Iron Viz Feeder Retrospective: Water

‘I should have got a stronger grip on her,’ wrote Lord Montagu in a letter home from his sickbed in Malta in 1916, after being rescued from the wreckage of the SS Persia which was hit by a German torpedo while crossing the Mediterranean.

But to his enduring pain, Eleanor Thornton, his travelling companion, personal assistant and beloved mistress, had not been saved.

Love at first sight: Eleanor Thornton and Lord Montagu. “Theirs was a great love affair. Although when he came back home he was badly injured, he spent days looking for Thorn, who had been thrown overboard, searching everywhere, hoping that somehow she would turn up.”

Of course, she never did. But though the affair between the aristocrat and Eleanor Thornton ended with her death, their love was immortalised in the most unlikely of places.

It was the inspiration for the Rolls-Royce flying lady, or ‘Spirit of Ecstasy’, whose soaring curves are modelled on Thorn and recognised by motorists across the world as a symbol of quality and distinction.

Montagu’s wife, Lady Cecil, not only knew about the affair, but condoned it. For her part, Eleanor had a child by Montagu but, knowing that as a single mother she would be unable to continue to work for Montagu, gave her daughter up for adoption.

Born in Stockwell in 1880 to a Spanish mother and an Australian engineer, Eleanor Velasco Thornton left school at 16 and went to work at the Automobile Club (now the RAC).Through her work, she met all the motoring pioneers of the day, among them John Scott Montagu.

Rolls Royce

Montagu was a charismatic figure, educated at Eton and Oxford, with a great interest in travel and transport.An MP for the New Forest Division of Hampshire, he was a great car enthusiast, who came third in the Paris-Ostend road race in 1899 and is credited with introducing King Edward VII to motoring.

The love affair of Eleanor Thornton and Lord Montagu was the inspiration for the ‘Spirit of Ecstasy’

But he was also married to Lady Cecil ‘Cis’ Kerr, with whom he already had a daughter. When he met Miss Thornton, however, the effect was instantaneous.

“I fell in love with her at first sight,” he later said. “But as I couldn’t marry her I felt I must keep away from her as much as I could. But she began to like me and realise my feelings as well.”

In 1902, when Eleanor was 22 and Montagu 36, she went to work as his assistant on Britain’s first motoring magazine, Car Illustrated, in an office on London’s Shaftesbury Avenue.

He explained: “Before long, we discovered we loved each other intensely and our scruples vanished before our great love.”

It was a love whose light never went out. When Montagu’s father died in 1905, John Scott inherited the title, becoming the second Baron Montagu of Beaulieu, and moved from the House of Commons to the Lords.

Miss Thornton was still very much on the scene, increasing her duties as his assistant accordingly. Montagu owned a Rolls-Royce and would often take her for a spin along with Charles Sykes, an artist and sculptor. This is how Thorn came to inspire, and model for, the Spirit of Ecstasy.

Rolls Royce

Montagu was friends with the managing director of Rolls-Royce and between them they cooked up a plan for an official sculpture, which at Montagu’s suggestion Charles Sykes was commissioned to design.

Sykes used Miss Thornton as a model and The Spirit of Ecstasy, or “Miss Thornton in her nightie”, as those in the know called it, graced its first Rolls-Royce in 1911.

Whether by this time Lady Cecil had worked out the truth about the relationship between her husband and his vibrant personal assistant is not clear, but by 1915, when Montagu had to leave for India to be Adviser on Mechanical Transport Services to the government of India, she certainly did.

It had been decided that Miss Thornton would accompany Montagu onboard the SS Persia. Before the trip, Miss Thornton corresponded with Lady Cecil. Her tone is tender and conspiratorial. “I think it will be best for me to make arrangements without telling Lord Montagu – so he cannot raise objections,” she writes.

Later in the letter she writes, tellingly: “It is kind of you to give your sanction to my going as far as Port Said. You will have the satisfaction of knowing that as far as human help can avail he will be looked after.”

According to Montagu’s biographer, the family felt that Lady Cecil “became resigned, with no feelings of bitterness to her husband’s affair and took the view that if he had to take a mistress then it was as well he had chosen someone as sweet-natured as Eleanor Thornton – rather than someone who might cause a scandal.”

P1040801 ss persia mir copy

But their days on the SS Persia would be the last Montagu and Thornton spent together. They boarded the ship in Marseille on Christmas Day in 1915. Five days later, on December 30, they were sitting at a table having lunch when a German U-boat fired a torpedo at the ship’s hull.

The massive blast was repeated as one of the ship’s boiler’s exploded. As the ship began to list, icy seawater rushed in through the open port holes, and in the mayhem, Montagu and Eleanor made for the decks, which were already beginning to split.

They considered trying to find a lifeboat but there was no time. One moment, Montagu had Eleanor in his arms, the next they were hit by a wall of water and she was gone. The port side of the ship was submerged within minutes and Montagu was dragged down with it. He was wearing an inflatable waistcoat and this, along with an underwater explosion that thrust him to the surface, probably saved his life.

“I saw a dreadful scene of struggling human beings,” he later cabled home. “Nearly all the boats were smashed. After a desperate struggle, I climbed on to a broken boat with 28 Lascars (Eastern sailors) and three other Europeans. “Our number was reduced to 19 the following day and only 11 remained by the next, the rest having died from exposure and injuries.”

They were eventually rescued, after 32 hours at sea with no food or water, by steamship Ningchow. Montagu convalesced in Malta, then returned home where he was flattered to read his own obituary, written by Lord Northcliffe, in the Times.

