Bella consults

Time only runs top down in sandglasses. But not in charts. There, time is from left to right. Because it’s like that. Take a chart with a horizontal x-axis and a vertical y-axis. The independent variable should always be on the x-axis. Then everybody knows: The dependent variable is on the y-axis. That’s helpful. Like most other conventions, too.

In the newspaper WAMS, the time is vertical. Not too bad. But also not very nice.

Source: Welt am Sonntag, Nr. 40, 04/10/2009, p. 36. Title: 26 percent minus. The German export to the Iran in billions of Euro, January to July for each year

Oh, ZEIT. I didn´t mean that. My scribbled design was born out of necessity. Yours is intentional.

Scribbling threefold from Die ZEIT. Not supposed to take it seriously?
Source: DIE ZEIT No. 19 from 2009–04–30, No. 20 from 2009–05–07, No. 25 from 2009–06–10.

Funny: We use a computer to make it look handmade. Even the handwriting is a computer font.

Handwriting and painting signal: Love letter. Poetry album. Children’s book. First-grader. Shopping list. Draft. Sketch. Holiday greetings. Note. So rather: Temporary. Hasty. Unfinished. Casual. Raw. Imprecise. Maybe faulty. Unchecked. Unique. Spontaneous. And young.

All want young. I don’t care. Data neither.

We are on vacation in France. In France they play Boule. With the neighbor’s dog I bet on the outcome. I watched five throws from each of them and measured.

I heeded his advice. And sketched the entire distribution. Not only the mean values. This is how close the boules landed to the Cochon. In centimeter.

With the naked eye I can see now:

Outlier
He made the very best throw of all. One swallow? A summer?

Happy made some respectable throws. But also the worst one. Could be a one-time outlier. Or a hint that his performance is not solid.

Averages
Hannes and Bimpfi are close to each other. So are he and Baui. Hm. That’s still not what I was looking for.

Spread
Bauis throws spread less than his do. And the averages differ only by 1 centimeter. Therefore Baui comes in third place, instead of fourth.

Bimpfi’s performance is solid, with a moderate spread. Hannes beats them all: His average and spread are the lowest.

I bet on Hannes. He won.

He says: If you have the data, show them. Instead of whining about “crash” and “upswing” and “slump” and “boom”. Otherwise, someone believes it is 70 % and someone else believes something else. So, the controllers may not copycat the newspapers. He says, they don’t have to: They have their data.

Matt says the same. Roughly. He has shown us a graphic. Sherman Kent drew it. 40 years ago. A secret agent from America. NATO officers were to estimate how much percent Sherman means when he says: something is “highly likely” or “probably” or “unlikely” or “highly unlikely” and so on.

Source: Richards J. Heuer, Jr.: Psychology of Intelligence Analysis, Langley 1999, p. 155.

Ah. Quite dangerous. After all, this was about military stuff. If the CIA tells the President, the bad guys are coming “probably”. And the President understands, ok, well, maybe with a probability of 25 %. But the CIA had 75 % in mind: Phew.

Sherman said, that´s why there have to be standards for words in the reports of intelligence agencies:

Maybe. Anyway: Rules are good. Whining is bad.

Tax revenues collapse. Tells us the Handelsblatt. He thinks, 9 percent minus and collapse don’t go together. I read the article, too. And nearly collapsed: Perceptive priority not understood. Once again: With columns you interpret differences in length as differences in value. Therefore: Chopping prohibited. Or: Use lines instead of columns.

The Handelsblatt didn’t. See left. But they should. See right.

“Beyond the zenith”: German tax revenues in billion Euro. Source: Handelsblatt no. 131, July 13th 2009, p. 3. Redesign: me. – The Handelsblatt is a major German economy newspaper.

Now I became a little skeptical and checked the reference (PDF). There are two more estimates. Hm.

Source: me.

Now I’m pinned to the office wall. I like that. My own small multiple.

Apropos of resolution: General Electric has got a web page with health data. You can play with the data. 300 stickmen symbolize America. They are in every picture.

Depending on “Diseases & Conditions” and “Risk Factors” they stand in different formations. Well, sometimes they just disappear. Here 120 stickmen have left (or 40%).

For a stroke it looks like this. The thin ones on the left and the fat ones on the right.

Maximum differentiation is 0.33 % (1/300) if 300 stickmen symbolize America as a whole. Within the segments differentiation drops. For the thin ones it is 4.8 % (1/21). But now listen: 1 % of the thin ones have a stroke. That’s one-fifth stickman (1 % of 21). But GE doesn’t paint this. Underneath the picture: The more fat, the more stroke. I can’t see that either. Everything is 2 %. Except the thin ones (1%).

Play with it yourself. I believe: Painting absolute frequencies while talking about percentages goes wrong.

I don’t care about all that: I am a Labrador. I wouldn’t mind being fat.

