This is my weekly newsletter where I share some useful tips that I've learned while researching and writing the book Hands-on APIs for AI and Data Science, which will be published by O'Reilly Publishing in April 2025.
Tip Sheet #18: Deep Dive on the NFL Big Data Bowl
|
Hello Tip-Sheeters, I hope your week is off to a good start. When I'm asked how someone can start or grow their data career, my short answer is generally "Build something and share it". We're in the golden era of cloud computing, with hundreds of low-cost ways to build data apps and share them with friends, colleagues, and potential employers for minimal cost. There's just no reason not to jump in and get started. One fun way to do this is by joining an online data science contest like the National Football League (NFL) Big Data Bowl. This week, we're taking a look at that competition. Let's get started! What's So Special About the NFL Big Data Bowl?The NFL is the most popular U.S. sports league, and TV broadcasts are full of stats and analytics. In the last few years, the NFL has worked with Amazon Web Services (AWS) to host a yearly data analytics contest with real prize money and great publicity. A special angle of the contest is that it focuses on real-time location tracking data. Here's a good explanation of this from Ron Yurko from Carnegie Mellon in a recent Sports Analytics Podcast. "The NFL has chips in the shoulder pads and the ball. So at every tenth of a second, we know where every football player is at on the field, on a 2D matrix...This competition inspires people to work on different types of problems by having a theme associated with it every year."
- Ron Yurko, Carnegie Mellon
Even though the contest is only 5 years old, the official website says that more than 75 competitors have gotten jobs in sports analytics. Yurko explains what a unique opportunity the competition is to show your work to sports teams: "Just being able to enter in something like this is incredible. Everyone is looking at this. Not just NFL teams: MLB teams, NBA basketball. That’s the other purpose of this, it’s a pipeline (for hiring)." What's in it for the NFL?The competition has been spearheaded by Mike Lopez, the NFL's Senior Director of Data Analytics. In an interview with CNBC, Lopez said that the league offices want to learn from the competitors, because this tracking data is so new they haven't had time to explore it all. And NFL teams want to hire those same competitors so they can get an advantage for their own teams. In a Q&A on the AWS website Lopez said that the data is so new that "when you’re filtering it or looking to apply it in a specific way, you're probably the first person in the world to ever do that with this data." This Year's FinalistsAt the time of writing, the finalists have been selected in the 2025 contest and their entries are listed here. Five teams in three categories will be traveling to the NFL Combine in Indianapolis in a few weeks to pitch their entries to NFL executives and teams.
The entries are available on Kaggle, so I examined one of the finalists in the metrics track that created a metric named Safety Entropy. They did a great job explaining the problem they were trying to solve and sharing the research behind their approach. Safety Entropy uses a neural network and has some good visualizations to show how it can be applied. (You'll also notice that Ron Yurko and a teammate made the semi-finals with their entry this year.) Why Not Enter This Fall?Anyone can enter the contest, which includes you. Past winners have included people from around the world and various professions. Matt Chang, a winner last year, shared some helpful advice here:
I hope some Tip Sheeters enter the contest this fall Keep an eye out on the NFL's Kaggle page in early October 2025. ----- That's it for this week. I hope you see the fun and value of creating data projects and finding ways to share them. Keep Coding. Ryan Day |
Ryan Day
This is my weekly newsletter where I share some useful tips that I've learned while researching and writing the book Hands-on APIs for AI and Data Science, which will be published by O'Reilly Publishing in April 2025.
Hi Tip-Sheeters, This week I'm excited to share a chat I had with Sebastian Ramírez Montaňo, creator of FastAPI and founder of FastAPI Labs. There's some good stuff in our conversation, including Sebastián's tips on skill-building and advice on staying positive. If you've had this email forwarded to you, I hope you'll consider subscribing weekly at https://tips.handsonapibook.com/. The Big Announcement For background, you might have seen the announcement a few weeks ago on Sebastián's...
Hello Tip-Sheeters, I hope you're having a great week. In this issue, I'll take you through some initial Python coding for Model Context Protocol (MCP), a recent open-source project led by Anthropic. I also share a behind-the-scenes look at book launch day, which was a lot of fun and successful. MCP and Your APIs In Tip Sheet #19, I shared 7 steps to getting up to speed on MCP (or any new technology). In that, I gave some background on what MCP is and how it relates to using APIs with AI....
Hi Tip Sheeters! This week I will take my first look at Model Context Protocol (MCP), which Anthropic rolled out during the holidays last year. Along the way, I'll share my approach to getting up to speed on a new open-source framework or tool. Let's get started! 7 Steps to get up to speed on MCP (or any new tech) There are some repeatable steps about how I get to know a new tool. One thing to remember up front is that just because someone proposes a standard or releases a tool doesn't mean...