The accident left him physically frail, but for years Montagu continued to search for his beloved Thorn. He also erected a memorial plaque in Beaulieu parish church beside the family pew, giving thanks for his own

“miraculous escape from drowning” and “in memory of Eleanor Velasco Thornton who served him devotedly for 15 years” – an extraordinary public display of feeling under the circumstances.

Lady Cecil died in 1919 and Montagu remarried the following year, to Pearl Crake whom he met in the South of France.

She bore him a son, Edward, who is now the Third Baron Montagu of Beaulieu. But the repercussions of the love affair did not end with the deaths of the two women involved.

The current Lord Montagu takes up the story. “My father died in 1929, when I was two and that was when the family discovered, by reading his will, that Eleanor had had a child.

“The will made provision for her, but was worded to obscure who she was. We always used to wonder who she was and were keen to find her.

“Then my half-sister Elizabeth went to live in Devon. She was standing in a fishmongers queue one day when someone said to her: ‘See that woman over there? She’s your sister’.”

The woman’s name was Joan. She was born in 1903, soon after Montagu and Thornton began their affair, and given up for adoption straight away. The curious thing was that while Eleanor made no attempt to contact her daughter, Montagu had, on occasion, met up with her.

He also wrote her a letter explaining the circumstances of her birth – “Your mother was the most wonderful and lovable woman I have ever met… if she loved me as few women love, I equally loved her as few men love…” – that she did not receive until after his death.

Joan’s behaviour was as discreet as her mother’s. She had attended her father’s funeral, but so quietly no one even noticed she was there.

Says the current Lord Montagu: “Eventually, I got in touch and took her for lunch at the Ritz. We had oysters and she said: ‘Your father always used to bring me here and we would have oysters, too.”

Joan married a surgeon commander in the Royal Navy and had two sons, one of whom, by sheer chance, worked for Rolls-Royce.

Lord Montagu did as he knew his father would have wished. “I recognised them as full family,” he says, apologising for the tears on his cheeks as he recounts the moving story.

And so, a century after Eleanor Thornton and John Montagu met, their story has now passed into history.

But the spirit of their feelings lives on, in the form of the figurine that still graces every Rolls-Royce.

abridged version of The Great Rolls Royce Love Story,

printed in the Daily Mail,1 May 2008


You’ll please forgive this rather random introduction. It was the above story (not the Daily Mail version midn you) in the Buckler’s Hard Maritime Museum (in one small corner of a much larger museum), discovered while on on holiday two weeks ago, that peaked my interest. [Not just mine either – you can find out much more at]

The SS Persia sank on was sunk off Crete, while the passengers were having lunch, on 30 December 1915, by German U-Boat commander Max Valentiner (commanding U-38). Persia sank in five to ten minutes, killing 343 of the 519 aboard.

The story of Lord Montague, his love affair, his inflatable waistcoat that had been purchased in advance of the trip and likely saved his life (the actual jacket is in the museum), as well as the trip of a British Steamer through openly hostile waters of the Mediterranean and the U-Boat Commander’s apparent disregard for civilian lives struck me as fascinating.

Having long held a fascination with U-Boats when I was deciding on a theme for my Water Iron Visualisation I decided to research U-Boats of World War 1 in more detail. You can see my visualisation below (click the picture to explore).

Data Gathering

Looking for data I found, this amazing site lists every ship attacked by U-Boats in both World Wars and also plots the location of the majority of them where know. It also includes details of the attacking U-Boats, their commanders as well as the fate of the U-Boat itself. It’s a literal tour de force and I spent a long time perusing the site.

I have difficult time with the morals of scraping data from third-party sites, clearly someone has put time into collecting this data and it’s theirs to share but I decided to use the data under the principles of fair use. I already have access to the data, made a whole copy, I am using the data for computational analysis and study and my work will increase interest to the site (well I certainly hope it will). I have also attributed the data. let me know your thoughts on a fairly grey area that I (and others) struggle with.

I used Alteryx Download tool to download the data

This looks complex but actually the initial starting point was just half of the first row of tools to download the ships hit. As I built the visualisation I added other segments that downloaded details of the U-Boats, their fates and then parsed out their sinking locations from the notes using Reg Ex, as well as joining on Ocean areas geographically. Each task was a discrete item consisting of a few tools after I’d spent time with the data set in Tableau building and analysing.

This back and forth workflow is key to the way I work and I talk about it a lot, I make no apologies for that though as it really is essential to the way I think and work. I had a few draft design ideas in my head but I never sketch things out or plan them ahead of time – I prefer to let the data guide my hand and eye, eventually using the analysis and visualisations I create to inform the final viz and design. I created 40 separate visualisations as part of this project and then stepped through them looking for stories in an effort to weave together a narrative.

Inevitably there are elements it’s frustrating to miss out, you’ll find no mention of the U-Boat commanders with the highest number of kills in the final viz for example – partly I didn’t want to glorify the horrific crimes they committed but partly they just didn’t weave into a coherent and structured story.

Design Decisions

Long form? Story-points? Wide form? Single page? I toyed with a lot of these in my head but eventually I opted for story-points as a way to tell the story. It sounds quite obvious but the decision was hard purely because of the design limitations story-points places on me as an author. The height and position of the story points for example can’t be controlled, nor can graphics be added outside the dashboard area. You’re also forced into a horizontal / left – right set of story points. All these felt like limitations that restricted me but with hindsight they probably made my life easier from a design standpoint and stopped me worrying about large aspects of the design.

