Hide and Freeze VR is a multiplayer game that based on the common children’s game mixed with the possibilities of interaction and space in VR. With open spatial audio and some additional rules, it creates a competitive and co-operative social game.
See project page on the website of Improvive
Hide and Freeze was a project accidently discovered while testing out the Unity Networking system in a different VR game from ImproVive. While something went wrong during testing, I started playing Hide and Seek with the lead game designer, Nikita Kayal. Turns out, quite a fun principle. So, Hide and Freeze was born.
Hide And Freeze
Screenshots
| Team Size | Engine | Platform | Role |
| 4+ | Unity 2017 | SteamVR, Oculus | Intern Game Programmer |
Personal Contribution
Migration from Unity Networking to Photon Engine
The development of Hide and Freeze began with a base structure created by previous interns at ImproVive, using Unity Networking. However, during multiplayer testing, we quickly ran into issues with stability. Since VR requires transmitting a large amount of data to synchronize controller positions and headset location with characters in the game, packet loss caused major problems. Even a few missed packets led to characters jittering or snapping back into place unnaturally, which broke immersion.
To solve this, we explored alternative networking solutions and eventually decided on Photon Engine. My task was to rework the existing code so that we could preserve most of the features we had already developed, but rebuild the networking layer using Photon. This was a significant challenge, as the project’s codebase had grown into a bit of a spaghetti mess. And on top of that, playtesting requires multiple people, so this would require every build that was playtest ready to contain as few points of error as possible, to avoid wasting colleagues time.
Despite those problems, I managed to untangle and restructure much of the networking code, integrate Photon, and build a stable system that supported smooth character movement across all players. With these improvements, we were able to run multiplayer sessions reliably with 4–6 players in VR.