Sprint 4 Blog
This sprint was a nice return to form where I got to once again work on level design work while still dabbling in interdisciplinary roles. This was also our first official playtest, giving us plenty of work to do.
This 7 point sprint for me was actually quite a lot of work that just barely didn't push its way into being full of 3 point cards. This came from a combination of designing annotated maps for future levels, modeling an extra piece of environmental assets, setting up animations, and the biggest, most time consuming aspect, covering the recorded gameplay footage of our play testers to do some rebalancing of the first level. Let's take a look.
Starting off with the modeled asset I added we have a chair to pair up with the desk I created in the last sprint. This chair was initially designed to be an LOD model that would allow us to maintain solid frame rate while giving the player a clear detail when they got in close with the zoom. While this worked in theory, the player can never actually get close enough to the chair to see the finer levels of detail that make it being a LOD model worth it. The high poly model was cut as a result and we wound up only using the low poly model. While this felt wasteful, there was no point in putting in extra polygons in the game that weren't discernable anyways.
Moving onto the textures for it the chair was originally much darker on the felt parts of the chair. This wound up not working great due to our darker camera settings hiding a large amount of detail. The plastic and the felt textures wound up looking the exact same. In order to fix this I went through 4 iterations of lighter and lighter fabric until I got this much, much lighter fabric color.
Finally the chair model looked a little too tall in unity when compared with the desk. While it had appeared to be fine in its Maya scene, it was noticeably larger when given the camera perspective we use in Nightwatch. I scaled down the Y scale to 80% and it looked perfect in game.
When we started discussing animations in game, we realized that our modeler did not have any experience with rigging or animation. This presented a small problem as I took it upon myself as the producer for the game to teach them. After we noticed there were issues with the rigging and animations coming through his branch on our github, we decided it was best for me to take care of it as we had used my branch as a test and found it working perfectly. We used Mixamo's rigging tool to create the rigs quickly and test out iterations of the guard and thief while it was in progress to make sure that it would look alright in the scene.
For the guard it was imperative that the guard have at least 4 animations: An idle for when they weren't moving around anywhere, a walking animation for when they were patrolling or moving to a player selected area, a run for when they were in pursuit of a thief, and a door opening animation for when they went to open the doors. Additionally, I decided to throw in a little dance for if the guards managed to apprehend every single thief on a given night to give the game a little bit more of the fun we try to chase as designers.
The thief animations needed to feel a lot sneakier and panicky than the guard's. Due to the functionality we see from the thief, an idle animation was actually not needed at all. Instead, I used a sneak animation for the thief's default animation, a running animation where the thief is looking back for when the thief is being pursued by a guard, a door opening animation in the event that the thief is spotted while opening a door by a camera, and an animation where the thief looks like they are working with a tablet to hack into cameras, sensors, and computers.
When looking at gameplay footage I noticed that the players were following my general design decisions I was making for the levels in some "optimal" set ups. The camera had a wider sense of vision and a shorter reach where the laser sensor had a wall to wall reach but provides nothing for a thief can walk just outside of its center. This lead to the players often (but not always) deciding to use the laser sensor on long, narrow corridors where as they would use cameras on intersections to see exactly how thieves moved around these pathways. This helped me immensely in seeing that my level design was easily intuited by the player, showing me that I could more or less continue with the same design decisions I had been using for the annotated maps in the future.
With that said, the players often relied more on cameras than laser sensors, leading me to adjust a couple of sections of the building to have even longer stretches of uninterrupted pathways for the laser sensor to have clear spots of power where it will outshine the camera in what it can accomplish per cost.
Moving onto the annotated maps for the upcoming levels, I wanted to create a combination of clean, grid like sections and more "round" sections like seen on the bottom right. This should help to create some more interesting choices for the player as they need to adjust to several very different sections of the building. Blocking off the extra areas (as seen by the X's) will set out to help with our night by night scaling of the building to help prevent the player from being completely overwhelmed on the first night.
The second annotated map set for this sprint is a 2 parter. This level is set up to introduce the player to multi-floored levels and help force the player to multi-task a little bit better while spreading their defenses even thinner. The top map acts as the first floor, split into two parts to help with the issue of the player having too much ground to cover with too little money to work with. The second map is the entire top floor, fully available to the player all at once. The design is the same for both floors intentionally as I want to take this as a chance to experiment with building design and symmetry. I think I would have liked for there to be a bit more difference between the two maps as the only real difference is where the stealing points are and how much needs to be covered at once. I expect to make these changes in the block out when I have time to see how it looks and feels in game itself.