My second difficult decision was how to entwine other charts into the visualisation and story. I built the map fairly quickly and knew that would be the centre piece of the visualisation but wasn’t sure how to bring other charts in without affecting that beautiful look Mapbox had helped me achieve with the map. I’m not sure I’m 100% happy with the result but by keeping colour schemes similar and also using a lot of annotations in both (and keeping the consistent top-left annotation) I think I just about got away with what I was trying to achieve. This consistency was a key factor at play in designing the whole visualisation.

Data Visualisation and Analysis

From a visualisation perspective the biggest think I struggled with was the occlusion of the attacks on the map. My original larger points offered a good differential to see the different tonnage involved but the smaller points, second image below, offers less occlusion thus giving more visibility of the true density. Playing with opacity wasn’t an option (see below) and so I opted for the smaller points in the end despite them being a touch harder to see.

Would the new Tableau Density Mark type have helped:

In this case no I don’t think it would have – I prefer to show every single attack and sinking rather than obscuring them into an amorphous blob. Density mapping works to tell me where the main theatres of war were but doesn’t help pull out the main attacks and help the story in my opinion. Likewise, the two combined is even worse – yuck (perhaps improving the colours here may help).

So for those dying to see the new Density Mark feature, as with everything in data visualisation, my advice would be it’s best used sparingly and for the right use case.

Other data visualisation matters were in the use of the rolling averages and rolling sums – how understandable is a rolling sum of U-Boats sunk, it allows a useful comparison measure when there are low numbers (much more than an average) but takes some understanding. Here I opted to include the visualisation with some explanation.


Storytelling was to be a huge focus of the visualisation and a key to pulling the whole thing together. A timeline of the war was to be the structure and I was keen not to deviate from that – using data to illustrate points I’d pulled from research (with limited time it’s hard not to focus on just one source but I pulled several together and actually ignored some claims that were made in some source as I couldn’t justify them with the data I had).

I wanted the timeline to focus on the human and political aspects of the visualisation, as well as some the famous, and not so famous stories. I also wanted to get across the full scale and horror of these attacks – 1000’s dying trapped in a sinking ship under a storm can sometimes get lost as an emotionless point on the map. These aims were not easy to bring together, especially in a limited time-frame and I’m most disappointing about the last point – I feel I’ve not truly got across the nature of those deaths and the true horror of the sinkings. That said I feel it’s very hard to achieve in visualisations without falling into cliched visualisation types, all of which have been done before, and so rather than go down that rabbit hole I stuck to telling the stories in the limited space I had and trust the user to imagine the horror.

Balance between the text, the visualisations and keeping the length of the story relatively short was another difficult aspect. Again I’m relatively happy with the result but I had to sacrifice a lot of the detail and miss out interesting stories and acts of heroism in an effort to keep it short (including my original inspiration above). I’ve no real advice here for others except that it is a constant juggling act and something you’ll never be truly happy with.

Wish list

I thought I’d finish with a list of things in Tableau that I found hard, or harder than they should be. Spending just 24 hours on data collection and visualisation of this scope is testimony to the geniuses who build Tableau and Alteryx but there’s always ways it could be improved – here were some of my pain points and how I overcame them.

Borders on Shapes – I wanted to keep the U-Boats sinkings a different shape from the attacks (circles) for obvious reasons, but the result was the circles lost their border….not ideal! For aesthetic and analytical reasons I think the border adds a lot.

The workaround I applied was to duplicate the data and add a tiny portion to the size of the shadow data point, then colour the two differently. I then ordered the points in the detail pane by an ID so the shadow appeared directly below the mark and no other marks were between them. Each “mark” in the viz was actually two, one real and then the shadow below it.

The real colour legend below shows what I had to do here. This was a reasonable amount of effort to work out and then removing the shadows for aggregations elsewhere was a pain (I could have used two data sources I guess).

Of course the result was I couldn’t change the opacity either.

MapBox – I had to add my own bathymetry polygons to the North Star map as their version of the bathymetry rasters didn’t work about Zoom Level 3, I needed to zoom out further! This took a lot of faffing in map box studio (probably because of my inexperience).

Annotations – please, please, please can annotations be easier in Tableau, losing them every time I add a dimension to the visualisation is painful. Not to mention using them with pages in Storypoints seems to be slightly hit and miss – at one point I lost all of them, whether it was my fault or not I’m not sure but it was a painful experience.

Storypoints – as an infrequent user of Storypoints I generally found them easy to use, however I look forward to the day I can customise and move the buttons as much as I can style the rest of my visualisation

Alteryx – I’d love some easy web page parsing adding to be able to select patterns in HTML and pull them out – I do this so often! I’m a dab hand now at (.*?) in regex but I’d love there to be more.

Iron Viz – I go on holiday for two weeks every August for two weeks, please consider using a different time for the feeder to reduce my stress levels when I return 😀

Thanks for reading – I’d love to hear your critique of my efforts in the visualisation, what worked what didn’t. Also consider giving it a “favourite” in Tableau Public – I’d appreciate it. Here’s the link again:!/vizhome/TheU-BoatinWorldWarIAVisualHistory/U-BoatsinWW1





IronViz Feeder 2 – Retrospective

An interview with myself looking back at the recent IronViz Feeder for Health and Well-Being. Below is my final visualisation, click for the interactive version.


Perhaps I can start by asking you what you thought about this Iron Viz theme, did it get you excited – did you immediately have themes that sprang to mind?

