[echohack]

I want to call particular attention to the author's point about Apple's pie chart.

Not only is Apple's pie chart used for telling lies here, I would argue that all pie charts are used for telling lies.

The only thing a pie chart gives you is it tells you that everything in a category adds up to 100%. It doesn't tell you what 100% is, or why 100% is good, or bad, or let you compare across charts, or set goals.

You can accomplish this same thing with a bullet graph, developed by Stephen Few. You make a bar chart, then put a line through it to represent your goal (or 100%). Then, it's simple to compare multiple categories and model very complex information in a simple to understand way.

So remember this: The next time you see a pie chart, know that someone is lying or trying to appear more impressive than they actually are.

The worst part about this particular way of graphing information is that it's so ingrained in our culture (it's taught in Kindergarden), that people don't even understand that they're lying, even when they are.

[cstrat]

I can understand how pie charts can be manipulated to show whatever the presenter wants to show...

However I fail to see how they are lying because they used a pie chart... back in 2008 they are saying that Apple had roughly 20% of the market share of smart phones.

They are vague in explaining whether they are measuring - devices sold, devices in use, devices pre-ordered... however that would be the case whether they used a bar graph, line graph, or just gave the raw numbers.

[echohack]

Here's an experiment: Get a random set of pie charts without labels on them. Get a large group of people. Now get those people to go through the pie charts and add what % of the pie chart they think each slice represents.

Now, get a set of bar charts and do the same thing.

In both experiments time the participants. You'll need two groups, a group that knows they're being timed and one that doesn't know.

What group was more accurate? What group was able to complete the exercise faster?

Here's a side by side of the same information in these two different forms: http://speakingppt.com/wp-content/uploads/2012/10/stephen-fe...

The reason that using pie charts is, in and of itself, a dark pattern is because the results of this experiment support bar charts as a more concise, more easily understood way of communicating information. In other words, the only real reason you'd use a pie chart is to obfuscate information. Or, well... to lie. If you're just using it because you don't know better, then you're negligent and misleading people without even knowing it.

[pests]

"then you're negligent and misleading people without even knowing it."

Is that their fault? I don't entirely think so. You might be able to claim they should have done the research before choosing that type of graph but the majority of people are not even going to give pie charts a second thought.

If there was an active campaign against pie charts and more discussion I might find fault in that persons ignorance to the matter.

[gwright]

Yes, pie charts are poor visualization tools. But most people who utilize them are simply unaware of their shortcomings and so they aren't 'lying' (a lie requires intent).

The misuse of language is another way to mislead and confuse people -- intential or not.

We should strive for accuracy in words and charts!

[hayksaakian]

OTOH you can shrink a pie chart without losing information.

Although I agree they are generally bad

http://www.stevefenton.co.uk/Content/Pie-Charts-Are-Bad/