How do you visualise chess games?

There are lots of ways you can interpret this question. As someone who asks and answers lots of data visualisation related questions, and themes an entire blog around it, I’ll touch on a few of them. Sorry about that …

First of all, chess is a battle game between two armies. So you could say that the game of chess is already a data visualisation. Each piece is a symbolic representation of a particular soldier in the army. Each army is represented by a different colour (which we always refer to as Black and White even if physical colours are different)


(image from – pieces above are the classic Staunton design)

If chess is a symbolised battle, then chess in its diagrammatic form is further visualised to another level. Unless you regularly play over the board games or tournaments, the chances are that most regular players in 2017 are more familiar with chess games on computers, in books or on websites. Below is a visualisation of the beautiful final position of Anderssen vs Dufresne’s “Evergreen” game of 1852.

Screen Shot 2017-04-02 at 20.19.37

For a long while now I’ve been considering a different way of visualising chess games. Not by way of improving traditional chess diagrams – I’ve studied more thousands of these over the years than I care to admit, and “if it ain’t broke, don’t fix it”. But is there a way to visualise an entire chess game in one diagram? Not to study individual moves, pieces or positions, but to represent a full game pictorially.

Discussions among chess players, experts and analysts has often centred around the question: “Is chess an art, or a science?” The fact that it is considered both by different proponents of the game means we shouldn’t be surprised that it is an interest shared by many data visualisation exponents. Many enter the profession as analysts and learn artistic skills, many enter as artists and learn analytic skills. Many have both talents, and more, in abundance. Edward Tufte, in some people’s eyes the most influential name in data visualisation, is, I believe, a keen player and has blogged on chess and chess-related visualisation before here. Of lesser influence perhaps, but of equal importance, my friend and fellow #MakeoverMonday participant blogged here, less than 24 hours ago, about the positioning of participants on the “vizzer” spectrum between art and analysis.

Moving away from art and science, as mentioned above, a game of chess represents the movement of two armies over the course of a tactical battle. This brought to mind one of the most famous and instructional data visualisations of all time, Charles Minard’s famous visualisation (below) chronicling the movement and size of Napoleon’s troops in their 1812 campaign against the Russians.


I won’t go into this in huge detail, suffice it to say that if you’re not familiar with this, it’s sourced from Minard’s entry in Wikipedia, and that would be as good a place as any to learn more about this. It tracks army strength, size and positioning and represents it as a flow on a diagram. Can we do something like this for a chess game? Instead of Anderssen’s finishing position in diagrammatic  form above, can we depict the flow of his pieces (armies) throughout his conquest of Dufresne’s armies?

Another visualisation type that has interested me recently has been the “centre of gravity” chart. These charts show how average position of something on a map has changed over time, thus creating a trail year on year (turn by turn?) To illustrate what I mean, here are a couple of examples:


The first chart (above) is from showing how the average position of the Stanley Cup winning team has gradually moved from Canada over the border towards Detroit.


The second chart (above), from the Economist, shows how the World’s economic centres have moved from Mesopotamia across the continent towards the North Atlantic (as the “gravitational pull” from the US takes effect, before a boomerang effect back over Northern Europe as the pull of far Eastern economies takes effect.

Combining these two effects gives me what I need. Taking the average position of all white and black pieces move by move, weighted by the size of each “army” will show the slow movement of overall mass of white and black pieces. I’ve loaded in data for three classic games, one from each of the last three centuries. The “Evergreen” game, mentioned above, is included, as is the first game of the iconic 1972 World Championship between Spassky and Fischer, and the decisive game in the most recent World Championship – Carlsen’s victory over Karjagan in 2016. I’ve included a slider so that the user can follow along with the game in traditional notation and diagrammatic form, whilst watching the visualisation unfold. More details are shown by hovering on the top right black pawn for information.



I’m pleased with these and the story they tell. The first game (Anderssen vs Dufresne) shows how thin the white line becomes as piece after piece is sacrificed as the remaining armies push forward towards ambushing the black king. The second game (Spassky vs Fischer) shows Black’s downward right push as his bishop grabs material in what ultimately is proven to be an unwise manoeuvre – the long thin trail towards the Queen’s side shows Spassky’s movement towards a winning ending advantage on the opposite side of the board. The thicker lines in Carlsen vs Karjagan show the continual manoeuvring of major pieces (the higher values queens and rooks) and the ultimate win when white’s pieces are positioned so much more on the king’s side of the board than black’s pieces to exploit a quick finish.

They showed what I wanted them to show – an “artistic” capture of a full classic game of chess. But I’m always grateful for feedback or suggestions. In this particular case, suggestions of future games to include would be great – so long as you appreciate that the input of data is a bit of a labour of love! One person I asked specifically for feedback was Nicholas Rougeux, whose work I mentioned was a particular influence here. I see these charts as being quite reminiscent of something he might produce and he expressed a genuine interest to see how these visualisations would turn out. His suggestion was to show a trail for average position of each individual piece, not the whole army en masse. Might that be easier to understand and interpret? Here are the results below, shown at the games’ winning move in each case:

anderssen grab

I love this! First, the evergreen game: the centralisation of white’s rooks to the centre and the diagonal ranges of his queens and bishops are obvious. The futile raid of the thick black line (his queen) and the small dots on both sides representing unmoved pawns tell the story of how short the game was.


Next, Spassky’s win. What sticks out is black’s bishop raid into the bottom right corner to pick off white’s pawn and the gradual move of white’s pieces up the board. Most pawns have moved forward to contest central squares here.


The Carlsen game shows all the manoeuvring of thick trails, particularly black’s queen and rooks jockeying for position, particularly all his rook moves on the (left hand) a-rank (column). Ultimately it was fruitless as White made his move towards black’s king in the back right.

I’ve written a lot about visualisation and chess – thanks for sticking with me if you’ve been following it this far. Thank you to enthusiasts of both for your suggestions, and thank you in advance for those yet to come, please do let me know as I intend to update my online Tableau version with future games and improvements.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s