After submitting the projects partner and peer reviewing will start in parallel.
To ensure quality results, each participant should review at least 15 other projects. However, since reviewing other projects is a lot of fun, you're free to keep going until you've reviewed all of the other projects. If your team does not participate in reviewing other projects, we reserve the right to apply a small penalty to your team's score.
<aside> 📢
Back Offline Bonus: You can also participate in face-to-face reviewing! Simply follow the table numbers to visit and review projects in person.
</aside>
Partners select the challenge winners and runner-ups.
Overall TOP 10 projects are determined based on the ranking on Gavel algorithm. TOP 10 teams are instructed for the pitches in the closing ceremony for 3 minutes each, then hackers vote for the main prize winner.
Video tutorial
How peer reviewing works https://link.junction.so/peer-reviewing-explained
Finalist voting
TBA
This event uses a modified version of Gavel to decide the winner. Gavel was originally created by the HackMIT team in 2015 to solve a large-scale competition winner in a matter of a few hours as fairly as possible.
We at Junction have used it at our events since 2016 and have since improved on the original, as we don't believe outside "experts" are the best people to decide what is a good hack. We want to leave that to the people who know most about hackathons (you, the hackers). Thus we made our own version of the algorithm which allows participants to review each other without being able to skew the results in their favor.
If you're interested in reading in detail about the mathematics and reasoning behind the system, please go ahead and view the original project on GitHub, as well as this excellent blog post written by the creator of the original Gavel, Anish Athalye.
Gavel is a fully automated judging system that both tells you which projects to look at and collects your votes. The system is based on the model of pairwise comparison.
You'll start off by looking at a single submission and then for every submission after that, you'll decide whether it's better or worse than the one you looked at immediately beforehand.
If at any point, you're unable to view the demo of a particular submission, you will be able to skip it and you will be assigned a new project to review.
Please don't skip projects unless absolutely necessary. The system makes it really simple for you to submit votes, but please think hard before you vote.
Once you make a decision, you can't take it back.