Introduction

This year marks the fifth VandyHacks, and we couldn’t be more proud of all the people that came and the awesome projects that they built! From November 2–4, we hosted a record number of 635 RSVPed hackers, plus additional walk-ins that pushed us over 700 hackers at the Wond’ry at Vanderbilt University, a 41.4% increase from last year. In the true spontaneous nature of a hackathon, our overall winner was “Seismic Data,” built by a team that had no previous hackathon experience and didn’t know each other before coming to the event.

This article seeks to break down statistics from the event, examine places to improve for the future, talk about successes, and finally present our outlooks for future years. If you just want to look at the stats, they’re grouped right below. All of the submissions from this year are here.

By the Numbers

The breakdown of this section is broadly divided into general stats, school, major, and then diversity. Directly below is a comparison of this year’s event compared to last year, followed by more detailed breakdowns combined with brief analysis against previous years. For additional comparisons, you can check out our recap from last year.

VandyHacks IV and VandyHacks V Changes

Overall, we saw significant growth from VandyHacks IV. While the number of applicants did not rise significantly, the number of submissions, schools, and submissions rose a staggering amount. The rest of the percentages were about the same as last year. Read on for more detailed analysis!

Breakdown & Analysis

Just from looking at the numbers, VH5 broke all previous records in numbers of attendees, submissions, unique schools, diversity in terms of gender and ethnicity. These charts indicate that the steps we are taking are moving us in the right direction. Following each chart is a brief analysis.

Attendance

The day before the event, we were up to over 900 RSVPs, a quantity that astounded expectations and sent our logistics team into a flurry. While the total number of attendees was lower — 635 not including walk-ins — the growth percentage was an astounding 41.4% from last year. This year, we also developed an NFC system in the form of light wristbands, allowing us to collect precise attendance data for meals, talks, and workshops. With this data, we will be able to plan more effectively for future years and streamline logistics pre-planning.

Submissions

We love seeing the projects that hackers produce, and it came out to an incredible 97 submissions, which you can check out for yourself at our Devpost. The number of submissions replaced our previous record and reversed the decline from VH3 to VH4.

Unique Schools in Attendance

This year’s number of unique schools nearly doubled from last year, increasing by 74.4%! We’re so excited that so many different schools are coming to our event!

Graduation Year

The total range between 2019–2022 graduation years was small, coming to 43 hackers. Once again, sophomores have the strongest presence, representing over a quarter of the total attendees.

Major

Our mission is for our event to be open to anyone, regardless of major. While Computer Science majors outnumber any other category, we’re proud that there are a variety of other majors listed. For the sake of legibility, majors with <5 people were not included on the chart, but are as follows: Political Science and Government, Music Performance, Chemical Engineering, Computer Programming, Cognitive Science, Engineering Science, Information Technology, Civil Engineering, Computer Software Engineering, and Biochemistry and Molecular Biology. While most of the non-CS majors are in STEM fields, we’re glad that 31% of our hackers are non-CS.

First-Time Hackers

VandyHacks strives to have our event allow anyone to create a project, regardless of major or previous hackathon experience. Our overall winner “Seismic Data” was comprised of a group that had no previous hackathon experience, a feat that compounded their achievement. We had an incredible 38.74% of our hackers have no previous hacking experience, and we’re elated that VandyHacks V was their first. This was a decrease from last year, where 46.10% of our hackers were first timers. We hosted a collection of workshops (expanded on below) designed to give these beginners a strong start, regardless of any previous computer science experience. We hope that hackers who perhaps never had any previous development experience were able to walk out of VandyHacks V with a project and newfound confidence.

Gender Diversity

This year, 25.63% of our hackers identified as female, with an additional 1.34% identifying as neutral, and 0.67% identifying as other. Those identifying as male represented under 3/4s of the total hacker population for the first time!

Ethnic Diversity

Once again, those identifying as Asian/Asian-American accounted for the majority of hacker population — just over 50%. This year’s split is very close to last year’s, and we seek to improve on increasing diversity. The steps that we have personally taken are in line with the rest of our mission: to make our event as open as possible to all, and provide tools for all levels to participate, despite any stereotypical images of “tech” or “computer scientists.” Industry-wide, there are plenty of other articles that further discuss this topic.

Key Takeaways

Putting on an event of this scale means that there are a lot of different things to cover, and there are according flaws. For future years and planning, we seek to pick apart these flaws to understand how to preemptively combat them and more effectively respond for future years.

Statistical Takeaways

While we’ve improved from previous years, we constantly strive to improve attendance, submissions, and diversity. To do this, we look past the statistics and try to understand how to make our event as friendly and possible. The improvements we can make are increasing more targeted marketing, strengthening our relationships with groups such as Women in Computing on campus, and hosting additional events throughout the year to make a “hackathon” a less intimidating term.

