest articles; please get in touch if you wish to contribute.
With the days until the General Election counting down, we look at tactical voting recommendations made by the Observer, before giving our own advice. This was updated following new information.
With yet another election called, we wheel out our General Election model, and get some early impressions. With values available for Northern Ireland, Wales and Scotland, we have a look at the key battles in each nation. Eventually, our full model is ready and we see how each party is looking. We use this to assess the chances of various independent candidates standing.
Ahead of the European Elections, we look at how tactical voting can be used to limit the number of Brexit party MEPs in Scotland, England & Wales.
The World Cup model is updated to give the team's chances of success ahead of the second round and quarter-finals.
It's World Cup month, so we've launched our World Cup simulator! This article explains how the simulator works, whilst this one covers how knockout fixture odds are created. The model is run 3000 times to given an idea of each team's chances. All this is used to identify high-value bets. During the tournament we had a look at which teams look likely to progress after the first and second round of group games.
I was very proud to be asked to present a public lecture at the University of Stirling on the use of mathematics in sport. The lecture was recorded and is available here. Later in the month I appeared on the show Campbell's Footballs, and to prepare for it I updated my xG model - the latest data is summarised here.
The xG model we've created is used to give an update on how the teams in the SPL are progressing.
Unsurprisingly, after the mayhem that was analysing the election, we have a bit of down-time. Part of this is because of me starting a new blog on Football Manager. Anyway, our first article in a while introduces a new area of analysis for us, and how it'll be used on the SPL. This model is used to check in on how the SPL sides are doing so far.
Ahead of the UK General Election, we unveil our new Election Live page, with updated coverage of the vote. This includes an article explaining all the details of the page. On the day of the vote itself we give our final predictions.
Following a (rather fun) election night, we review how our model performed. Firstly, we look at our predictions for individual parties, then we look at individual seat probabilities.
After gathering the full vote data, we are able to see how the election would have looked under a proportional system.
Two years to the day of the 2015 General Election results coming through, we unveil our model for the 2017 contest! We begin by introducing the model's methodology and making some initial predictions. This is followed by a study of how a Progressive Alliance could prevent a Conservative majority. Firstly, the arithmetic of what is needed is calculated, which we then use to provide a guide to the south of England, the north of England and Scotland, Wales & Northern Ireland. We look at vote shares to see what happens if, however unlucky, the election finishes with equal votes for the Conservatives and Labour.
Additionally, TGIAF now has its own Twitter page, which you should definitely follow!
After doing some online digging, it appears that an article from this website has been used in two academic papers! The papers 'Intrinsic Classes in the Union of European Football Associations Soccer Team Ranking' and 'Seasonal Linear Predictivity in National Football Championships' used our article on predicting football results using odds. Additionally, our guest article on the game theory of F1 appears to be used in lecture notes at the prestigious Cornell University.
Meanwhile, we continue our look at Premier League transfer policy, by identifying which clubs have done the best business over recent seasons.
We then have our second guest article, with a look at the different points systems used in Formula 1.
With my PhD finally completed, I can get back into writing articles. Using data on Premier League transfers we look at a series of aspects of transfer policy, starting with the effect of the fee paid. This is followed by looking at which nations have provided the best-value players to the Premier League and how well players have adapted when travelling over from different leagues.
With the US election entering the final stretch, we unveil our simulator for it.
Following a work-related hiatus, a new simulator is introduced on the eve of Euro 2016, with its pre-tournament predictions analysed.
On the eve of the Oscars, a full guide to the ceremony is provided.
Much like 2015, the year is kicked off with a look at the running for the forthcoming Oscars, which is followed up by introducing a new simulator for the Six Nations, which is available in full. The simulator's output is compared to the long-term odds provided.
A return to sports, with a look at the possible outcomes ahead of the final round of games in the Women's World Cup (Groups A-C, D-F). We begin to track the Labour leadership contest, and also investigate the idea of the "nervous nineties" in cricket.
Ahead of the General Election we introduce Election Live, the hub where our election model will be used to analyse the vote as the results come in. It also means we publish an article explaining all the various facts and figures we'll be using. A new update of the model allows us to see how the parties have changed in the week before the election, and which constituencies have caused that change. We follow that with a look at the marginal constituencies which will decide the election. On the morning of the election we post our final predictions.
Following a manic night keeping up with results, we are able to analyse how our model performed both in terms of predictions and method. This meant we were able to see whether our earlier configuration of the model was accurate and discover the effect that introducing swing would have had.
The lack of proportionality in the General Election results is compared to elections across Europe, and the D'Hondt method is studied as a potential alternative.
After uncovering odds from the 2010 election, we use these to recalibrate our General Election model, using the resulting model to make our latest predictions for each party and make recommended bets on the election. Continuing updates on the model allows us to look at the fight for seats in Scotland, as well as the chances of potential coalitions forming government. We also look at to what extent bookies weight their odds, and look to see whether the Grand National is well-priced or not.