To be honest I have a love / hate relationship with these feeders, on the one hand I love the open-endedness of the theme, yes it’s a guide but you can go almost anywhere with it, but for me it still feels too open to mean my imagination struggles to settle on one particular idea.

I was already doing some work for a Data Beats article on Parkrun and their accessibility and I initially wanted to cover this but what I had in mind didn’t fit nicely into a single visualisation or set of dashboards. I also had in mind some work on Agony Aunts – comparing the Sun’s Dear Deidre to The Guardian’s Mariella Fostrop based on the words they used – but the analysis started taking too long….

So that’s an interesting point – how do you balance out the various aspects of visualisation when choosing a subject? Do you choose subjects that require little data preparation so you can maximise data visualisation time or look for more complex subjects? 

When choosing a subject I’m primarily interested in choosing a subject that interests me. If the subject do that then it isn’t going to make me want to stay up til 3am working on it, or dedicate hours outside work. Let’s face it, I’m not in IronViz to win the thing, although I’d love to, there’s just too much talent and competition out there for me to compete, and so I’d rather just have fun doing a visualisation.

That said, I also don’t want to pick a subject that feel’s too easy – I like to work at my data and perform some analysis, I want to be able to say “This is what I found” rather than “This is what the dataset I found said”. The difference is subtle but I see this as a direction my public visualisation path is taking more and more lately. So I want to build and define my own analysis and say something with it – I do take inspiration from other sources, after all very little is new or novel today, but for me the analysis is as important as the visualisation itself.

This is where too the “data journalism” aspects of data visualisation are important, in the IronViz scoring criteria this is labelled as “Storytelling”.  However you label it I interpret it as not just showing the numbers. Anyone can show numbers visually, they can show the highest and the lowest and the relationship between them. They can design a dashboard and they can publish it. That isn’t data storytelling though, it’s data presentation. I want to convey why someone should be interested in the numbers, what do the numbers tell us and why is it important, and what should we do because the numbers show what they do.

So you mean adding commentary?

Well yes, but that’s only part of what I’m talking about. What I’m getting at is that this storytelling goes right back to the data choice, the subject choice and the analysis. And it’s not about presenting numbers back that people should care about either; it’s about doing some meaningful analysis and telling a story that is different, not the same old numbers presented in a different way.

It sounds like you feel the way you approach IronViz now is perhaps different to the way you’ve approached it in the past. What’s prompted it do you think?

Certainly it’s been a journey to get to this point, probably starting with my Springs got Speed visualisation in last years Iron Viz. As to what has prompted a shift towards this more analytics direction, well I suppose it’s the same things that prompted Rob and I to start Data Beats. Sometimes you look at the Iron Viz entries and you feel like you’re in a game where everyone is kicking the ball when suddenly someone comes in, picks it up and starts running with it. Over the last year or so the norms in the Tableau community certainly seem to have shifted; what was considered good a few years ago is now very, very average and people are pushing boundaries left right and centre.

When people start pushing boundaries you really are left with two choices; you can either find your own boundaries to push or settle down and try to do the basics really, really well. So while perhaps in the past I was happy to push boundaries, there are now others who do much wilder stuff than I ever could – and so I really need to hunker down and do the basics as well as I can.

So tell us about your Iron Viz. Where did the idea come from and how did you choose to approach it?

I decided to look at how deprivation is linked to the number of takeaways in an area, looking back I think like any good idea it didn’t come from any one source, instead it came from several seeds over time. Certainly walking around my own town, which is in a relatively deprived area I see a lot of takeaways, we get new takeaway leaflets every day and, where once the town centre was made up of lots of different stores now I see about twelve takeaways (contrast this to just two as I was growing up – in perhaps 500m of shops). There’s been some similar research on these links already, this article sticking in my mind recently.

Having explored other ideas and failed I knew I could get this one off the ground quickly – I’ve played with the Food Standards Agency’s Ratings data before and knew I could download that to get a classification of takeaways, while deprivation is easy calculated from the Index of Multiple Deprivation. So the problem seemed relatively simple given my limited time.

Speaking of time, how long did your visualisation take?

I didn’t have long enough, the world cup, TC Europe and several camping holidays meant that I really dedicated just the last day of the time allowed to this. I started about 3pm and, with several breaks to see to the family and eat, I was working til 3am.

I wouldn’t recommend this approach, it meant I had very little time to iterate on the visualisation, I had no time to get feedback from any peers and very little time to step back and consider what I was doing.

What took the longest time?

I settled on the data and story fairly quickly, using Alteryx to pull the data together. However the design was something that I hadn’t worked out before starting, and well over half the time was spent on trying to come up with ideas.

I started off the idea to put the article on a newspaper, partially covered by fish and chips (that’s how we eat them from takeaways traditionally in the UK); there were however several difficulties. First and foremost, I need any design to use images I have created or that are free to use. Finding an image of fish and chips at the right angle and with a transparent background was hard with no copyright was hard, also I wanted to have the article crease like it was folded which would have been quite a bit of work.

I quickly returned to the drawing board with very few ideas as to how to approach my visualisation design. I’d wasted 2 or 3 hours looking for images and I needed something quick. In desperation I googled “Takeaway” to look for related images and that’s where the takeaway menu hit me – and the idea was born.

The design looks quick complex – what software do you use?

I actually have a licence of photoshop I use for photography but I’m not a very good user, I can understand layers and some elements so I used that to piece together the design.

