Postmortem - Light Speed

Results are out! Being no. 16 isn't bad at all!

Still I have to be brutally honest with myself. This project was definitely overly ambitious in scope and I burnt myself out during the process. I think the crucial problem with this game: its a puzzle game but with many actions required by the player - to pick up objects, to jump and move around, to aim, to zoom and even do bullet time. The complex gameplay was hard to fit into a bite-sized game jam entry, and it definitely detracted from the theme of mirrors. I spent too much time developing useless features, such as the tractor beam mechanic. I even programmed a refraction mechanic in detail where if the beam was angled at the critical angle it would reflect instead. However these mechanics weren't explored fully in the 3 levels.

Looking at the winning entries, the mechanics were simple but highly polished to a tee, and were focused strongly on the core theme. I guess I should approach a game jam from the perspective of making a Mario Maker level - just big enough to explore a single gameplay mechanic, but not too big that it overwhelms with complexity. 

Also, I made a slip-up by not clearly introducing the laser gun's key mechanic of having a limited range. Players were getting stuck on the room right after the laser was introduced as they did not appreciate that the laser beam had a limited range, and it was difficult to convey this in first person view. Conventional wisdom would expect that a laser beam will go all the way to the end, so this was not very intuitive design too. I could have communicated this better. A playtest would have revealed this issue, so note to self: playtest early and often!

Being my first 3D game, it was a real challenge. Used a bunch of cheap ways to get away with less, by using UE4 primitive shapes on the various objects, so I was quite surprised to get a relatively high score on graphics given that I had only 2 assets from blender - the robot and the gun. I guess the postprocessing made it look really good. But level building in 3D - that was such a chore. It really requires an eye for 3D architecture. Also realized that puzzles in the first person view is really difficult to make fun if it's not about shooting and immersion. The primary game mechanic of lasers (and mirrors) was a challenge to implement in the first person view and adding the third person view didn't help much. 

My primary purpose in this game jam was to learn UE4. And there was just oh so much to learn.....

There's a lot to like about UE4, from the way lighting just works out of the box and postprocessing is able to make everything look really pretty, to how the material editor is just oh-so-awesome. Blueprints is also a snap to learn although it takes some getting used to with the context sensitive BPs. However the greatest downside became apparent during submission - exporting to linux and mac was really tedious and errors popped up, apparently because of some custom code (could've been the shaders), while Unity exporting to multiple platforms was a breeze. And I was travelling right after that with a more basic PC with no graphics card, reviewing the other entries - UE4 games were basically unplayable while the Unity games were still chugging on fine. Also, blueprints can get really annoying after a while and not wanting to delve into C++ just yet as it looks really intimidating. 

Still, I'm really happy with how some of the puzzles came about - in particular the puzzle with mirrors. It was just beautiful - to see that aiming at the mirrored object would result in the beam hitting the object as expected. Also, I just love the bullet time effect and in combination with the laser beam it just looks really pretty. And looking forward to trying out more UE4 with suitable games.

Onwards to the next game jam and my own games! (Thinking of a game with just refraction as a mechanic - now wouldn't that be cool?)