Forbidden to forbid

Tuesday, December 15th, 2009

Today it’s getting didactic. I did some cutting.

Source: Handelsblatt, April 30th 2009, no. 83, p. 1, redesign by me. Click for original. Graphic displays increase of highway transportation in billions of ton-kilometers (1997 to 2008).

Many will say: left is correctly scaled. They believe: Null value must be shown. Regardless of line or column – never cut.

But: The null value is not sea level. On the right the full space is utilized to display changes. That’s important, often. All other scales more or less steepen the slope. They show more or less details. That the variation is bigger than 50 %: neither visible on the right nor the left. Better write that down.

Anyway: To cut lines’ feet is not forbidden. Snip snap.

Competent with media

Thursday, April 30th, 2009

It’s the school’s fault. They say so. But not here: most information designers create junk charts when using areas. In 7th grade kids learn how it is right and wrong. It makes them competent with media.

Fl├Ąchengrafiken in der 7. Klasse
Lambacher Schweizer: Mathematik für Gymnasien 7, p. 122. (click for translation)

Teachers have their book with solutions. On page 81 it says:

left: essentially a column chart – one car symbol represents 10,000 cars; reasonable
right: length and height increase by 66 %; perception is distorted, because the area is doubled

Now, I wish more competent media to all kids competent with media.

Listen to the pattern II

Wednesday, April 15th, 2009

Johannes Kreidler gives time series a sound. I did that two years ago. In fact, I copied it from him*.

I guess Johannes wants to make fun of the crisis. Businesses take this very seriously. Your human ear is a lot faster than your eye. The eyes see what the ears have already heard. Such as a ranking with one high, some middle and a lot of small values. A very characteristic sound pattern. Possibly irritating. But that is good. You think about it.

This is the sound of some of the data from his video without fun:

Lehman Brothers

General Motors


My old sparklines with new data:

Click on the speaker symbol to hear the crisis.

* Source: Bissantz, Nicolas: Innovative Produkte: DeltaMiner. In: WIRTSCHAFTSINFORMATIK 43 (2001) 1, pp. 77–80.

Double spam

Thursday, January 15th, 2009

At November 18th I received this email. Unasked. They cannot blame Excel’s diagram wizard for that. It likes to chop axes. But not always. And not with this data.

Flashy? Not with the Journal

Sunday, November 30th, 2008

Obama times: Yes we draw. Bars as long as they actually are. No problem at all in Obama country:

The Wall Street Journal draws them over the whole page (click for full view).

In the German newspaper Die Welt, the chart has been struck by lightning (left: original, right: me). The flash makes the chart absurd. A table would be better. On the right I drew how it should be.

Wall Street Journal, 2006–11–21, p. C4; Die Welt, 2008–08–13, p. 26.

A picture triggers more than 1000 questions, not?

Friday, May 30th, 2008

From the chart I read that from 1999 till last year the number of passangers increased to sparkline_passagiere.png 2.2 bn. During the same period there have been lately per 1 million flights  0.75 total write-offs for airplanes. The German journal “Der SPIEGEL” sticks to rule 13 vor demolishing charts: Embellish whatever you want to say with numbers that tell a different story!

In the text SPIEGEL says: For the first time since 1998 (yes!) there are more total write-offs than the year before. “Flying is secure but starts to become insecure is the statistic’s story. And some flight experts see a dangerous new trend”. I never fly. I am only interested in graphs. That should look like that:

The difference between national/international is ignored by the author, so we don’t need it. Scale and guiding lines don’t help either. Values all the more. AND INCLUDE THE MESSAGE. Not bad.

(Chart: SPIEGEL 22/2008, p. 147, Redesign: ME)

Good old Times

Thursday, May 15th, 2008

Many cut axes just because of Excel. And go to charting hell. Because when you cut you distort. And when you distort you lie. At least with your graph. The graphical change in your data is no longer proportional to the change in values. He explains it quite well with a demolished graph from the SportAuto magazine. Here is a positive example. From the German newspaper “Die ZEIT”. This newspaper is off the mark sometimes, too.

Strike at the German Post, information on market share and letter volume
Die ZEIT, 2008–04–30, p. 37, market share of Deutsche Post (left) and number of letters in Germany in billions (right)

All my rules are observed. Time runs from left to right. Scale starts at zero. The graph is proportional to its values. No exaggeration. No gadgets. No unnecessary percentage signs. Letters in billions, not in single pieces.

In the same issue another good graph. Structure is shown top-down. Labeling where it belongs: next to the columns. At least for the values. I would have left out the series “andere” (“others”). The dots, too.

The demand for academics rises, data on the automotive sector in Germany
Die ZEIT, 2008–04–30, p. 78, academics in the automotive sector, for automobile manufacturers and suppliers

Bella Reporting Standards

Wednesday, April 30th, 2008

The day before yesterday Rolf and I raged again. We took apart annual reports of large corporate groups. We made fun of tachometers. We showed how charts lie. We established rules. We defined standards. Some of the rules are in the example. Time runs from left to right. Only structure is shown top-down. We don’t use funny patterns. We label directly. We never label twice. We avoid legends and scales.

Rules for charts

Charts show profits (Gewinne) of TUI AG, a major German company for tourism (Touristik, red), shipping (Schifffahrt, blue), and logistics.

“You can’t have the pie and eat it, too”

Sunday, March 30th, 2008

Appetite always signals which piece of the pie is biggest. Always? Sort the pieces. How long did it take?


Sort the bars. How long did it take? Values are the same in both cases.


Eric has the same example. But while scaling the bars he was tricked by Excel.

Guess what the circular area represents. And?

