Hi,

Do you have any thoughts on using grid maps where humans take up 2x2 tiles instead of 1? And do you know about any good examples of it?

The kind of games I have in mind are tactical or single character RPGs(Either turn based or real time). It would probably be overkill for games with larger armies, but it might still provide some benefits.

It gives more options for weapon range. For example: Dagger - 1, Sword - 2 Spear - 3.

Diagonal movement and attacks would be much easier to handle too.

And you can add monsters that are 1.5 times instead of twice the size of the players.

I imagine having a human body fill up 1x2 tiles, but usually standing in the middle of an area of 2x2. She could still squeeze herself through a 1 tile wide crack, or walk on a ledge(without too much armor), but it would reduce movement speed and balance. And my favourite feature coming from that: It would be possible and look natural walking past people in a less than 2 man wide hallway or walk through a crowd. It might just be a source of unnecessary complecity.

Pros, Cons? How to implement it? Derived features?

Tile-by-tile. If by that you mean that it is impossible to stand between the tiles.

True. 2x1 would make things more complex. I imagine that you would design the game with a 2x2 character in mind, but implement 2x1 (If you can be bothered) for special occasions like ledges or hiding in the shadows of a wall. And if you don't mind people looking creepy when passing each other you can use the same animation for all three of those cases. Not my biggest priority though.

All the tile-by-tile tactics games I've seen(Not that many) use a tile the size of one character to measure range and movement. The attack range for melee characters always felt weird. It doesn't seem right to allow characters to attack diagonally, but when he can only attack in 4 directions it feels very limited. And the difference in reach between a dagger and a spear has to be implemented without using actual distance.

All that being said. Those games worked well and I can't imagine that making the tiles smaller would improve them much.

Is it even worth having small increments rather than sticking to a safe and simple 1-tile-one-character or even going all the way to a grid-less system?

All that being said. Those games worked well and I can't imagine that making the tiles smaller would improve them much.

Is it even worth having small increments rather than sticking to a safe and simple 1-tile-one-character or even going all the way to a grid-less system?

I don't think it's worth it unless you're going for more graphical realism than a tactical simulation. Mostly all it does it let characters look a little less lined up on the screen.

I'm not sure of the relevance of turning sideways to walk on a ledge; as far as game mechanics are important, either you can walk on it, or you can't. If you want to make ledges special, just make them a special tile-type.

Could possibly work nicely, but I think it depends partly depends on how you display it.

Part of the value of a grid in strategy game is that it allows some visual simplification (you can tell at a glance who's where and how far apart they are) and conceptual simplification (it limits the search space of strategies so you can look deeper rather than get caught on minutiae). For example, you can know whether your unit will fit between two other units, you don't have to "measure" or use trial-and-error. You can say, well, that unit moves 3 spaces, and this one has already moved, and they're five spaces apart, so there's going to be space for my unit to squeeze in no matter what. You don't have to wait for them to move and hope it works, you can know it in advance, and therefore think more moves ahead.

Having a grid size that's a fraction of the unit might support this kind of eyeballing and reasoning, or it might obscure it; it depends on the shape of the units, the perspective of the player, etc. If you've got an isometric view, and tall skinny units, and the units are different heights, I can imagine that looking at an unruly crowd of them it'll be difficult to know exactly where everyone is, who can reach each other, etc. On the other hand, if you've got blocky units and you're looking at them from above, then it'd be trivially easy to know these things.

So it depends. Consider X-COM, which is in the middle of these but works. It's isometric, and the 2x2 terror units are of different heights, but at the same time the terror units tend to be low-to-the-ground and are pretty beefy in the sense that they take up much of that 2x2x1 space. I don't wonder too often what space they're in -- and in any case I'm hopefully not engaging in melee combat with a bunch of them in close quarters.

It could work with 2x2, especially if you ensure every space is an even number of squares wide. Melee would work better in my opinion. As someone else said it would also allow smaller 1x1 units which would be cool. I would advise against 2x1 ledge situations, it would just make pathfinding confusing. You don't want to go to motion planning unless you really have to.

Maybe you could just make the character 1x1, collision-wise, but have the character's appearance be larger than 1x1, and make the animation state react to nearby walls (switching to shimmying along cliffs, squeezing into narrow spaces, moving through crowds). You hopefully wouldn't need to change your collision methods, just make your character's appearance more reactive to the surrounding environment.