A tiny bit of a media blackout going on because there isn’t much that is new to show off right now. So, here’s a non media update on what’s been going on with Turnover.
Level Planning
I’ve been doing a lot of thinking lately about where the second half of Turnover is headed and I think I have a pretty good idea. A level flow document is completed, so I have the setting and the story for the remaining levels. I then moved on to sketching out the layouts for them.
A level for Turnover usually goes through a few steps before it is “set”.
1. I sketch out a number of gameplay scenarios. These usually consist of a room or an area with enemy, item, and entity placements. These scenarios are a sort of puzzle. I try to think of a situation that the player needs to get out of, along with a few methods of doing so. This formulates the core gameplay.
2. I’ll go through my notebook of scenarios and choose a few that connect together well and are thematically similar. These scenarios will lay the foundation for a complete level.
3. I’ll create a playable, barebones layout using the selected scenarios. As I connect the scenarios together, I’ll account for scale, feel, and how well they work together. Sometimes I’ll chuck out a scenarios and select another one, but often time things come together fairly easily.
4. I’ll play and revise. Since I spent most of my time thinking about these layouts, this step goes pretty quick. I’m just trying to get a tangible approximation of what I sketched working. Once I’m satisfied, I’ll play through it a few times from start to finish, trying to fix any adaptation problems.
5. I’ll play test it. I’m lucky enough to have a family member who tests everything that I make. He’ll often play through it a few times and give his feedback. During these play test sessions, he often finds engine and game play bugs that need to be addressed. This outside feedback is an absolutely crucial step.
6. I’ll play through it one more time. After my tester and I have discussed (argued) about what’s best for the level (who’s right and who’s wrong), I’ll give it one last play through. At this point, the level becomes gameplay complete.
7. I’ll move on to another level. At this phase, I’ll begin on another level and I’ll begin decorating the one I just finished. Aside from the engine work, level creation is by far the most involved and time consuming part of working on Turnover.
What has been interesting is that each level is fluid. New enemies, items, ways of playing end up emerging as new levels are made, so I’ll often revisit a “completed” level, only to revise it further. All in all, this workflow has served me well.
Profiling
I had a little idle time lately, so I chucked Turnover through a compiler to see what kind of horrifying discoveries I would encounter. To my delight, things seem OK.
I spent a few hours doing miscellaneous optimizations, and based on my benchmarks, I saw a 13.7% increase in performance on the CPU side (thinking) and 2% increase on the GPU (drawing). For a couple hour’s work on a quick profile, I can’t complain. The game does run great right now, but there’s still some room for improvement.
