Compared to previous LDs, I didn't go for any crazy or complicated mechanics but rather settled for something simple and clone-ish (Space Invaders). The idea was I could spend more time on polish and less on the complicated mechanics and squashing the bugs in it.

https://ldjam.com/events/ludum-dare/38/small-world-attackers

The Good

I went for a 3D version of a space invaders type of gameplay. It was cool but not necessarily exciting. However, when I need to make the mothership flyby feature I discovered that was actually nicer to play because you could use the third person/first person switch mechanic pretty well. So I made that the primary game, and added a "classic mode". The lesson is: Thanks to not spending huge amounts of time on mechanics code, I was able to playtest more, find a better use of what I had, but still kept the original. More time means more flexibility.

The Not-So-Good

Even though I was going to have more time for polish, I used some of it on playtesting and adding the other mechanic. I also was getting a bit sick so actually went to bed early (11pm) and got up later than usual. My Mac also broken down - I usually use it for music in Garageband. I have a Ableton Live "Lite" version but am not very familiar with it. It worked fairly well after some messing around, but I didn't get the vibe so the music is sufficient but not that good I think. As far as polish I forced myself to do as much as I went, and not wait to see how much time was left in the end. It's still programmer art in a way, and I used the arcade setting as an excuse to keep it simple.

The Bad?

Don't know if there was anything really bad. I guess I could have tweaker the gameplay a little more but I'm fairly happy with the result.

https://ldjam.com/events/ludum-dare/38/small-world-attackers