When trying to clear a broken wall, we found that if you messed up once, you would often get stuck in a loop of hitting the wall until you died. Was this because the player resumed moving too soon after getting up? Or because the jump doesn’t “trigger” quickly enough after recovering? Are broken walls simply too hard to clear?
It turns out, the animator transitions were simply too rigid, leading to the awful results we see above. The player would try to go from “Jumping” to the default “Running” to “Death” to “Getting Up” over a very short period of time. By the time it cycled through all those transitions and states, the player character Helix was already moving again and the user has little time to react to jump in time.
Better transitions – more fluid and interruptible. Enabling the “Next State” to interrupt both Jump -> Run and Run -> Jump transitions with a new Jump -> Death transition leads to a result that is not only smooth visually, but also offers the player a few precious milliseconds of reaction time. This made a huge difference in how easy it was to recover from a failed jump, from nearly-impossible to relatively easy while still requiring a decent reaction time from the user.