Game Feel
As mentioned in an earlier devlog post, I’m now starting to work on game feel – specifically the movement of walking up a wall. This is what the first version of the movement looks like in the editor:
And here it is in the game:
This is what I wrote back in early 2013 when I wrote the first prototype. It was sufficient for the purpose of designing puzzles, so I basically left it alone until recently.
There are two problems:
1) It feels very clunky – Notice how the player actually stops right after rotation. This happens even if you’re holding down the joystick button or pressing ‘W’ to go forward. The reason why it happens from a technical standpoint is that during rotation, the code actually takes control away from the player, and rotates the player object. After that, force needs to be reapplied to get the player moving again. Having the player stop and go prevents the movement from feeling like ‘flow’.
2) You can’t look around during rotation – because control is taken away during rotation, none the player’s input has an effect. I’ve had several playtesters telling me they wish they could look around during rotation, and also having the game control the camera during rotation makes the experience very jarring.
After the conversation with my friend Kyle, I decided it was time to start looking at how I can improve this movement.
I’ve never played the game Prey, but in my research for first-person games in which players can walk up walls, Prey was one of the few results that turned up. The mechanic is quite different – you can’t just walk up any wall, only along certain paths. I won’t go into the details, except to show you the movement in action:
I’ve actually never played Prey before, but look at it how smooth that movement is!
I decided to try something similar by using a quarter pipe, and having the player walk up it to transition between walls:
Of course, the quarter pipe wouldn’t be visible in the game.
This looks and feels pretty good. But there is a problem, which is that I can’t always afford to have an invisible quarter pipe that big. Right now, that quarter pipe is 2 units tall and 2 units deep. A lot of my geometry is much shorter than that, and the player still needs to be able to walk onto those. 
Some issues:
1) The player doesn’t begin to feel the rotation until much closer to the wall – visually it doesn’t look as good to have the entire screen fill up with a view of the empty wall.
2) Because the radius of the pipe is smaller, the rotation speed is faster – maybe a little too fast. When the quarter pipe is larger, it feels a little better. This could be fixed by decreasing the speed during rotation.
However, another issue with the quarter pipe is that the boxes are now too small to step on:
Look at that, the player just goes right over the box, as there isn’t enough room for the player to fully rotate. I could make the pipe smaller, but that was make the above problems worse, and there is a lower limit to the size of the pipe. If it’s too small, then it’s not walkable at all.
The solution is to make the box bigger.
Currently, the boxes are 1 x 1 x 1. I could make them 1.5 x 1.5 x 1.5. Of course, this would have repercussions elsewhere with regards to design. Maybe not such a big issue though.
With a bigger box, you can see that it works a little better:
Would definitely need some tweaking to get it right.
However, it’s still nowhere near as precise as the previous method of rotation, in which the player would land directly above the box:
Alright, so I still have yet to find the perfect solution. Seems like the methods I’m using are a compromise between precision and feel-goodness.
I’m glad to be finally diving into this aspect of the game though. Looking forward to it evolve over the course of the next few weeks.