This is a guest post from James Mullins and Bertie Kennedy (Performance Advantage Ltd), summarising the key points of the webinar we hosted with them on 14th May 2020, which can be viewed in full here
Our goal at the start of this project was to seek to understand each player’s independent style of play, and assess how this impacted upon the point/match outcome.
Using the updated Graphic Descriptor tool released earlier this year in Nacsport V5, we were able to plot shots based on the XY coordinates at the start and end of each point. Our tagging panel gave each point context, seeking to provide a better understanding for the questions coaches would like to answer.
Figure 1 shows our final panel, which was the result of 7/8 iterations of the build process.
Figure 1 – Nacsport Tagging Panel including XY
Automatically added descriptors enabled us to keep a consistent descriptor order within the XLS export, with subsequent descriptors each relating to independent XY plots. This enabled logical data management within Alteryx.
Figure 2 – Nacsport XLS Export Format 2
The Nacsport XLS export format 2 (including XY values) was the most appropriate format for this project, as we wished to collect the same descriptor multiple times and maintain the collection order.
Alteryx Data Management/Manipulation
Using Alteryx to handle the data management made it possible to convert the data from a ‘horizontal’ to a ‘vertical’ representation.
Figure 2 shows that each row of data refers to an entire point, with XY data hosted within a single cell and other descriptors moving across the page from left to right. After changing this from point-by-point to shot-by-shot, it was then possible to conduct the same process with the XY cell, by parsing each coordinate out and then transposing (similar to text-to-columns in excel).
To maintain the links between shots, origin and destination XY columns were added, where the destination = the origin of the subsequent shot XY within the same point. This made it possible to plot the XY shot map in Tableau.
Finally, Alteryx permitted the addition of further contextual data (like new fields) to each point, by associating other descriptors sourced via Nacsport with the XY data. This process resulted in a shot-by-shot structure seen in Figure 3.
Figure 3 – Post-Alteryx Excel file (Shot-by-Shot Structure)
Tableau Error Checking
By building extensive filters into Tableau worksheets, it was possible to develop ‘impossible scenarios’ to check for any errors within our data. The structure of the Nacsport/Alteryx files ensures the descriptor count is consistent so that cross-checking is possible. An example of this is discussed in the webinar here:
Visualising Nacsport Data in Tableau
Building the graphics in Tableau (Figure 4) provided a basic shot plot with a heat map option. However, setting the data in Tableau allowed greater management of the data and the application of multiple filters.
Figure 4 – XY Shot Plot and Heat Map in Tableau
We were now starting to produce actionable information for coaches to use.
We added more value by producing multiple calculated fields and sheets which were combined in dashboards (examples within the webinar from 46:25).
Linking Data to Video in Tableau
Posts in the Tableau Public community have inspired the integration of video within many of our workbooks, including the badminton workbook. The visualisation which sparked this is below:
Knowing that Nacsport had generated timecodes, it was possible to produce a point time and subsequently a shot time for each row of data. Alteryx automatically produces a unique URL with the timecode for each row and by using URL actions, it’s possible to manipulate the data and link directly to the relevant video. Moving this on a level from the Astros example, it was subsequently possible to fully integrate video within a Tableau dashboard rather than a pop-up browser window. This, along with building a playlist function, now allows a coach to manipulate data, draw and test conclusions and view video all from a single screen on a single software platform.
A marriage made in heaven – Nacsport & Tableau!
For full explanation and demonstration of all the points made in this blog, please take time to check out the full webinar. Should you wish to follow any elements up, please contact us on phil@perfadvantage.co.uk or contact us via twitter @PerfAdvantage
Authors: James Mullins & Bertie Kennedy
FAQs
Which version of Nacsport do I need to use the Graphic Descriptor tool? Graphic Descriptors are available from the Scout Plus version of Nacsport.
Do I need Tableau to visualise my Nacsport data? No. Nacsport has the Dashboard tool to help you easily visualise your data within the software. Of course, you can export your data in lots of different formats to use in other third party software like Tableau, R, and Excel though!
Can I trial Nacsport? Yes! You can trial any version of the software free for 30 days. Contact us for more information about Nacsport and to help you find the right package for you!
コメント