This is the main thread for the AthTech trHackathon.
Do you want to participate? 
Just post a message here indicating your interest and introducing yourself (+ your team, if you are more). You don’t need to disclose your great idea.
You can find the rules and more information on the official website.
Feel free to ask any question about the challenge here!
Hello,
Unfourtunately I haven’t been able to come up with something I’d want to do with the awesome datasets that you guys have treated us with. Instead I have a fairly plain and boring idea that I’ve had before. And I don’t even know if I’ll have the time and motivation to accomplish it. But there’s no harm in applying, right?
Scoring road running.
WA scoring tables are a tool that converts an athletic performance in a numerical score and somewhat allows comparing performances across different events. However, it’s only applicable for a predefined set of events. If you’ve run a 3.6 km race or an 8 km morning run, they are of no use. Sure, you can extrapolate the pace and compare it to your 10 km pace, but why not make it more precise and let you gauge how good or bad was that 8 km run?
My idea is to:
I don’t have a team. If anyone is interested to join me, hit me up in the DMs.
I am aware this is not very needed for competitions. It’s more of use for recreational runners or training work. So let me know if this is too far outside the intended scope of this challenge!
Hello @Dzhuris,
I wrote a post about parsing the World Athletics scoring tables PDF here to get the data one would need to do what you describe. It might be relevant for you.
I believe most of the information and data you need for your project can be found in the technical documents under “scoring tables” here however they are in PDF format so you will need to parse the tables and store them (e.g. in json or .csv).
This year, I wrote a PDF parser for this and created an API to access the PostgreSQL tables storing the data for different events (see here). For now the API is not published and it is for my personal use.
You will probably need to do something similar and use the data to calculate the coefficients required to then compute the scores for your distances and times.
I was planning to work on computing the coefficients as a side project later this year! Also, have a look at the work done here by Jeff Chen. You might get some clues about the type of regression you will need to implement to identify the coefficients.
Good luck!
I have removed my message and started a new thread instead. Thank you!
Hi everyone,
I’m really excited about TrHackathon and wanted to signal my intent to participate, even though I don’t have a team or concrete project idea just yet.
I’m a keen runner myself, with a growing passion for athletics, particularly track events, and have competed at a low level for my university. This has sparked a deeper interest in how technology and data can help us better understand performance, competition dynamics, and fan engagement.
I’m currently exploring the field of data analytics, with a special curiosity around sports analytics and the potential of working with emerging technologies like AR/XR, and to create real-time insights, and intuitive visualizations to revolutionize the athletics experience for both athletes and fans is an unmissable opportunity.
If anyone is forming a team or has an idea brewing and is looking for collaborators, particularly around data storytelling, visualization, or athlete performance insights, I’d love to connect and contribute!
Looking forward to seeing the creativity this challenge brings out.
Best,
Alexandre Mauger
Hi everyone,
Apologies for being absent from the forum. Please consider all ideas above “approved” if you wish to pursue them. The Rome dataset is just a starting point, and we welcome all innovations and ideas.
Alexandre, have you had time to explore the data set yet? I’m just throwing out a few ideas here…
With the athlete data, we have just about every race that 1200 athletes did in their careers up to the championships. So this has the potential for much more interesting commentary. For example, if we pretend it’s just prior to a final, one could explore among the finalists…
One might try to predict who is going to win, from their past form.
Then with the IsoLynx data there are many fascinating coaching-type questions…
For steeplechase and hurdles, who slows down least over the barriers, and presumably has better technique?
In all the longer events, who ran what unnecessary distance? Was it tactically necessary to run 8m extra in an 800, or did they pay the price?
Who in a race is going out at a pace they can’t sustain based on their PBs and recent bests?
What live metrics and graphics could be shown in future to help a TV audience quickly grasp these things?
Going out beyond the data set, one can mine Instagram for photos taken in the stadium at the time, and Twitter/Facebook for what people are saying about the athletes.
I’m sure there are many more ideas.
Hi everyone,
Apologies if I’m late to post, I’m also interested in the TrHackathon. I would like to explore some combination of data visualisation and predictive analytics as they pertain to athletes’ careers i.e. can we display in an interesting and intuitive fashion an athlete’s career trajectory and make predictions about their future performances? I am a thrower myself, so I might focus, say, on the shot put to start with.
I am considering using some sort of statistical model trained on the dataset (and maybe additional data e.g. weather data), and maybe other machine learning techniques.
Marley, that sounds great! It should be quite easy to find the throwing events, and then the files in the athletes folder for all of them, which will pretty much give their careers up to that date. Definitely bonus points if you can link to external data.
If it sparks any ideas, here is a graphic a friend and I made up from some discus data in 2017 - you can get a real feel for who throws their best at major games, and who is how consistent. Gerd Kanter was “Mr Consistent”… What it lacked is the time dimension - when they did their bests…
Thank you @Alexandre_Mauger , we are registering your interest in our challenge. We will see if you can be part of team, so they benefit from your ideas and your insight!
Nicolas Launois, from European Athletics