Wait, Photoshop? and you use Alteryx? Other people don’t have those advantages?

No, but let’s be clear I only use them because I have licences and have put effort in to learn them. All the work I did in Photoshop I could have done in Gimp, SnagIt Editor or even Paint. Likewise the data prep work could have been done in Tableau Prep (aside from joining on the spatial files which could have been done in Tableau) or I could have used other free software like EasyMorph.

So back to the design, where did the actual Takeaway design come from?

I copied took inspiration from this design from a menu designer, I then played around with the sizes and colours until I was happy.

What blind alleys did you go down in producing your Iron Viz?

Lots! Isn’t that what Iron Viz is all about. I really wanted to add an extra geographic element to the visualisation, and look at the relationship at perhaps 1km grid level. I did the analysis but the relationship I wanted to see just wasn’t there due to geographic anomalies i.e. town centres have a lot of shops but not many people. I tried extending the analysis out to 3km or 3 miles but there was too much noise in the data, remote areas were completely distorting the story and there were no patterns I could. In the end I settled for the simple analysis.

What was the hardest part did you find?

Having done so many Data Beats projects lately, I found it incredibly hard to limit myself to a single dashboard. I’ve got so used to using words to tell my story and explaining it over several paragraphs with visualisations to help me a long the way then this was incredibly frustrating – I had so much to say but not enough space to say it.

You said in the last Feeder you were too intimidated by the competition to enter. What changed your mind to enter?

I regret not entering the first feeder. My thought process came from my competitiveness – I really want to win this thing and I feel the competition is such that I might not be able to. Coupled with the fact the time has increased to a full month I really struggled to create enough time to compete with some of the stronger entries. Before a few hours was enough to compete, now it’s not even close.

But my thought process was wrong, trying to win Iron Viz is like trying to win the London Marathon – it takes hours and hours of practice and training in the build-up to get even close. Does that mean it’s not worth it? No. The fun is in the taking part – I’d encourage everyone to take part and just give it a go. It’s a fun project and something that only comes around 3 times a year.

What about the other entries, any favourites?

For analysis I have to choose my good friend Rob Radburn’s: Who Care’s. Rob has an instantly recognisable style and his commentary and analysis really shine in this piece.

For Story-Telling I’d say Mike Cisneros’: Last Words is just beautiful. Mike pulls together visualisations that might be just “show me the numbers” but binds them with stories of last letters home which just break the heart.

For Design Curtis Harris’: If I was Kevin Durant wins the day for me, it’s just a beautiful piece of  work, not over reliant on imagery – just all about the data.

There are lots more I could pick out but these are some of my favourites.

Those are all amazing pieces of work, thanks for sharing Chris and good luck with Iron Viz.







What happens at The End?

This is a post is a small thought about one piece of Data Visualisation best-practice.

“Jeez Chris, get a life”. Yes, I know, here I am again. Get over it. I happen to think this stuff is quite important.

This is a reply to and critique of the chart made by Jeff Plattner and recreated by Andy Kriebel in his blog post Makeover Monday: Restricted Dietary Requirements Around the World | Dot Plot Edition

I’d originally fed back on best practice to Jeff on Twitter but given Andy has chosen to recreate the chart and has a huge audience for his blog I felt it was worth a blog post in reply to point out, what I think, is a small best practice failings in the chart.

Let’s compare the three charts below, the last being Jeff’s original:


Bars – Andy’s Initial Makeover


A Lollipop Chart based on Andy’s Makeover of Jeff’s

Dot Plot

Jeff’s Original

I love Jeff’s take on this subject. I immediately fell in love with the design and loved the “slider ” plot which I’d not seen used so effectively.

However there is a subtle difference between this last chart and the two above.

All three have axis that shows the range of the bars / lollipops / sliders at 50%. This is a design choice which both Andy and Jeff (for the first and third charts) both said came from wishing to make the chart look better.


Now here comes the rub for me. In the first two the shrinking of the axis doesn’t take away from the audiences understanding of the chart. However in the last “slider” chart it does. Why? Because the chart has an end.

Why “The End” matters…

The end of a visualisation mark / chart is important for me, because if it exists then it implies something to the reader. It implies that

a. the data has a limit

b. you know where the limit is and can define it

c. you have ended the chart at the same place as the limit of the data

Let’s look at the three aspects here with our data

a. ✔ the limit of the data is 100%

b. ✔  no region can be more than 100% of a given diet

c. ✖  the line ends at 50% in Jeff’s chart

Why doesn’t this matter for the first two charts? Well these two charts don’t have a limit set by the chart. Yes, the bar and lollipops end, but we’re forced to look elsewhere to see the scale. With the “slider” chart, then in my opinion, the reader feels safe to assume that a dot half-way along a line means that half the people in that area follow the diet. They don’t go further to look for the scale – despite the fact Jeff has clearly marked the limits.

This perceptual difference between the charts is important for me, and a good reason not to limit the axis at any other value than 100%, as I have done below by remaking Andy’s Remake. Dot Plot 100%

It this the biggest faux paux in the history of Data Visualisation? On a scale from 0 to 3D Exploding Pie Chart then we’re at 0.05. So no, not really, but I thought it was interesting enough to share my thoughts on what was an excellent Viz by Jeff.

As ever these are only my thoughts, they’re subjective, and many of the experts may not agree. Let me know what you think. Is the visual improvement worth the sacrifice of best practice?

