Napoleon’s Moscow march never seems to end

Well I’ve managed to hold off for 53 posts but finally I’m compelled to discuss Minard’s 1869 graph depicting Napoleon’s 1812 march on Moscow, a graph famously described by Edward Tufte as possibly being “the best statistical graphic ever drawn“.

It has come up now because Andrew Arbela on The Extreme Presentation Method blog has published a post presenting excerpts from an essay written by Gene Zelazny, McKinsey Director of Visual Communications and author of Say It With Charts, amongst other titles. The purpose of this essay is to present an argument in favour of Microsoft PowerPoint in the face of common negativity that exists towards it (I’ve talked about this increasingly popular and hysterical attitude too).

Zelazny decides to frame his argument about the sophistication of PowerPoint by taking on Tufte (famously anti-PowerPoint) and his belief about the effectiveness of this graphic, in turn offering alternative designs. The basis of his disagreement with Tufte, on both fronts it seems, is a belief that the Minard graphic is difficult to read and is ineffective when displayed in a presentation:

I’ll grant that this is indeed a highly informational chart, especially considering that, when it’s published in a book, you can take as much time as you need to study it (with the help of a magnifying glass in this case)… however, it doesn’t work anywhere as effectively in a presentation… In this situation, the presenter must describe the chart one step at a time so everyone in the audience focuses on the same point at the same time. Clearly, the chart’s complexity makes this a challenge. Secondly,… the chart must be legible to each and every member of the audience. Clearly it’s not.

Unfortunately, I think Zelazny really misses the key argument lines in both his critique of the Minard graph and also in his defence of PowerPoint.

Whilst I don’t necessarily fall under its charm in the same way many do, you can’t help but marvel at the then sophistication of Minard’s solution to successfully display such a complex story into a single graphic containing a total of six variables. And the key word is ‘then’ because it was created in the 1860s with primitive tools, designed specifically for the viewer in printed format and at a time when using statistical graphics to present such information was an absolute rarity. To compare its effect on a PowerPoint slide is absurd. Its perhaps a rather tenuous (but topical) comparison but you wouldn’t challenge judgements about Pele by asking him to prove himself and turn out for the current Brazilian side at the age of 69, would you?

The problems with this essay continue as Zelazny attempts to reinforce his views by offering a series of “additional options to tell the same story [which] rely on the sophistication of PowerPoint as a production tool”.

Let me concede that I have never attempted a reworking of the Minard graphic nor have got an immediate solution, so I’m not speaking from a position of strength here. However, whilst I don’t mind the first attempt - perhaps removing the clipart type images would be advisable – the other two solutions are totally inferior. They completely fail to enhance the communication of the story’s complexity, they fail to create a more cohesive or elegant visual display than Minard achieved and these examples simply do not demonstrate the purpose, utility and value of PowerPoint.

I hate to criticise individuals but, given his prominent visual communications position in such an influential organisation like McKinsey’s, it is disheartening to see that Zelazny himself is credited with creating the second image. Compounding the failure, in my mind, of this essay in the context of its intent Zelazny leaves us with this rather unfortunate parting line:

Contrasting any of these with the original chart, I leave you to decide on the value of PowerPoint as a tool for visualizing the march.

6 Comments

Alex KerinJune 22nd, 2010 at 11:28 pm

The problem with Gene Zelazny’s interpretation is that Powerpoint alone can never be used to replace what was a descriptive printed document. The key to presenting (rather than printing) the Minard document will be the skills of the presenter.

Take the original print, animate portions of it, use some paintings made of the march (http://thephoenixsun.com/wp-content/uploads/2009/06/napoleons_retreat_from_moscow.jpg), and (most importantly) add an enthusiastic presenter and suddenly you have an engaging piece of work.

I don’t like the suggested replacement slides at all – clip art, as mentioned, dreadful, color choices, 3d-effects and so on.

You can’t suggest a Powerpoint replacement of Minard (or Tufte’s Columbia example) with just images – you need to show the audience how the information should be presented, through the addition of an audio track, or even better video showing the presenter’s body language.

If you just show the slide images, then you may as well just use bullets….

Andy KirkJune 23rd, 2010 at 8:48 am

Thanks for your comment Alex, agree with everything you say – you’re first paragraph encapsulates the issue perfectly.

Steve KassJune 23rd, 2010 at 6:57 pm

Not to mention that Zelazny got nearly all the numerical facts (such as five out of the six date-temperature data points) wrong, and that the graphics and numbers are inconsistent (look at the bottom row of army size values, for example). I’ve written about this here.

Michael MacAskillJune 23rd, 2010 at 9:38 pm

I thought at first that those PowerPoint slides were parodies. They seem to be exactly the sort of examples people use when they deride PowerPoint, not defend it.

Apart from being ugly and ineffective, they betray the creator’s lack of understanding of actual data communication. In the first slide, the number of dead soldiers decreases at the 1500 km mark! Was there a sudden influx of the zombie undead at that point? Minard’s graphic makes it perfectly clear what happened there, with the merging of several forces.

The creator of this graphic, meanwhile, has completely confused absolute and proportional quantities, which makes his figure meaningless. That, and the loss of informational dimensions (as opposed to simultaneous display of location as well as troop numbers in Minard’s graphic), complicates rather than simplifies the account. That is, he requires multiple figures to try to make the same points as Minard’s figure, but fails, as this is inherently a multidimensional account.

True, one can’t display enough detail of Minard’s graphic on a 1024×768 pixel PowerPoint slide. So make use of that medium’s advantages (colour, interactivity, multimedia, etc), but also give the audience a one page handout of the graphic, making use of the much higher resolution and information density of paper. That’s Tufte’s point: paper is still superior to a video screen for some purposes. That’s not restricted to a complex graphic like this: try to display a useful table with more than a few rows on screen. Again, supplement PowerPoint with paper when required.

This doesn’t have to be a religious war. Just use each medium for its strengths.

Andy KirkJune 24th, 2010 at 1:13 pm

Steve/Michael many thanks for your respective comments, unearthing inaccuracies in the redesigns and generally helpfully supplementing the initial observations I made above.

Andy CotgreaveJuly 8th, 2010 at 8:39 am

I’ve not come across Zelazny before, but it appears he’s not moved on from the web techniques that were pioneered in 1995. His slides, his website and his books all look very outdated and are far behind the current thinking, as far as I can tell.