Here's a recap of some of the activity related to Sport Analytics at rstudio::global(2021).
Q/A: What's the next step you want to take with Sports Analytics?
Tony: I want to make work that is more accessible. I'm primarily thinking of having a shiny app or some dashboard for a project. That way others can interactively play with the data/model that has been collected/generated
Jarren: For me, it would be starting my passion project and rolling with it. I have been hesitating for a while because I want the process to be perfect, similar to the way that we envision data analysis to look like. However, I need to approach it in the same way I do every other data analysis task I tackle, which involves a lot of problem solving and perseverance.
If only we had more time in the day dedicated to passion projects ...
James: I think advice or guidance that may be helpful is for those DS professionals who are interested in the field, is how to get your analysis or ideas noticed. Specifically in sports who are traditionally "entrenched" in analytical approaches. It seems like there are so many great resources and content out there, how does one standout?
Tom to James - I think that some of "meta collection" sources like OpenSourceFootball.com, hockey-graphs.com, etc are great for getting "exposure" at the beginning as you are trying to build an audience or get noticed.
Chris suggests: In terms of content to get there, perhaps some kind of weekly or monthly sports analytics "challenges" could be a fun way to kick my butt into getting into the habit of using sports analytics to then kick off a side project.
Chris: I'd like to echo what Jarren said in terms of I think sports analytics would be a great side project. For me personally, it might be a good way to learn Shiny, which is on my "to-learn" list.
My name is Tom, and I write guides/tutorials for R-based workflows at themockup.blog, mostly with NFL datasets.
R by Ryo
My name is Ryo and I do a lot of soccer analytics with R as a hobby. You can check out my work on github - https://github.com/Ryo-N7/soccer_ggplots which has links to my blog posts, soccer data, and R code.
Ex. tutorials for web scraping soccer data, visualizing StatsBomb data sets, deeper analytical dives into the Premier League, J.League, Copa America, etc.
Anybody else doing soccer analytics stuff? I've met a couple of people from RStudio::Conf and useR in the past so would be nice to get to know more of you!
Domenic - I recently wrote a couple of blog posts on probabilistic xg: https://www.allyourbayes.com/post/2020-12-10-uncertainty-in-xg-1/
It would be great to hear your thoughts.
https://telaroz.netlify.app/ I just started a blog which right now only has one entry on how to draw a handball court in ggplot: https://telaroz.netlify.app/ You can also contact me via https://twitter.com/telaroz.com.
I have written a couple of blog posts (https://www.allyourbayes.com/). Only on Bayesian models of xg for now, using data from the StatsBombR library. I'm hoping to add some more analysis soon.
Hi! As a side project, I analyze my Strava data: https://www.datannery.com/post/2021-01-07-scrape-strava/
Nate to Sport Analytics - January 21st, 11:24 AM
Hi everyone, I'm Nate and I (get to!) work with volleyball data full-time. Volleyball certainly isn't as mainstream with data analytics as many other sports, but we're steadily improving with what we've got. I've personally drawn a lot of inspiration from other sports, particularly football/soccer and Expected Goals and developed an initial frame work for an Expected Kills (point won on attack in volleyball) model that I've detailed here: https://volleyviz.netlify.app. I'm always looking to learn more and improve my skills with R and GitHub (my very desolate page: https://github.com/natengo1), and would love any feedback or advice on what next steps I might be able to take to expand my knowledge base. Feel free to connect with me via Twitter https://twitter.com/natengo1
there is a data scientist in New Zealand that started an open source project called OpenVolley (https://openvolley.org) that is building machine learning tools in R which is incredibly exciting.
Jarren - January 21st, 11:28 AM
I love the xK use case that you went over in your blog post here: https://volleyviz.netlify.app/post/expectedkills/. Thanks for bringing these awesome volleyball data use cases to the table.
workflows for eSports analysis using R?
Jarren to Sport Analytics - January 21st, 8:03 AM
Are there people out there that have designed workflows for eSports analysis using R? I have been wanting to tackle a project for quite some time using Oracle's Elixir (https://oracleselixir.com/) and other similar sites that aggregate data, but I haven't been able to find too many examples of ongoing/continuous eSports projects.
Jarren Santos to Sport Analytics - January 21st, 1:13 PM
I am curious to learn more about what eSports data sources and what people are tapping into for mass-multiplayer games like League of Legends (LoL), Counter-Strike: Global Offensive (CS:GO), Fortnite, just to name a few.
My go-to for LoL data is Tim Sevenhuysen's Oracle's Elixir (https://oracleselixir.com/), but I know a lot more exists within the League of Legends community and outside of it.
Thoughts (potentially a Tidy Tuesday dataset of interest)?
Benjamin Farkas to Sport Analytics - January 21st, 12:46 PM
I work for Crescendo Technology, which is a Sports Analytics firm based in Toronto, ON. Crescendo works on modeling of many sports (including eSports) and we provide our insights to partners. We have a team of ~20 Data Scientists and Machine Learning Engineers, working on cutting edge Sports Analytics. We have released several packages to CRAN and maintain an internal R package ecosystem w/ ~70 packages. We are hiring for all positions and are looking for folks who have a Data Science background and are interested in a career in Sports Analytics. Remote work is possible. If you are interested, please message me directly here or email
hr at crescendotechnology dot com.
Also, our team member, Aaron Jacobs will be giving a talk