Logistics

In our retrospective meetings, many points of improvement for the logistics team were brought up. These criticisms were fair — managing an event caring for several hundred people in an efficient, quick manner requires careful planning and setup, and [the author’s] team did not prepare sufficiently in some instances. However, some of the problems were because of insufficient communication from caterers, which led to last-minute adjustments. For the future, we plan on taking all the experience and data we’ve collected from this year’s event and channeling them into making next year’s event a strong improvement.

Judging

This year we decided to use Gavel, a system from HackMIT that does pairwise comparisons. It’s supposed to be more fair than having judges rank everyone and then averaging the results. However, we ended up having to fight with the import format from Devpost, and only got it at the last minute. As a result, people were assigned tables incorrectly and there was a bit of confusion which took about a half hour to clear up.

Mentoring

We also decided to revamp our mentoring system from VH5, and use a software solution. We chose to use HelpQ, also from hackMIT because it had been well-tested and seemed to have lots of features. Unfortunately, we didn’t foresee that everyone needed to make an account beforehand, and using an ad-hoc username/password assignment right before the event led to a lot of confusion, and may people didn’t know the system existed, or that they had been assigned a unique username/password. As a result, very few people (both hackers and mentors) ended up using it. In addition, waiting in a queue for a long time is pretty frustrating for hackers when they could just walk up to a mentor standing several feet away and get help immediately.

Dayof page

Having a system that let hackers know what was happening at every moment of the day was a huge priority for us in order to lead to better hacker engagement and to ensure everyone knew what was going on (eg. food, fun events, etc.). In order to achieve this, we decided to create an event notification system (using a slackbot, an SMS sending mechanism based on Twilio, and a browser push notification option on our “dayof” page). The dayof page was built in two versions — one for browsers, and one specifically to be rendered on the TVs. Although we worked extremely hard the days leading up to the event, it proved to be somewhat buggy and the TVs didn’t connect to the wifi (we used Raspberry PIs that hooked up to each TV running Raspbian with Firefox). It did have cool moving animating clouds, so that was a huge bummer. Oh also our slackbot got hacked due to various bad configurations in our Express/Node.js server serving an unprotected HTTP endpoint (protip: don’t do that :)).

Successes & Cool Pictures

Not only are we proud of our hackers, we’re proud of the team that made it happen. In this section, we’ll show off some of the things that helped our event run more smoothly, the work of our dev and design teams, and share pictures of some events that helped everyone destress. Janet Fang took awesome pictures for us all weekend, and we’d like to display some of them here.

NFC

Our dev team created an NFC system that made everything significantly more efficient. It streamlined meal, event, and workshop check-ins, making everyone’s lives better. As a result, we have precise data for every single event with hackers, meaning that future planning and expectations will be much more precise.

Design

One of the core components of a hackathon is the awesome swag from us and sponsors. When selecting different items, we really put a lot of pride into getting things we hope people will enjoy. This year’s swag included tumblers, beanies, toiletry bags, stickers, and, of course, t-shirts.

Food

This year, we had the chance to try out new caterers and revisit old favorites. Also, we experimented with making cold brew alongside hot brewed coffee this year. Creating cold brew for a few hundred people was quite an experience, and our work paid off. We had warm responses to our cold brew, and will adapt our approach for future years.

Workshops

At the start of the event, we began with a 2 hour workshop for beginners introducing the event and providing starts in an array of areas. That was followed by intros to Front-end Web Development, Flask, and NodeJS. Throughout the rest of the event, we had additional workshops (e.g introductions to Unity/VR) and more given by sponsors such as Fulcrum GT’s Intro to ReactJS. Learning is a key goal of hackathons, and we hope that we inspired people to experiment with new technologies.

Events

After our midnight snack of too much Jeni’s ice cream, hackers gathered for Karaoke. From classics like Mr. Brightside to System of a Down, we had a blast destressing as the event wrapped up. Also, MLH cupstacking was pretty hype.

Looking onwards to VandyHacks VI

We will continue to adapt and grow from our experiences. On the fields of diversity, logistics, and more, we will fine tune our approach to perfect not only VandyHacks as an event, but as an organization.

Thanks

Here’s to VH6! Thank you to all of the awesome people that came, we really loved seeing the projects that you built! Thank you to all of the mentors,

To see all the pictures that our photographer Janet Fang took for us, check out the pictures from Day 1, Day 2, and Day 3.

Be on the lookout for more detailed articles coming out later!

VandyHacks V Recap was originally published in VandyHacks on Medium, where people are continuing the conversation by highlighting and responding to this story.