Circular area as measure

Its even worse in Excel. You can use the diameter to show the value.

Diameter as measure

It’s enough to drive a person mad.

Pimpcharts in his beloved SportAuto newspaper

Saturday, March 15th, 2008

His day was ruined already. His beloved SportAuto newspaper had graphically demolished 500 yards of safety fence. I wanted to comfort him and skimmed through the paper. Issue seven 2006. Insiders know. Ooh…

SportAuto Manual correction of the chart

Page 102 in the same issue: Formula One lap times from two seasons. Smaller engines yet faster again. Very interesting comparison. Good data. Good legend. Beautiful Evidence.

But the chart: a total loss. Reason: Gross negligence. On the right side: That’s how it should have looked like. The two Grand Prix which are not comparable (rain in Australia, new race track in Imola) are left out. During the race they have been faster twice and slower three times. Just the opposite for the training.

Correct display

In the last two issues of SportAuto: not one chart at all. Good.

Good luck is not enough to understand business charts

Monday, December 31st, 2007

Business charts are losers. Its only good luck if they work at all. He has shown some basic problems of them. With mine here, you need more brains than luck to find the truth.

Three time series

Each series increases with the same amount per month. The first increases by 5, the second by 25 and the third by 30. All increase with the same relative dynamic. From January to February by 50 %, next months by 33 %, then 25, 20, 17, 14, 13, 11, 10, 9, 8 %.

The eye is sure to see something different. The increase of the bottom most series seems dull compared to the top most. And even a difference in dynamics between the two top most series appears to be plausible.

All wrong.

Sparklines are much better suited. They present the first series as series 1 65, the second asseries 2 325 and the third as series 3 390. More general: Want to compare different time series? Then scale them individually. It doesn’t work in one single chart.

Why not Mosaic Displays!

Wednesday, May 30th, 2007

Worse than radar charts? Mosaic displays!

Table mosaic displays - raw data

Data on hair and eye color of 592 students is transformed into this:

Mosaic display

The relative frequency is displayed as an area. However, the eye isn’t good in comparing areas of different sizes.

All three areas are of the same size: Comparison of areas

The labels are hard to read, too. You have to count. To display the frequency a simple graphic table is enough:

Graphic table as an alternative to the mosaic display

Stop! Is green greener than yellow?

Friday, March 30th, 2007

There were nice charts in the German journal “Bild der Wissenschaft” 3/2006. They showed risk maps for Germany. Sadly, they weren’t readable. A little trick and they would have been brilliant. A color gradient from dark red via yellow to dark green doesn’t represent a natural order. Neither for the human eye nor for mine. Green is no better, larger or warmer than yellow and yellow no better than red. If colors with identical intensity are used for lowest and highest values you cannot identify patterns.

Earthquake and winterstorm risk Germany

Left: Risk of earth quakes in Germany, right: risk of winter storms, source: CEDIM Risk Explorer

Cognition of colors has to be proportional to displayed values. It’s best with a gray scale. If color is required, different hues of the same color are easily distinguished by the eye. For differentiating positive and negative values a combination of two colors is o.k.

Traffic light vs. color hue

Traffic light colors vs. color hues – geo analysis example from DeltaMaster

What a pity that most designers of weather charts don’t know that, either. Zero degree Celsius is very blue. Plus one degree Celsius is only a little less blue. But never yellow.

weather map

Example of a weather chart

Rich tables for high resolution reporting

Thursday, February 15th, 2007

What is better: tables or charts? That was discussed in the past. Some said a chart is worth a thousand words. I think a picture is worth a thousand words. And most charts need a thousand words to be explained. They are so awful, mostly. Because it is so hard to standardize a chart and there are lots of things that can go wrong. Bissantz loves graphic tables. Me too. Charts are tightly integrated into tables. You have the best of both worlds in one concept. Graphic tables are easy to standardize and you can integrate almost every chart type in a table. Sparklines provide time-series. Horizontal bars are good for almost all types of comparisons. Graphic tables maximize data-to-ink ratio. They are fine grained and with high resolution. The graphical elements provide easy orientation for your eyes. The numbers provide all the details. Typical problems of legends and labeling never arise. I provided some examples in the “Radar chart trap” and in “Small things that make a big difference”. Today, I explain how to create waterfall charts in tables.

A waterfall chart is a variant of a bar chart. It shows how an initial value is changed through other values which lead to a final value. In Excel, you can use invisible columns but this is tricky. Rolf has a nice Excel template for free. Data entry is not so easy. Formatting is very nice.

GM scheme Hichert

Jon has programmed an Excel add-in. Formatting is not so easy. Data entry is very nice.

GM scheme Peltier

In both cases a chart is the result. If you turn it around it is still readable, perhaps even better.

GM scheme turned

Turned around it fits perfectly into a graphic table. Subtotals are also easy. This is how Bissantz has done it in DeltaMaster.

GM scheme in Pivot

It is possible in Excel, too. Just use the REPT function and a few tricks. I show you how in an example Excel sheet.

DB-Scheme inExcel

Radar chart trap

Wednesday, December 6th, 2006

In our company, people think a lot about how to present figures well. ‘Well’ means that everybody is able to quickly grasp what they mean. That gives us more time to brood over their impact.

Things become difficult when it comes to radar charts.

Radar chart

The same data is much better off in a graphical table. The most important fact emerges immediately: I am always above average!

Graphical table

Here at the office, they make software that draws these tables automatically. In Excel you would use the Repeat function. That’s quite simple: a vertical bar is repeated as many times as the corresponding data value prescribes. That yields a nice bar inside the cell. You can even label it if you concatenate the value with the “&” character. Here is a sample how that works.