Day 1 of Dashboard week saw the Data School enter the wonderful world of betting as we celebrated the Melbourne cup in the only way we know how… In a uniquely Tableau style!
First off, we were given 3 datasets by Craig. These datasets covered (a) all horses competing in the 2018 Melbourne cup and info about the horse and their rider (b) all horses that have previously won the Melbourne cup since 1983 (c) all horses competing in the 2018 Melbourne cup and info about their performance. After looking at the datasets, figuring out what was cool and worth visualizing, I decided to focus on datasource (a) because it had the most relevant information to create a radar chart that shows the performance of each horse within a race so that the audience could make an informed betting choice at a glance.
Having narrowed the scope of what resources I had access to, and decided on what I wanted to create with it I began my dive into Alteryx and parse out the info I needed to get me there.
Alteryx
It was a lot of parsing. A lot and a lot of parsing.
BUT the concept of what I did is pretty simple, I systematically cleaned each data column based on what was required, where patches of blue represent my data cleansing stages with a focus on feature engineering new columns. The green represents using the Regex tool to parse out particularly info-dense columns so that I could use them in my analysis.
The other interesting thing in this workflow was using a join to attach repeatable info about the race (URL, track length, track name, prize money) to itself so as to avoid using a multi-row tool for every column.
This took me up until lunch, and following that I jumped into Tableau with great determination to create a radar chart.
Tableau
Creating a radar chart was simultaneously more difficult than I expected, and far easier than I expected. A youtube video by Andy Kriebel gave me a good understanding of the concepts behind a radar chart, whereas some awesome charts from tableau public gave me guidance in taking my own data and massaging it into a form suitable for a radar chart.
This meant jumping back into Alteryx, using a transpose tool, and binning my values to normalize the scale before jumping back into Tableau to create a polygon joined by the values in the radar chart.
After that, I decided I wanted to keep some information on the dashboard at all times (The horse and jockey info) – and with a bit of formatting I came to the end of day 1.
Click here to check out my dashboard!