The promise and perils of sleeping at hackathons

Vadim Liventsev
4 min readDec 17, 2018

If you don’t want a civil war to break out spontaneously in your little friend group, you better not discuss politics, religion, frontend frameworks and sleeping at hackathons. Which is exactly what I recently found myself doing. I suggested that maybe (just maybe) you shouldn’t go to an event where they literally hand you caffeinated protein shakes and energy drinks to keep you coding for 48 hours to then pretend like you’re a responsible adult™ and sleep 8 hours a night. My beloved opponent argued that clear thinking is your main asset and it is foolish to flush it down the drain by sleep depriving yourself even if the environment encourages it.

Work in progress at Junction 2018 © Aino Röyskö

Shots were fired, arguments were exchanged, growling was had, so in the end we decided to go get some hard data and asked the participants of Junction 2018 if and how much they slept during the event. The survey also included questions about team size and whether said team won any award to see if team’s results correlate with sleep. The hackathon started on Friday evening and concluded Sunday morning: in normal circuimstances, one would sleep for 14–18 hours in that timespan.

Results are in: 52 people filled out our questionnaire. So, without further ado…

All participants

Result #1: people do sleep at hackathons! A lot! 34% of respondents had pretty reasonable healthy amount of sleep and 76% slept for more than 6 hours and only one guy (girl?) claimed that (s)he didn’t sleep at all.

Non-winners

People whose team didn’t receive any award.

Result #2: the person who didn’t sleep at all didn’t get anything. Oh well.

Challenge winners

People whose team won one of the 45 challenges

Looks like challenge winners did not in general sleep less than others. Also, 2 out of 3 people who said they slept “whenever they felt dozing off” ended up winning challenges. I guess “follow your heart and fall asleep on your keyboard when you feel like it” is a viable strategy.

Track winners

People whose team came first in one of 11 tracks. 4 track winners filled out our survey.

Looks like track winners did sleep a little more than challenge winners.

Conclusions

Overall, we can see some evidence that sleeping more improves your results. However, since the difference isn’t too big and we only have 4 track winners, statistical significance of this result is highly questionable.

What isn’t questionable is this: it is possible to have a good night’s sleep and still win!

Data

Don’t like this analysis? Do your own: we release an anonymized dataset with all responses.

Acknowledgements

Everyone who participated in the survey: you rock! You helped us get some answers to the question that bugged many a hackathon-dweller.

Junction organizers: you rock as well! Thank you for running one of the best hackathons out there.

Special thanks to Maria Sheyanova for the debate that sparked this research and to Alexander Prosolkin for reviewing it before it was published.

--

--