Comment here or tweet me @ChrisLuv

Spring’s Got Speed

I must admit when Iron Viz was announced for this qualifier I had mixed feelings. On a positive side I know nature and animals, as a photographer and bird watcher I’ve spent a long time gathering data on animals. I even run a website dedicated to sightings in my local area (sadly it’s been neglected for years but still gets used occasionally). On the negative side though I knew I would be faced with competition from a bunch of amazing looking visualisations, using striking imagery and engaging stories.

Did I want to compete against that? With precious time for vizzing lately (especially at end of quarter – you can tell the Tableau Public don’t work in the sales team!) I only wanted to participate in Iron Viz if I could be competitive, and for those who don’t know me I like to be competitive….

So, as you perhaps guessed, my competitive edge won and I pulled some late hours and risked the wrath of my family to get something that did the competition justice.

A Note On Visualisations and creating a Brand

I’ve noted above I expected a lot of pictures and text from people in this qualifier, after all Jonni Walker has created his own brand around animal visualisations, stock photography and black backgrounds. However I have my style of visualisation,  I’m not Jonni Walker, what he does is amazing but there’s only place for so many “Jonni Walker” vizzes. I couldn’t replicate what he does if I tried.

In the past I’ve deliberately combined analytics and design, treading the fine line between best practice and metaphor, staying away from photograph and external embellishments and preferring my visualisations to speak for themselves through their colours and data. The subject this time was tricky though…was it possible to produce an animal visualisation without pictures?

Finding a Subject

I could turn section this into a blog post on it’s own! . I trawled the internet for data and subjects over days and days. Some of the potential subjects :

  • Homing Pigeons (did you know their sense of smell affects their direction)
  • Poo (size of animal to size of poo) – this was my boys’ favourite
  • Eggs (had I found this data I’d have been gazumped:
  • Zebra Migration
  • Sightings Data of Butterflies

Literally I couldn’t find any data to do these enough justice, I was also verging on writing scientific papers at points. I was running out of ideas when I found this website: – and I dropped them an email to see if I could use their data. Phenology (studying natures cycles) has always interested me and getting my hands on the data would be fantastic. There was even a tantalising  mention of “measuring the speed of spring” on their news site with some numbers attached but no mention of the methodology….

Now, I’m impatient and so….using a few “dark art” techniques I didn’t wait for a reply and scraped a lot of the data out of their flash map using a combination of tools including Alteryx.

Thankfully a few days later they came back and said I’d be able to use it (after a short discussion) and so all ended well.

Measuring the Speed of Spring

Now I had the data working out how to measure my own “speed of spring” was difficult. Several options presented themselves but all had their drawbacks…the data is crowd-sourced from the public, mainly by people who can be trusted but amateur outliers could affect the result (do you want to say Spring has hit Scotland based on one result). Also the pure number of recorders in the South East, say, could affect any analysis, as could the lack of them in say, Scotland. Given we’re expecting to see Spring move from South to North then that could seriously sway results.

In the end I played between two methods:

  1. A moving average of the centroid of any sightings – and tracking it’s rate of movement
  2. A more complex method involving drawing rings round each sighting and then tracking the overall spread of clusters of sightings across the UK.

In the end I opted for the latter method as the former was really too likely weighted by the numbers of sightings in the south.

Very briefly I’ll outline my methodology built in Alteryx.

  1. Split the country into 10 mile grids and assign sightings to these based on location
  2. Taking each grid position calculate the contribution to the surrounding grids within 50 miles based on a formula: 1-(Distance/50). Where Distance is the distance of the grid from the source grid.
  3. Calculate the overall “heat” (i.e. local and surrounding “adjusted” sightings) in each grid cell
  4. Group together cells based on tiling them into groups dependent on “heat”
  5. Draw polygons based on each set of tiles
  6. Keep the polygon closest to the “average” grouping i.e. ignoring outliers beyond the standard deviation

I then did the above algorithm for each week (assigning all sightings so far this year to the week) and for each event and species.

These polygons are what you see on the first screen in the visualisation and show the spread of the sightings. I picked out the more interesting spreads for the visualisation from the many species and events in the data.

Small Multi Map

The above process was all coded in Alteryx.


If you look closely there’s a blue dot which calls this batch process:

Alteryx Batch

which in turn calls the HeatMap macro. Phew, thank god for Alteryx!

Now to calculate the speed, well rate of change of area if you want to be pedantic! Simple Tableau lookups helped me here as I could export the area from Alteryx and then compare this weeks area to the last. The “overall speed” was then an average of all the weeks (taking artistic licence here but given the overall likely accuracy of the result this approximation was okay in my book).

Iterate, Feedback, Repeat

I won’t go into detail on all the ideas I had with this data for visualisation but the screenshots will show some of what I produced through countless evenings and nights.


“Good vizzes don’t happen by themselves they’re crowdsourced”

I couldn’t have produced this visualisation without the help of many. Special mentions go to:

Rob Radburn for endless Direct Messages, a train journey and lots of ideas and feedback.

Dave Kirk for his feedback in person with Elena Hristzova at the Alteryx event midweek and also for the endless DMs.

Lorna Eden put me on the right path when I was feeling lost on Friday night with a great idea about navigating from the front to the back page (I was going to layer each in sections).

Also everyone else mentioned in the credits in the viz for their messages on Twitter DM and via our internal company chat (I’m so lucky to have a great team of Tableau experts to call on as needed).


Getting things to disappear is a nightmare! Any Actions and Containers need to be in astrological alignment….

