How to win a sports data hackathon

Winning an on-site sports data hackathon requires four key ingredients:
Preparation, simplification, technical skills and slick presentation.

There are hackathons for almost everything. Hackathon winners have chances to take home prizes, hold bragging rights and can use their competitive badges on their way to job interviews.

In a hackathon you work like crazy for 6 to 72 hours and then have 2–5 minutes to present, often being disappointed because the best technical team did not win, or the simplest idea did.

If you in the Toronto area on March 11, 2023, join us for a 7 hour baseball sports data hackathon (strictly in person and FREE) and flex your collaboration skills.

So how do you increase your chances of winning the prizes?

Rule #1 for winning a hackathon: Prepare by building your team before the event

Ideally your team of 3 to 5 people is made up of 4 roles:
- Frontend and UI/UX
- Backend/ Full Stack developer / Data Engineer type
- Business subject savvy visuals techie
- CUO (Chief Urgency Officer)

Form, storm and norm your team as early as possible before the event. Attend with your cohort or people you have worked with before. Don’t waste time on niceties.

Your CUO will keep the pressure up focusing on ideas, flow and timekeeping ….. You just cannot afford to run out of time.
In a hackathon, every minute counts.

Rule #2 for winning a hackathon: Simplify everything

Sure, coding a blockchain based voting system is cool and technical. The problem is that explaining your incredible design to a set of judges can take more than 3 minutes. You lose.

When time is limited, you need a simple story flow to convey your message. Problem/Challenge, Solution, Why and How, Thank You. (Optionally What We Would Do Next)

For example, imagine a Baseball sports industry hackathon where you need to solve a problem using data.

Problem/Challenge: Major League Baseball (MLB) needs to build up it’s younger demographic fanbase over the next years and wants to emulate the success of the NBA which has massively increased it’s younger following by leveraging and supporting the social media of it’s most popular basketball player personalities throughout the season and beyond.

Solution: An AI based system to learn about, and monitor the social media sentiments of the “biggest most engaging” baseball players and teams and feed that information into MLB targeted marketing campaigns toward younger demographic groups of baseball fans and casuals.

Why and how: The MLB average age is 57 and the game needs to evolve to bring in the younger demographic, who view the game in completely different ways to their grandparents. This new demographic wants excitement, shorter games, athletic plays and see players showing emotion and personality. They want to be more connected to baseball players behind the scenes via social media. Ultimately the more connected they are, the more money they will bring into the sport.

Thank You and Next Steps: We will implement a pilot as below to test our predictions for three players: Vladimir Guerrero Jr, Mike Trout and Shohei Ohtani.

• Find the Top 10 baseball players by levels of fan engagement eg tweets, IG.
• Build a model to predict future levels of fan engagement
• Recommend MLB marketing campaigns timed around high levels of fan engagement

Rule #3 for winning a hackathon: Use your strongest technical skills

The competition winners don’t usually learn new techniques during the hackathon. Because no-one has time to learn about new shiny objects during the hackathon.

Do your coding and learn what you need before the competition…. And know your strengths and weaknesses.

Ideally, you should mould the hackathon goals to fit your best skills and techniques. For example if you are an expert on data science NLP, then make the presentation about semantics.

If you are wizard at Tableau presentations, then deliver an incredible Tableau Story presentation to the judges.

If real-time big data analysis is your speciality, then throw together a fast data pipeline and front-end solution to showcase your skill, but make sure that the front-end is user-friendly

For example, a recent hackathon winning team won a UI/UX hackathon not because they were the best designers, but because they had a team member who was a genius with Apple Watch application development.

Rule #4 for winning a hackathon: Slick beats dull in the 5 minute presentations race

If you have listened to eight monotone poorly presented hackathon presentations, you start to look for things that stimulate your brain.
As a judge, you remember the slicker presentations because it gives a feeling of an organised team, regardless of the content.

Too often teams think that everyone has to say “something” at the presentation, but not everyone is comfortable with public speaking, or has experience with QUICKLY explaining the work and concepts. This is a strange consequence of the hackathon format and they (of course) run out of time.

Too often, elaborate technical demos are created but are too complicated, and eat up the time without being explained.

Winning teams generally start preparing their presentations outlines within the first 20% of the hackathon time, and fill out the content as they go along. They practice before the cutoff time.

Don’t forget to include a slide at the end showing the individual team faces, QR-Codes and one line bios that can be displayed during the Q&A.

If you love data, lots of data and the sport of baseball look out for the Baseball HackDay Toronto happening in person on 11th March 2023 and I will see you there.

Bonus Rule: ENJOY YOURSELF

Confession: I won an AI hackathon and these principles worked perfectly.
I was part of Team “Synthetyk” that won a competition organized by Aggregate Intellect (Toronto) to build a product based on Generative Adversarial Networks (GAN). We built an AI data augmentation product to synthesize positive breast cancer tissue images to help improve accuracy of classification of the disease.

And yes, I sometimes do organize hackathons.