Who should be your judges?
The following are all groups you can select judges from, though it is recommended to sample from each pool available to you.
- Sponsors (They may also be judging to give away sponsor prizes)
- University Faculty
When selecting judges you should aim to have enough judges to cover all of your hacks in a reasonable timeframe. Bias should be avoided if possible.
How to structure judging?
Allocate adequate time given your number of judges and attendees to allow the judges adequate time to consider each hack they need to judge.
This is the simplist approach to judging, every team presents their hack to all attendees and judges with a small, 3-5 minute, slot. Judges will see all hacks but this can become incredibly time consuming for large hackathons.
Here all the hacks set up in effect a stall to show off their hack, judges will visit a subset of all hacks and report back to select a top "n" that then present their hacks to the entire hackathon. Benefits from immense parallelisation but can create unruly and biased judging reports if the judges marking subset A are all marking higher than those marking subset B. Gavel can alliviate this by using comparative judging. In any case how to organise the judge's routes should be considered with simple assignments based off of Devpost being available here.
How should you score demos?
It's a good idea to decide your judging criteria before the event. CovHack has shared their judging criteria for CovHack2020 here!