Concentrating on one story is hard – it took supreme will and effort to concentrate on just one aspect of this data

Size – I need so much space to tell the story, this viz kept expanding to fit it’s different elements. I hope the size fits most screens.

The Result

Click below to see the result:



Why we’re going to #opendatacamp

screenshot-2017-02-18-11-42-58On Saturday and Sunday fellow Tableau Zen Master, Rob Radburn and I will be attending Open Data Camp in Cardiff.

So why are we spending a Saturday and Sunday in Cardiff away from our families and spending a small fortune on hotels?

Well sometimes data visualisation can be frustrating. We’re both prominent members of the Tableau Community and we’ve spent countless hours producing visualisations for our own projects as well as community initiatives such as Makeover Monday and Iron Viz. There’s lots of fun and rewards for this work, both personally and professionally and so why is it frustrating? Well shouldn’t there be more to data visualisation than just producing a visualisations for consumption on Twitter? How do we do produce something meaningful and useful (long term) through data and visualisations?

Open Data seems a suitable answer however with so many data sets, potential questions and applications it’s hard to know where to start. The open data community have done a great job at securing access to many important datasets but I’ve seen little useful visualisation / applications of open datasets in the UK beyond a few key datasets. How do we do more?

tableau_logo_crop-jpg_resized_460_Tableau Public on the other hand has done a fantastic job of ensuring free access to data visualisation for all, but few in the community have worked with the open data community to enable the delivery of open data through the platform.

Rob and I are hoping that our pitch at Open Data Camp will facilitate a discussion around bridging the gap between the Tableau Community and the Open Data Community. On the one side we have a heap of engaged and talented data viz practitioners on Tableau Public looking for problems, on the other hand a ton of data with people screaming for help understanding it….on the face of it there seems some exciting possibilities, we just need to pick through the .

Oh and while we’re there if anyone wants us to pitch a Tableau Introduction and / or Intro to Data Visualisation we’d be happy to facilitate a discussion around that too.

Would love your thoughts

Chris and Rob

MM Week 44: Scottish Index of Multiple Deprivation

This weeks Makeover Monday (week 44) focuses on the Scottish Index of Multiple Deprivation.


Barcode charts like this can be useful for seeing small patterns in Data but the visualisation has some issues.

What works well

  • It shows all the data in a single view with no clicking / interaction
  • Density of lines shows where most areas lie e.g. Glasgow and North Lanarkshire can quickly be seen as having lots of areas among the most deprived
  • It is simple and eye catching

What does work as well

  • No indication of population in each area
  • Areas tend to blur together
  • It may be overly simple for the audience

In my first attempt to solve these problems I addressed the second problem above using a jitter (using the random() function)


However it still didn’t address the population issue and given the vast majority of points had similar population with a few outliers (see below) I wondered whether to even address the issue.


Then I realised I could perhaps go back to the original and simply expand on it with a box plot (adding a sort for clarity):


Voila, a simple makeover that improves the original and adds meaning and understanding while staying true to the aims of the original. Time for dinner.

Done and dusted…wasn’t I? If I had any sense I would be but I wanted to find out more about the population of each area. Were the more populated areas also the more deprived?

There have been multiple discussions this week on Twitter about people stepping beyond what Makeover Monday is was intended to be about. However there was story to tell here and I dwelled on it over dinner and, with the recent debates about the aims of Makeover Monday (and data visualisation generally), swirling in my head I wondered what I should do.

I wondered about the rights and wrongs of continuing with a more complex visualisation, should finish here and show how simple Makeover Monday can be? Or should I satisfy my natural curiosity and investigate a chart that, while perhaps more complex, might show ways of presenting data that others hadn’t considered….

I had the data bug and I wanted to tell a story even if it meant diving a bit deeper and perhaps breaking the “rules” of Makeover Monday and spending longer on the visualisation. I caved in and went beyond a simple makeover….sorry Andy K.

Perhaps a scatter plot might work best focusing at the median deprivation of a given area (most deprived at the top by reversing the Rank axis):



Meh, it’s simple but hides a lot of the detail. I added each Data Area and it got too messy as a scatter – but how about a Pareto type chart…


So we can see from the running sum of population (ordered by the most deprived areas first) that lots of people live in deprived areas in Glasgow, but we also see the shape of the other lines is lost given so many people live in Glasgow.

So I added a secondary percent of total, not too complex….this is still within the Desktop II course for Tableau.


Now we were getting somewhere. I can see from the shape of the line whether areas have high proportions of more or less deprived people. Time to add some annotation and explanation….as well as focus on the original 15% most deprived as in the original.

Click on the image below to go to the interactive version. This took me around 3 hours to build following some experimenting with commenting and drop lines that took me down blind (but fun) alleys before I wound back to this.



Makeover Monday is good fun, I happened to have a bit more time tonight and I got the data bug. I could have produced the slightly improved visualisation and stuck with it, but that’s not how storytelling goes. We see different angles and viewpoints, constraining myself to too narrow a viewpoint felt like I was ignoring an itch that just needed scratching.

I’m glad I scratched it. I’m happy with my visualisation but I offer the following critique:

What works well:

  • it’s more engaging than the original, while it is more complex I hope the annotations offer enough detail to help draw the viewer in and get them exploring.
  • the purple labels show the user the legend at the same time as describing the data.
  • there is a story for the user to explore as they click, pop-up text adds extra details.
  • it adds context about population within areas.

