Martynas Jočys

How pie charts encode data.

How bar charts encode data
#dataviz

Image: visual encodings for numeric data - Position, Aligned length, Angle, Slope.

The image shows the best ways to encode numeric data in a chart.
The most obvious charts using these encodings are:
- Bar chart (uses Length)
- Dot chart and Scatterplot (use Position)
- Line chart and Slope chart (use Angle and Slope)

There are plenty of resources on how to choose the right chart for your data visualisation. I simplified all of it for business intelligence people creating dashboards, who need to make fast and efficient decisions.

We need encodings, I'm afraid. chartplanet.net/choose-chart...

I believe this is the best advice in data visualization and dashboard design.

– Wait, but what if they think they know what they are doing, but actually, they don't?

Humans are not bad at reading angles.
They are not terrific, though.

I'm not saying pie charts are good. I'm saying this usual argument that humans are bad at reading angles so you have to avoid using pie charts is unrelated to pie charts.

If you say that pie charts are OK – beginners will abuse them in outrageous ways.

We can agree that it is safe to never ever use a pie chart. Same with Comic Sans, dashboards with dark background, travelling to other countries, trekking in mountains, diving, skiing, and driving a car.

If pie charts are BAD, we have to find alternatives.
Are Treemaps and Pareto charts better?

Every author writing about dataviz has to have an article about pie charts.
Here is mine with a twist – I analyse how confirmation bias influences suboptimal advice in dataviz, and why, occasionally, I give such advice myself.

chartplanet.net/pie-charts-c...

My session about spaceships in Power BI are back to Power BI and Fabric Summit 2026!
Mostly spaceships, but a few bits on bookmarks are also included. Actually, all you need to know about bookmarks will be there.
globalpowerbisummit.com

How every complaint about Data Quality sounds like:
"The dashboard is broken!"
#data #dashboard

I have finished reading Data Management Body of Knowledge book.
This immediately makes me a Data Governance expert.
Data Governance now is my second specialization after Data Visualization, and will be the first one after AI takes my job as Power BI Consultant.
How do you prepare for AI takeover?

3 things data analysts will do in the age of AI:
1) Set up the infrastructure for AI to work properly.
2) Orchestrate AI agents.
3) Check if AI output makes sense. (Like in the image below)
Data analysts never analysed data and never will.

How do we interact with data in business?

1) We ask questions.
2) We do ad-hoc analysis.
3) We look at ready-made data visualizations.

Which of these AI will disrupt first?
Which one will stay there forever?

AI will kill dashboards
the same way how dashboards, big data, data science, business intelligence killed Excel.

Since I write about data, I have to write something about AI. If I wait for too long, all my predictions might just come true too soon because everything is evolving so rapidly.
So, Will AI take over the Dashboards? Read how it will happen before it happens.
chartplanet.net/will-ai-take...

Enough with the quotes. The next post will be a full article!

Every pixel of colour in data visualization carries information.
Use colour wisely and carefully. #dataviz

So I wrote a nice sentence in my article and decided that it will become a quote worth posting on its own.

If you start creating a dashboard with grey, or toned down pastel colours, it becomes so easy to add highlighting later.
You won't need to invent a colour that shines more than the shiniest one and cuts through the noise.

If you just care about the contrast – at the same time you are solving plenty of accessibility issues.
#dataviz #accessibility

Some people will try to convince you that black text on white background is minimalism, and minimalism is good for dashboards, but I'm not convinced.
It is as bad as another extreme – text subtly fading into the background, which is barely readable.

#dataviz #dashboard

I've seen lecturers spending a full hour trying to convince the audience to spend more time on #dashboard #accessibility. And at the end there are comments like: "I have no time for this".

Let me sneak in one accessibility tip you can't resist: chartplanet.net/accessibilit...

If you work with Power BI and still don't have an ultra-wide screen – make yourself a gift this Christmas. Imagine all your panes fitting, and there is still space for the report view.

Two monitors just ain't good for Power BI.

#powerbi

Make use of a colour wheel when selecting additional colour for your dashboard. Here are a few methods to consider when you already have the basic colour and need additional:

Don't settle with the most basic colours (like basic "blue" or #0000FF). Drifting a bit off those will be a good improvement. If you have any sliders to adjust it in your software – use them.
#dataviz #dashboard

There are two cases for using colours in #dashboards:
– Encode categories - because colours immediately make a clear distinction
– Guide people's attention and highlight interesting data points - because #colour grabs attention and stands out.

#dataviz

You cannot just trust your eyes, they say what is bad, but they do not say what to do. #colour #color

The art of colour choices for dashboard – there is no magic, just colour theory - chart planet For those coming to Data Visualization field from Data Science or Computer Science fields, all colour related stuff might seem like magic – either you have an eye for it or you don't. And if not, then...

I'm fascinated about the colours, and I already know too much about them. This significantly slows down my dashboard creation process because I tweak HEX codes to get that perfect shade every time. If you want to learn just a bit about colours, check here:
chartplanet.net/colour-for-d...

I'm not saying that I'm not a fan of minimalism. But variety helps navigation. Use variety to your advantage. #dataviz #dashboard