I’m in France. The water bill has a chart. The chart is all right. Consumption is displayed compared to the years before. The EU wants it that way since May 2008 (Directive 2006/32/EG, PDF). Already in 2006 the EU said: The energy companies have to bill in an “informative way”. The real prices have to be displayed. And the real energy consumption. Also “Comparisons of the final customer’s current energy consumption with consumption for the same period in the previous year, preferably in graphic form” are needed.

Source: mine

So what’s the conclusion for a European dog?

1. If you like your customers you give them the possibility to compare by themselves. Before the EU tells you to do so. Customers like that. Even when they recognize that they have consumed more.
2. In times of crisis the government feels strong. It will dictate many regulations: who shows whom what to do and how to do it. I still don’t believe the state is a better manager. Nor that the state is a good graphic designer. So, a smart dog takes matters into her own hands.

One of my rules: Graphs must be good to read. Good means fast and clear. That applies to all measuring instruments. For watches, too.

Classical watches are beautiful because they are mechanical. They show all values with analog scales. Not good for readability. But nothing else worked. So far. This watch breaks the rules. The stop function looks digital but works with a mechanic drive. That’s new and difficult to produce. Grand complication. You need 800 individual parts and 4 barrels to build it.

The Porsche Design INDICATOR (P’6910) with a mechanical digital stop function display –
click for movie

The stop function is much better readable. Better than with individual indicators for hours, minutes, seconds and tenths of seconds. I see: ANALOG WAS NOT THE CONSEQUENCE OF MISSING READABILITY BUT A CONSEQUENCE OF THE MECHANICS.

However, I like clocks with analog indicators. But not in controlling. There, indicators remain old and stupid.

Apropos competent with media. Leo from the Süddeutsche Zeitung (SZ) got width, height and position. But that doesn’t mean anything. The 100 % value can be at the top middle, bottom middle or in the middle right. No meaning.

Source: Süddeutsche Zeitung, 2009–04–29, p. 31. The red dots are labeled with the names of the most important shareholdings of BayernLB (Bank of the state of Bavaria) and with the percentage held by BayernLB. The article suspects that some of them may be sold. Click to enlarge.

Same issue: The annual average temperature is rising, I thought. Not the temperature is rising but each year. Not cold at the bottom but 1901. At the top it’s pretty warm but above all 2008. The temperature is next to the year. But no clue what it was like in other years. This chart stinks.

Source: Süddeutsche Zeitung, 2009–04–29, p. 6. On the vertical scale: the 10 warmest years in Germany since 1901, the long-time average is 8.2 degree Celsius. Click to enlarge.

That makes me sad. The journalists of the Süddeutsche Zeitung do a good job. But who believes in good research with such bad drawings?

Frustrated, I turned to watch television. No help either. The TV guide, “TV Today”, was from last week. The charts were as bad as in Süddeutsche Zeitung. But not even worse. Oh my goodness.

Source: TV Today, 2009–04–11, p. 4. TV consumption in 2007 and 2008 in minutes on easter holidays (green) and average for the whole year (red).

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.

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.

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

Microsoft

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.

The magazine CHIP from Munich reports:

“Believe it or not, most discount stores put beer and diapers fairly close to each other.”

That’s due to Data Mining. Because: Young men come. Sent by young women. They forgot to buy diapers. The men find that inane and take beer with them. Wal Mart has known that for a long time. And put the beer close to the diapers. And made a fortune.

I don’t get it. Beer and diapers used to sell together. Before they stood together. Why should they be together now? Earning money only with beer and diapers? Why not beer on every shelf? What do women do now? How do they find diapers?

He says (and he has won again because of Data Mining) that it’s all a lie, CHIP is hallucinating. So I explored a discount store in Nuremberg.

In this discount store (Aldi), beer and diapers are placed together. Because of Data Mining?

Aha. Believe it or not, beer and diapers are located just opposite each other. Bicycles and toothbrushes, too. Then I went to the next discount store. There it is 23 meters from diapers to beer. I wanted to take a photo there as well. But I was thrown out.

Then I asked him. He says: Data Mining is about correlations in shopping baskets. But that doesn’t say anything about placement. Some things bought together are better placed far away from each other. People then pass by lots of things that they might also buy.

That sounds good: a bicycle for example.

Good news: More people are reading Bella, fewer people are reading Germany’s boulevard newspaper Bild. The German magazine Der Spiegel seems happy, too. So much so that –24 % is steeper than –31 %. I’ve drawn how it should look (click to enlarge)

Der Spiegel, No. 10, 2009–03–02, p. 87

I like the German journal Auto Motor Sport. First I liked the chart from Thursday (6/2009, p. 130), too. Then it felt strange: arrows for security at Mercedes and quality at Opel indicate nearly identical relative changes (-26 % and –25%). But they are so different. I painted them correctly in orange. I used 45° degrees for a change of 100 %.

I fear for the designer. He won’t get a job with Daimler. Although with Opel. But they don’t have any. And Toyota’s Ninjas are on his heels.

In case he survives: arrows are reserved for trends. What was in 2006, 2007, 2008? And: all arrows seem to start from the same level. But that is not true. I offer 5 $ compensation for this wreck.