Following the end of the Cricket World Cup, we look at whether the performance of the Associate nations has increased or not compared to previous contests. We also return to our General Election simulator, firstly with an update with the latest odds, followed by analysis on which parties are likely to be the decisive factor come May. The model is used to analyse target lists made by Labour, the Conservatives and the Greens.
Some more non-sports research, as a simulator for the 2015 Oscars is unveiled. An interactive version of the simulator is available here. The General Election simulator was also improved with the addition of a map showing the results of each constituency. With the Cricket World Cup starting, we begin tracking the probabilities of each team winning the tournament.
A move away from sports, for the first time, as we look at the 2015 General Election. Firstly, we use methods similar to our World Cup model to create a simulator for the election. We are able to run the model 1000 times in order to make some predictions for the upcoming vote.
The output of the second Premiership simulator is analysed.
In a diversion from mathematical models of football, we quickly look at the Commonwealth Games, and see if success in these Games predicts success in the Olympics. The first Premier League simulator is unveiled, although with a number of issues. A second simulator is thus created, using a different approach. Meanwhile, new pages are created to provide overviews of the work done on World Cup and Premiership modelling.
With the World Cup over, we have been able to review the quality of the TGIAF World Cup simulator, testing both the method of converting odds into probabilities and the method of creating odds. We also see if our recommended bets made any money. With the Premier League looming, a database of results and odds have been collated, which has allowed us to look at the success of basic betting strategies. It also means we can begin looking at creating a Premier League simulator, starting off by finding a method to create odds. This method is then used to look at the comparative strength of the Premiership and the Championship.
The simulator from last month is used to show which long-term odds on offer are good (and bad) value for money. During the World Cup, the permutations for the final group games are found for teams in groups A-D, and groups E-H.
Leading up to the 2014 World Cup, we have created a simulator for the tournament. This leads to many articles, firstly showing the mechanisms behind the simulation, namely how to use odds to find the probabilities of teams winning and how to predict bookies' odds. The simulator is also used to find the probabilities for each reaching any given stage.
The 2012 analysis I did on Paralympic versus Olympic success in Summer Games is followed up with a look at the same results for Winter Games.
TGIAF moves into new territory with our first article on snooker, as we look at which players are the best at coming through deciding frames.
To the shock of the nation, I have finally got around to creating a new article. With the Ashes approaching I'm looking whether Australia really are the underdogs they've been cast as. There is also the creation of a new page for the site, the Articles page, where you can view the profiles of authors on the site, as well as some other titbits.
With the start of the Ashes comes the TGIAF Ashes diary: Day 0, Day 1, Day 2, Day 3, Day 4, Day 5.
If you scroll down to the bottom of this page, you will see that the first entry is exactly one year ago! And to celebrate this, we had a very exciting (and tricky) competition for readers to enter. You can view the answers and the results of the competition here.
The site has a new look! Hope you like it! The banners of each page feature snaps of famous footballing moments, see if you can name them!
After a hectic month of, er... watching sports... with Wimbledon, the European Champions and especially the Olympic and Paralympic games making for a memorable time. Anyway, with the incredible summer of sport drawing to a close, here is an article comparing various countries' success over the two Games this century.
Here is the second part of the analysis of Premier League stats. Here we look at what affects league position the most - corners, penalties or shot accuracy? And after a confusing finale to Group A in Euro 2012, here is our guide showing what is needed for teams in the other groups.
With University out of the way, this is the first article breaking down the database of Premiership statistics, and answering a key question - is it more important to score goals, or avoid conceding them?
I am absolutely delighted to say that TGIAF's penalties and game theory article has been reproduced (with permission) on the Youth Football Scotland website! Check it out! There is also a wordy look at why new managers often seem to have an instant impact.
We have our very first guest article! Another article on game theory, but this time looking at competition between different racers in Formula One. Also added is a catelogue of the best sources around to learn about sports, as well as an introduction to future work on Premier League statistics.
A brief study of the footedness of goalkeepers and outfield players, as well as an explanation of logarithmic scales, which are used to assess whether the media perception of Premiership wages is reactionary. Additionally, the phrase "award-winning" is added to the game theory article. Why? Because it won an award.
There is also an expansion into cricket articles, with a study into how we can compensate for the opposition faced when looking at player's batting averages, followed by a look at how age affects batting averages, which will be of interest to both fans of cricket and normal distributions. Also a new page, which shows concisely the information found from each article is added.
Website goes live! Starting with two articles, one a brief look at newspaper transfer rumours, and a very long-winded look at the link between penalties and game theory. Followed up with a look at which positions England are best at producing, as well as my idea for a restructuring of UEFA qualification for major tournaments.