Five tips to make your visualisation more effective

Overview

As we have seen in my previous blog, visualisations are powerful tool in the business world, but they must be used with care as this can lead to misinterpretation of the results being visualised. In this blog, I will give you five tips that will improve your visualisations and make them more effective to the reader.

The examples used are for general cases and are my own personal preferences, and the visualisations do not always need to be exactly like them.

As always, please note that the data used to generate the visualisations in this blog are either fake or open-source data.

In the example that will be used in this blog, we’ll visualise the goals scored by the top 7 football scorers of all time. The data consist of the total number of goals, also broken down as goals scored at club and national levels. The data consists of the goal to game ratio (Total Goals Scored/Total Games Played) as well. This is basically the analysis of how good an attack-minded footballer is/was.

Do not visualise too many things on a single plot. This can be overwhelming to the reader.

Imagine sending this plot to your manager for review, and it is to be included in a client’s report:

This plot is bad at many levels. First of all, there are too many things being visualised. There are 2 y-axes as well, which can easily confuse the reader. To make the manager’s (AND the client’s) life easier, the line plot being used to visualise the “Goals:Game Ratio” can be removed from the plot as it is not really needed by the client in this context. We should never plot everything we have in a single plot (EVEN IF WE CAN!).

Now that is an improvement! 1 type of visualisation and only 1 y-axis! Let’s see what else we can do now.

Use clear and sufficient labeling in the charts.

As we have seen in the previous plot, there is no title or axis labels at all. Some of the x-values (the names of the footballers) also overlap each other, making it challenging to read their names.

Here what we have done is adding an informative title to the plot, labelling the axes properly, and rotating the x-values (the names of the footballers) slightly so that they become more readable. You can rotate the values at any degree that you want. Now the reader will know what this plot is about easily and what the values represent.

Avoid the use of distracting fonts or elements on visuals.

Did you notice that the font used in the plot was quite unusual? This is because it is not the font that is used by default. When it comes to the business world, the use of fancy fonts should be avoided (EVEN IF WE CAN!!) to ensure that the results are clearly presented (unless the client really wants a specific font).

And while I was at it, I noticed that the legend was blocking part of the bars. I relocated it to just outside the plot so that nothing gets blocked by it.

It just keeps on getting better!

Select colours appropriately. Be consistent in the use of colours, do not overdo it, and avoid patterns.

In this aspect, we have got something right, and something wrong.

What we have done right is that we stayed pretty consistent with the colours. They are all light colours, and the colours being used do not have a high level of contrast between them.

What we have done wrong is that there is a pattern in the colours of the plot. The Red/Yellow/Green (RYG) is more intuitively used in cases where the results are related such as High/Medium/Low or something similar. But in our plot, the values being displayed are point values and do not represent such things. Different, but consistent, colours can be chosen.

Always aim for comprehensible visuals. Take a look at the bigger picture to check what can be added, removed, or tweaked so as to make the visual more informative.

As it can be seen above, for this step, the plot was tweaked so that the Grouped bar chart was converted to a Stacked bar chart. This transformed the 3 bars that each player had into single bars. The comparison on the bars are between goals scored at club level and at national level.

The total goals scored by the players have then been added on top of their respective bars. This makes it much easier for the reader to get all 3 information (Club/National/Total).

Conclusion

As you saw, the plot was changed as shown below.

Before and After:

There has definitely been a great improvement in the quality and interpretability of the plot after doing all these steps. Whenever you have to make a visualisation, it is good to first take a moment to think about what you really need to communicate and what would be the best type of visualisation for that. If you are not too sure about what I am saying, please have a look at our training course which is called Data Visualisation in Python. This will help you massively in getting a good understanding and control of visualising data using Python.

Thank you for taking the time to read my blog! I hope it helped!

By Parwez Diloo

Leave a comment

Your email address will not be published. Required fields are marked *