What doesn’t work well:

  • the user is required to explore with clicks rather than simply scanning the image – a small concession given the improvement in engagement I hope I have made.
  • the visualisation take some understanding, percent of total cumulative population is a hard concept that many of the public simply won’t understand. The audience for this visualisation is therefore slightly more academic than the original. Would I say this is suitable for publishing on the original site? On balance I probably would say it was. The original website is text / table heavy and clearly intended for researchers not the public and therefore the audience can be expected to be willing to take longer to understand the detail.

Comment and critique welcomed and encouraged please.

Makeover Monday Week 43: US National Debt


This weeks Makeover Monday tackles National Debt. Let’s start by looking at the original visualisation.

Apparently the US National Debt is one-third of the global total. Showing these two values in a pie chart is a good idea as it quickly shows the proportions involved. However the pie chart chosen does have a strange white think slice between the two colours and a black crescent / shadow effect on its outside edge which add no real value (in fact the white slice added a bit of confusion for me).

The visualisation then goes on to show $19.5 trillion dollars in proportion to several other (equally meaningless) large figures. The figures do add some perspective on just how big that figure is and the use of $100 billion blocks in the unit chart does allow an easy comparison. One slightly critical feature, if we were to pick holes in the visualisation, is that half-way through the view starts showing the shaded blocks to compare to the 19.5 trillion, whereas before it doesn’t.


with shaded blocks


no shaded blocks

Achieving consistency is important in data visualisation as it lets the reader know what to expect and gives them a consistent view each time to aid comparisons. So making a design decision to add shaded blocks across each comparison would perhaps have been a better choice as opposed to switching half way through.

Visualising Small Data

The dataset provided for the weeks makeover has simply two rows, showing the debt for each area (US and Rest of the World).


Clearly this presents a visualisation challenge. Visualising small datasets is hard, as there are limited choices. One can attempt to include secondary datasets to show the numbers in context, as the original author has done but another, simpler choice, might be to show them relative to each other – similar the original’s pie chart. One might even attempt to show how the data corresponds to the population of the US or the world, attempting to bring the figure down to something manageable (in the US the debt is a more comprehensible $61,000 per head).

Before we attempt to visualise something though we need to think about the audience and message we want to provide. Are we simply trying to show the figures without any comment? or do we want to focus on how large they are? or are we commenting on how large the US debt is to the rest of the world and making a social / political comment?

With a dataset so small any editorial comment is difficult though. For example we have no context on the direction of movement of these figures. The US might be quickly bringing it’s debt under control, while the ROW grows, or the opposite might be true. The ROW figure might be dominated by other developed countries, or might be shared equally. How can we comment without further analysis on temporal change or the context of this figure?

If we can’t comment editorially then we are left with simply showing how huge these numbers are. My criticism of the original is that the number it shows in comparison are equally huge, and equally incomprehensible for a lay person. Given this visualisation is published on a website Visual Capitalist perhaps their audience is more familiar with global oil production or the size of companies but for any visualistion published away from the site a more meaningful figure is needed. Personally I think the amount per head is an especially powerful metaphor. In the US $61,000 dollars each would be required to clear the debt, the ROW world would just have to pay a little over $5.

To Visualise or not to Visualise

Now there is an important decision here, how to effectively show those figures in context. However with such small data is there any point in doing so? Everyone can quickly see $5 is much less than $61,000 – we don’t need a bar chart or bubble to show that, and we certainly don’t need a unit chart or anything even more complex. This is the problem with small datasets, any visual comparison is slightly academic given we can quickly mentally interpret the numbers.

One might be tempted to argue that a data visualisation is needed to engage our audience. Perhaps a beautiful and engaging data visual might do a good job of this, however so would the use of non-data images like the the below.


Defining Data Visualisation

Makeover Monday is a weekly social data project, should a visual that includes only text be included?

What if the pile of dollars in the image above had exactly 61,000 dollar bills would that make it any more of a data visualisation than one that contained a random amount? What if, instead, we added as a unit chart with 12,200 units of $5 bills? These accompanying items don’t help us visualise the difference any better than the text. One could argue where the main purpose of a visualisation isn’t to inform or add meaning or context, and is instead used as a way of engaging the user, then it becomes no different to any other image used in this way. Therefore adding any more data related visualisations to the above text wouldn’t make the image any more of a data visualisation than the one above.

Semantic arguments that attempt to define data visualisation are interesting but academic. Ultimately each project that uses data does so because it needs to inform its audience, and it is the success of the transaction from author to audience that deems how successful the project is.

So should we define a data visualisation as more (or less) successful because of its accompanying “window decoration” (or lack thereof)? In my opinion yes. Accompanying visuals and text help provide information to the audience and can help speed up the transfer of information by giving visual and textual clues.

Do charts / visuals that make no attempt (or poor attempts) to inform the audience add any more value to a data visualisation project simply because they use data? In my opinion, no. This isn’t the same same thing as saying they have no value but simply producing a beautiful unit chart, say. with the data for this Makeover Monday project would add no intrinsic extra value in educating the audience and therefore would be no more valuable than any other picture or image.

Is the above image a successful Data Visualisation? Let’s wait and see on that one. I’m intrigued to see what the community makes of a purely text based “visualisation”.

Does it do a better job at informing the audience than the original? Again this is hard to answer but I believe I understand more about the size of the debt when it is visualised in terms of dollars per head. By bringing these numbers down to values I understand I did’t need to add any more visualisation elements in the same way as the original author, therefore you might say mine is more successful because it manages to pass across information in a simpler, more succinct transaction.