Last time around, I had prepared a map for the Seattle Indies playtest night. It was incredibly fun! Having a bunch of people, some I knew and some I met for the first time, play my game was a wonderful experience. So what did I learn?
Well, first of all, nobody played the map the way I had envisioned. I had conceived of players standing behind the barrier and shooting the targets one after another, and then maybe throwing around the targets afterwards. Instead, the moment the door opened, nearly every single person leeroyed right into the pit in the center of the map. That center pit has two passageways hidden behind buttons, one that takes you back to the beginning of the map and the other takes you to the last room. The passages were only there for my own convenience… but of course others found them.
One player went to the last room and tried to shoot the targets before they were active. Another player got excited from finding one of the passages and then was disappointed when it just went back to the start of the map. On one level, it was a reminder that players who had never seen the map didn’t carry the preconceptions I had after staring at it in TrenchBroom. On another level, it was interesting that players had a strong desire to explore the map, even if it was a bland set of rooms with purple prototype textures. This was the part of the playtest I enjoyed the most: seeing everyone play the map in a different way, and none of them doing what I intended.
Another thing I noticed was that players took some time to understand the interaction model. Quite a few people attempted shooting the button to engage it, others jumped on top of it. I anticipated this, especially since it turned out most folks who tried it hadn’t played Half-Life, so it’s good to know that the planned tutorial messages would put me on the right track. A friend who had played Half-Life had tried to air-strafe at the end of the playtest. That build didn’t quite nail air acceleration, so it wasn’t successful. The latest build should behave much more like Half-Life’s physics, bunny hopping and all.
After the playtest, I got some great feedback on what the map was meant to test: the revolver. Much of it was what I had anticipated: it lacked impact. Some suggested more bass to the sound, others suggested muzzle flash, another suggested some kickback for the camera. All of them are great ideas that I’ll get around to doing once I come back to more asset work.
I also found it interesting that no one attempted to interact with the targets once they were shot down. Those that tried it had to be told it was an option. Granted, there was no reason to, but I still found it interesting.
In general, I learned a few major things:
Since the playtest, I’ve been working on a tutorial level and a new test level. The tutorial isn’t anything special, it just sets up a series of obstacles the player must clear to progress, each one teaching the player how to interact with the world (e.g. jumping, moving physics objects, shooting targets). Once the player clears the tutorial, they’ll find themselves in a new test map that I’ve been building.
This test map is going to lean less on the literal shooting gallery I had built, and instead lean more into physics interactions and exploration. I won’t be showing all of what I’ve built just yet, but I did want to show the latest thing I’ve whipped up:
I’ve stolen Half-Life 2’s seesaw puzzle. It was an opportunity to expand the physics capabilities of my engine, primarily by leveraging hinges. It also gave me the opportunity to overhaul the collision system, as the precomputed collision hulls of Quake and the original Half-Life don’t handle the types of runtime rotations physics puzzles were going to produce.
I won’t be able to go to the playtest night this month, but I’m hoping to have a much improved game next time around!