Judge feedback

Judge feedback is anonymous and shown in a random order.

• All I can really see of the project is of the GIF where the node is dragged around, so difficult to give an accurate opinion of the project. From what I see of the project, it looks like it draws what the path will be in the Unity editor, which is very cool, other than that I just see a lot of nodes in screenshots. Not much more to say about it without trying it out myself
• Quite hard to judge without actually being able to download and demo it myself. The gifs do show that you implemented pathfinding but I can't really look at what scale and level complexity it works up to while maintaining a solid framerate.
• The lack of presentation made this project extremely hard to assess (1 gif and some images isn't sufficient). Large scale wasn't well demonstrated, as distances were unclear, and there's only ever one agent. Performance wasn't demonstrated either. A playable demo would have helped substantially.
• Very light presentation... Large terrain: how large? 1ms per agent: how many agents in the scene? on what kind of hardware? Has your implementation something more special than the standard HPA? How did you apply it in your context? What was the most challenging technical problem?
• - Hard to assess; very little presentation, no technical design documentation explaining approach or comparisons - Benchmarking, video demonstration, dev diaries, executable with test cases or examples would aid presentation and visability
• A nice pathfind project. I bet you had a lot of fun researching and implementing all of the differenet pathfinding techniques - what made you pick this one? It would also be awesome to see what you thought of the other effects and why you didn't choose them. Were they too expensive in terms of processing time/memory? Did they require too much work to setup, where the results too robotic, etc? I like the little video demonstration - consider adding some voice over to talk us through the idea and the project :)
• "The final version of my implementation of the algorithm allows for hundreds of agents to find near-optimal paths over exceptionally large distances with sub-1ms calculation per agent" - I would have liked the material to show that and go into more detail on how that was achieved. The path itself doesn't always look great / natural, seems to hug corners.
• This project seems impressive but is lacking the right documentation to show off how good this pathfinding algorithm is. It sounds efficient but I want to see some sort of demonstration that shows off the pathfinding. A small maze made with some cubes showing it can calculate around obstacles doesn't feel like enough of a demonstration of what it can do. Have a demo project showing it can work in a build of unity have lots of AIs traverse over a large scale terrain. If this algorithm doesn't require any building in the editor and can work in runtime of the project show that it can work with level streamed/loaded scene with the terrain. Overall this project seems impressive but the presentation is lacking the right presentation needed for a tool/feature.