I find it interesting how production factors and game design interact. I was going to say I didn’t mean to talk about cost effectiveness but about creating different possibilities, but then I realized these are really two sides of the same coin.

I should probably play a UE game before saying more about it. You are the second person to speak highly of it. Personally, I am hoping they succeed in making it easy to get into…

There are various reasons why 2D can be the better choice. When I was at JoWooD Vienna and we were asked to produce several smaller titles in one year, it made more sense for us to stay with 2D, as we had a lot of experience there. We made games that worked well in 2D, and they turned out very nice. We had a shared engine, so that kept down costs. The games ran on a wide range of PCs. In the case of ‘Neighbors From Hell’ we would not have been able to do that in real-time 3D in that time and for that budget. However, we would have been able to make a different game in 3D, even with a 2D camera.

I tried to list some concrete examples where we did things in 2D that were hard or problematic, and which would have been much easier in 3D, or - and this is really just saying the same thing - we decided not to do certain things because we were working in 2D. I don’t think teams make the exact decision you describe: I think they try to make games that fit their technology, basically.

I don’t know if 3D games always take longer to produce than equivalent 2D games. Perhaps you are right, although I would insist on comparing teams with equivalent experience in either approach. My point is that it doesn’t make sense to make an equivalent game, but that you should make your game design fit your capabilities. Typically, your team’s capabilities are much harder to change than your game. In my experience, going from 2D to 3D is a high-level, long-term strategic choice, more so than which game you are going to make. You need to buy new tools, license engines, train pretty much your whole team, etc.

In the case of UE3, they might get the ’sexiness’ value of 3D, some additional freedoms, some production optimizations and a more beautiful game. (Note that I said ‘might’ and ‘in the case of UE3′ - these are not general statements about 3D vs 2D.)

The most labor-intensive art tasks these days are 2D painting (textures or sprites) and 3D character animation. If your game has neither of those, 3D is probably faster and easier. Most top-down and side-scrolling indie shoot-em-ups (shmups) make these days use 3D because of this. Introversion’s Darwinia also has minimal character animation and fairly stylized 2D art.

On the other hand, 3D engines generally take a lot more time and skill to make (Darwinia took ~3 years, a lot of which was tech development) than 2D engines.

If you like 3D games with 2D gameplay, I highly recommend the Klonoa games for PS1 and PS2 respectively. In gameplay terms they’re similar to Yoshi’s Island, with neat quirky environments.