My First Datathon: questions, lessons and preparing

Photo by Annie Spratt on Unsplash

“This is bad, this is really bad, I suck at this and I have been applying to jobs on it for the past 6 months. I thought it will be easy to conquer. it turns out I have absolutely no clue how to deal with it. I suck and I am sad”

This what I wrote hiding in the washroom 2 minutes and 1 hour before presentation time at my first datathon. If you reading this you are either going for your first Datathon, interested in data or medium recommended this to you (thank you medium!).

If you are not sure what a datathon is, let’s just say it is the closest thing to a pressure cooker for data scientists, data analysts or anyone interested in data. You have to come up with insights within a limited time period.

One of the people in my network recommended that I do it to get more practice and increase my job likelihood as a data analyst.

Before I left for my datathon I researched how to prepare. There was very little information other than this one LinkedIn article by Mari Takashima that gave me some foundations of what to expect. So I wrote questions before going to the event. I decided that this would be an experiment and I needed to make some assumptions. So think of them as exploratory data analysis on the event. The answers are what I experienced, so take them with a really healthy grain of salt. Remember it’s my first🤷🏽‍♀️

Okay let’s dive in

🏊🏼‍♀️

Should I bring a Laptop?

Yes! especially if the organizer made no promise of a computer. It also allows for flexibility. Datathon’s tend to be intensive because analysis has to be made in such a short time. Having a working laptop would allow you to be a contributing member to whatever team you join. Even if you decide to work alone, you can work with ease. Unless the organizer says no laptops, taking yours is always a good idea. Remember your charger too!

Do I need to download any software?

This depends on the needs of your project and your event organizers. My experience was that python programming on the online jupyterlab was fine. If unsure, ask your organizers what programming language, analytical software or statistical package most other participants will use. That will give you a rough idea of what may be used for analysis and. If you need to download anything.

Team formation: by me? chosen for me?

Most Datathon participants work in teams. It is better to be in a team, to be honest. I have read tons of articles that the lack of collaboration in data analysis is a big issue. I was glad to be in an environment where collaboration was encouraged. My experience was that we were given the liberty to choose our teams.

One way to pick a good team is to partner with people that complement your analytical weakness. Don’t be afraid to ask if you can join a team. You may be feeling nervous but remember you are all there for the love of data(or for the love of money, if your Datathon has a monetary gift😏).

How much expertise should I have?

Okay this is where should pay attention, so you don’t end up writing sad poetry in the toilet. Focus on your data cleaning skills. So you can get the data cleaned and ready to analyze in a shorter time than you usually do. Figure out how to write functions on a whim that clean your entire dataset by just changing a few variables. If your dataset is already structured, cleaned and manipulated to your taste by your organizers, then maybe learn how to do other things faster like machine learning techniques,or data visualization. I doubt your data will be organized to the extent that you don’t need to clean any spec of it all.

Are there anything else to bring to make my experience easier?

Yes! something to munch on, a water bottle. Locate the bathroom once you get there. Locate a quiet space to chill out, if you feel overwhelmed. Locate a whiteboard. Locate an outlet or many outlets for your team. Locate good sitting space and tables. Locate a datathon newbie too. Trust me, it’ll calm your nerves. Ask for the WiFi. Ask to be added to the communication channel for the day. Ask for whatever you don’t understand.

Lessons

If where to go back in time , what would I tell pre-datathon me:

1. Pressure teaches you a lot about your skill set. For that datathons are a good test of your ability, areas of growth and your muscle memory. I was struggling to make a simple excel chart under that pressure.
2. Team is for supporting your weakness. In the words of my sister “Be confident in your strengths and don’t feel sorry for your weakness”. Be upfront about it. It sets the tone for what is feasible given the time constraint.
3. Growth mindset or Analyst mindset,what ever you prefer to call it, is what you need. If you did great, great! Learn from others presentation. If you feel like you didn’t, that’s sad but also a great opportunity to learn. If you start nursing emotions of success or failure it would be hard to pick up on something someone else did that could increase your skill set. Treat your first as an experiment, don’t get caught up in emotions!
4. Your backgrounds are not all the same. Some people do this analysis stuff on repeat everyday. Some have jobs in it. Some are newbies. It adds for the diversity.
5. If you are a newbie, partner with an experienced person. They sort of know where this is going. If you are partnered with newbies, no fretting. It gives you lots of room to do whatever you want. Get creative
6. Whiteboards. Whatever you do, start by detailing out how you want to go about analysis. It could be done on paper, whiteboard, Ms word, google docs e.t.c
7. Narrative is everything! Remember that insight involves storytelling. Ask yourself what do I want the decision maker or audience to take away from my presentation? This helps you build a presentation around that. I found that even though time constraint didn’t let me dig as deep into the data as I would have liked. Narrative save the day!
8. Presentation is as important as analysis. Spend some time planning it. My team chose to create presentations 1 hour to presentation time.

Preparing for the next one

Would I do it again? YES!

I am off to more coding to prepare for the next one.

Enjoy your datathon 😊