Glad you enjoyed 'em.

I've posted the third part of my SMB 3 articles if anyone's interested: http://www.significant-bits.com/super-mario-bros-3-level-design-lessons-part-3

This one veers away from typical level design to focus on how the overworld hubs are structured and linked to the core stages, but I think it's still relevant to the topic.

In case anyone's interested, I posted a follow-up to the SMB3 article that goes over the remaining worlds: http://www.significant-bits.com/super-mario-bros-3-level-design-lessons-part-2

Deconstructing SMB 3 is a huge task, but if anyone's interested, here's my first stab at it: http://www.significant-bits.com/super-mario-bros-3-level-design-lessons

I've never heard of Future Tactics, but it seems pretty neat. We actually wanted to implement a bunch of its features, but only a few (such as destructible objects) will make it in as we had to pair-down the scope of the game.

Like some people mentioned, I think it's mainly an issue of "bang for the buck."

Side-view games show visually more interesting environments while requiring fewer assets -- left/right character poses can be flipped (without worrying about up and down), Flash-esque bodypart animations work much better, it's easier to fill out the scene with 2D props while utilizing effects such as parallax, etc.

Of course it also depends on the type of game you want to make. Platformers work well with a side-view, whereas top-down games usually avoid complex jump mechanics due to the skewed/limited perspective.

Thanks for the compliments!

Last time I took a stab at it I played through all of SMB3 and captured a bunch of screenshots as visual aids, but when I returned a week later to write the actual post, I forgot what a lot of the screenshots represented. I'll probably do the same exercise again, except this time with a notepad in hand.

I wrote the above article (What Made 2D Platformers So Great), and I'm actually doing a post about some some of the level design lessons gleamed from SMB3. It's a little time consuming, but it should be worth it as the platforming articles always prove popular.

Progressively shortening animation times for effects might be a bit tricky, but it's a neat idea.

I'm not a huge fan of gaining new powers that take forever to play out either. I recently checked out FFIV DS, and its battles were so much slower (not even counting the summons) than the original's. Combat in Tactics games is already pretty lengthy, so our goal is to make the battles a bit more snappy all-around.

I always did want to check out the environment-changing aspects of Suikoden Tactics, but for whatever reason I always assumed they were in a Disgaea game I just never played. It might be a bit hard to come by these days, but if I ever spot in a store I'll probably pick it up (along with Ring of Red).

No worries, I was just curious about what everyone enjoyed. Thanks again for the feedback.

We decided to go with a unit-by-unit approach for the reason you mentioned, i.e., it seems a bit more believable, and because it prevents the two armies from slowly inching towards each other in order to gain the advantage. This is somewhat combated by most units being able to counter-attack, but I wanted to avoid that mechanic (at least for most units).

As for the animations, I do agree that they add a lot to the atmosphere, and I thought that they were particularly great in Shining Force. With that said, these animations seem to have only gotten longer with time, with no real gameplay additions. They're still entirely passive, and after seeing them the first couple of times, I tend to turn 'em off/skip 'em (like I do with Final Fantasy summons).

Since we're dealing with less technical limitations (resolution, memory, processing, etc.), I'd rather implement the unit actions in the map instead of a separate-state screen and avoid that board-game piece vibe.

In a way, don't those stats give more information and eliminate some of the risk, though?

As for the example of ranged enemies attacking weaker units, that should be something you could predict provided you checked everyone's ranges. Any randomness would come from the enemies deciding to go after a different target, which is actually something we'd like to try out.

On one hand, you can go for a FFT approach that feels like a loose team-deathmatch system where each enemy can have its own unique behaviour. On the other, you have Advance Wars where all the units coordinate and exploit your weaknesses as if they were really led by a single general.

I'm personally leaning towards the first approach as it makes for a less predictable game and allows all the units to have a bit more personality via their actions, but which one do you guys prefer?

• In order to keep the pacing brisk, units will go down fairly quickly provided their weaknesses are exploited.
• Like I mentioned in my points, the ranges of your enemies will be indicated as you move a unit a la Valkyrie Profile for the DS, so you won't even need to do any manual checks.
• We'll probably be avoiding building cover and other components to keep things moving quickly while having one less variable to balance.
• Action Points are fine, but a rather abstract concept, i.e., how does running 30 yards take less time/action points than squeezing a trigger? Not that Tactics games are a very realistic portrayal of battles, but I think a move+single-action approach is easier to grasp.
  

Thanks for the link -- I did a quick search for Tactics before posting, but I didn't see that thread.

As for the other comments, yes, we'll be doing lots of work to make sure the units are very different from one another. The element of risk is a bit trickier. On one hand, you want the ability to predict and outsmart the enemy AI, but on another you don't want every battle to play out the same way or think 20 moves ahead to be successful. To combat that, we're making our battles quite short and going to experiment with a bit of randomization.