Let’s recap:

I’m Falkreon. I’m a prolific Minecraft/Bukkit/Forge/Spout modder and I also develop games as standalone cross-platform JARs.

My goal this time around is to shade in a lot more atmosphere and polish than my previous entries, and continue upping the WTF factor by abandoning or eliding features people generally expect in a genre. Every Ludum Dare, get about ten games I can’t live without, so I definitely agree with the keynote, it’s kind of like getting a hundred xmas presents at the end.

Behind the cut: I relate some of my LD history and share some wisdom received over the course of the production.

My first Ludum Dare was a painful failure. I didn’t know how to kill my babies, and the project was just completely untenable as envisioned. Editing is really important in game design. Ever since then I’ve had a particular approach to my LD design:

• During the final voting round, I rearrange the top themes into a series of boxes based on similarity of tone or theme. This helps me both pare down the number of early-stage designs, and stay organized and straightforward about my design process.
• For each box, I write a couple words to a sentence worth of design. Don’t write too much at this stage; just let the possibilities swirl in your head. I usually like all of the design possibilities, but I can only do one; the final theme selection is a sort of useful design roulette. I don’t feel bad about abandoning the other designs because I forced myself not to write too much, and only the particular theme box’s design will work on the selected theme.
• When the theme comes out, I reserve some of the precious competition time to write more about the design, and think more about the design. This is extremely important, building a game that fits the scope of the time limit, and to some degree you’ve just got to get a feel for it. I don’t want to put an exact number on it, but in practice I usually take a good ten or fifteen minutes just to make sure the design is extremely clever.
• Most importantly: Once I have a design you’re 99% confident about, I kill my darlings. I take the idea and ask myself: What are all the various things I need to do to make this happen? Then I rank them in order of importance and for time management. During this process I usually find that some of the clever features fall like a rock to the bottom of the list: They’re amazing features, I love them, but they’re totally unnecessary to the design. I doubt I’ll get to them in time, and maybe the design will actually be better with them streamlined out.

My second Ludum Dare, Erased (for #22, the Alone theme) was a tenative success: I released a product. I spent the next month refining it, and it was much more playable in the post-LD version. Even though there was much more that could have been done, at some point I finally admitted that nobody wanted a roguelike that wasn’t rogue-like, and it wasn’t worth polishing. That’s kind of a corollary to “kill your darlings.” Sometimes code is only digestible if you cook it on a rock and then eat the rock instead. Well-meaning programmers can create that code. Don’t feel bad about a failed project, and especially don’t feel bad about the risks you took in the design. Code is cheap if you write it often. Take new and different risks, put in a more thoughtful and skillful effort, and you’ll find that eventually, the risks pay off.

My third Ludum Dare, mini-LD #37, I’m calling a huge success. Little Stories was *exactly* what I envisioned. I removed game mechanics, control systems, and a lot of planned content never made it, but the design wasn’t something I expected to be feature-rich. It wasn’t a platformer, even though it looks like a platformer. For one thing, you can’t jump. You don’t kill anything. The whole goal was different. Not having a template was freeing to me, and suddenly I didn’t have to include something because games of genre X just always have Y. So my final piece of advice is, even if you have a genre, keep cutting. Kill their darlings too.

So, in review:
Kill your darlings,

Kill your darlings,

Kill their darlings too.

All the darlings.