Incline Hexes or Squares?

[Matticus]

My studio is going to work on a turn-based isometric game next and I was thinking hexes but gridless sounds good too

 

[I'm Big Into Anal]

very interesting read this thread

opened it up expecting the usual codexian blatherings and shitposting but refreshingly found a reasonable discussion

bravo

 

[vonAchdorf]

One disadvantage of grid less TB is that it's less obvious what is blocking moves and what is not (Can I squeeze past these two obstacles / mobs / characters - will my tank block this doorway?).

 

[Ocelot]

Hexagon is the most efficient shape in nature, so they must be superior compared to squares.

 

[ushas]

Truncated octahedrons, for 3d movement of course (in a mathematical way, they are the true hexagons in space). Some day I will make a game around this. BTW, they look like octagons from the top, and moving into the smaller squares means also moving up or down half a level. And also, they are the voronoi cells of the bcc grid.

Heh, also briefly thought about the Voronoi diagram - adapt the grid to any dungeon needs

Spoiler

But you both do know, that depended on the on the dots (vertices) distribution and the dimension the result can be a square or a hex grid or something else? What you both correctly described is that you can use Voronoi partition growth to a higher dimensional room than R2 or R3. Btw. Truncated octahedron (you aim here at the bitruncated cubic space-filling tessellation) cells have two different sides and therefore are do not regular, but at least they are uniform.

That's true, but at least the "diagonal" movement distances here are not that different from the linear ones (sqrt(3)/2 vs 1) so they can be approximated as being the same (0.866 ≈ 1) for a game. Then you get 14 roughly equal neighbors instead of having 6 direct, 12 2d diagonal and 8 3d diagonal neighbors in a 3d cube grid. One disadvantage is mapping, however, you can't easily use a piece of grid paper anymore. Also, imagine moving in such a voronoi-blobber - how would be the control scheme?

By mouse selectable cells? Looking forward to your game.

Not deadly serious with this, but in general can imagine all sort of things. One can limit cell shapes or sizes, anything from Delaunay triangles up to what not -gons. Yeah, it can also lead to just fitting usual square or hex grids on the map mask, eventually. Dunno, if adaptible mesh has much use as a grid. Am thinking maybe it allows to have different take on moving through various terrain. Instead of costing different TUs to move through terrain of variously difficulty, the cells/hexes have different sizes. Distance definition takes into account the walk-time in here. Which could be also used to represent vertical movement on topdown 2d grid, a sort of like when elevation contours are more nearby...

Yeah, gridless-like is also cool.
Tbh. when the game mechanics are build around it, then anything works well. See chess, for example.

 

[DraQ]

One disadvantage of grid less TB is that it's less obvious what is blocking moves and what is not (Can I squeeze past these two obstacles / mobs / characters - will my tank block this doorway?).

That's what you get for going with "rigid blocks that physically block the path" model.

It's better to assume that solid core of your character is surrounded by non-solid halo of their reach, which doesn't prevent friendly movement or own movement by obstacles, but is used for actual blocking. Then you can block passages by overlapping weapon ranges with obstacles and each other, without any need for being pixel exact.

Of course with gridless you do need some extra UI aids:

• Ability to queue any reasonable sequence of moves not limited to current turn (to make sure that your brilliant plan doesn't require you to go a pixel further than you actually can) - reasonable means just short of "fuck, if we do that we'll run out of RAM"
• Relevant ranges and passabilities overlaid (possibly on demand) on both current and planned/expected tactical situations
• Unambiguous presentation when movement crosses into range or hazard, and when ranges and hazards directly cross into obstacles.

Otoh you get to move and reposition arbitrarily and have a lot of subtle mechanics involving reach and movement.


As for voronoi cells, the only reason for using them is when you want location/feature rather than grid based positioning. For example "stairs", "NE alcove", "SE alcove" and "that suspicious looking platform in the middle" all getting their own discrete tiles, with stretches of homogeneous terrain too big to unambigouously describe positioning being broken into more conventional grid.
This is going to be coarse as fuck (blergh) and require flexible and opaque movement and action cost mechanics, but may yield itself to textual descriptions and situational modelling quite well.

 

[Matticus]

I think with UI indicators for showing character block radius and consistently sized passageways gridless could be pulled off in an intuitive way. One could even go so far as to show whether a passageway was blocked with an additional indicator, if that was deemed necessary.

I still prefer hex tiles but entertaining an alternative seems like a good idea.

 

[J1M]

Gridless.

Every gridless game I've played, including the amazing Temple of Elemental Evil, has resulted in a pixel hunt for positioning/aiming becoming the primary interaction. That's not what players should spend most of their time doing. If you have a counter-example, please provide it and why you think so many other games have failed to solve this flaw.

There is real and tangible value to the certainty, readability, and user-interaction benefits that a grid offers with regards to positioning, spacing, reach, etc.

 

[Matticus]

Gridless.

Every gridless game I've played, including the amazing Temple of Elemental Evil, has resulted in a pixel hunt for positioning/aiming becoming the primary interaction. That's not what players should spend most of their time doing. If you have a counter-example, please provide it and why you think so many other games have failed to solve this flaw.

There is real and tangible value to the certainty, readability, and user-interaction benefits that a grid offers with regards to positioning, spacing, reach, etc.

I agree with those points. The interesting thing to me about gridless would be expanding possibilities for the player if they want it -- and I like the idea of players being able to find things that "weren't meant to be there". At the same time, like you said, it might just be too cumbersome to use the majority of the time. I recently installed ToEE and circle of 8 modpack and will be playing it this weekend for research to see for myself.

 

[mihai]

Imagine gridless chess. In realtime. With float-valued hitpoints for each piece. For me, it fundamentally changes the type of game from a tactical evaluation of distinct possibilities to something highly intuition-based, because it gets extremely complex. In RTwP, this often results in trying to slice up the problem into millisecond- pixel-sized fragments that are easier to understand.

I like a high degree of granularity. In space (grids), time (TB or phased), actor state (low scale of integer hitpoints and such) and actions (not too many similar spells, for example). Because it produces a feeling of control. And because the AI gets a chance to act intelligently by sampling a larger section of possibility-space.

 

[Cael]

Imagine gridless chess. In realtime. With float-valued hitpoints for each piece. For me, it fundamentally changes the type of game from a tactical evaluation of distinct possibilities to something highly intuition-based, because it gets extremely complex. In RTwP, this often results in trying to slice up the problem into millisecond- pixel-sized fragments that are easier to understand.

I like a high degree of granularity. In space (grids), time (TB or phased), actor state (low scale of integer hitpoints and such) and actions (not too many similar spells, for example). Because it produces a feeling of control. And because the AI gets a chance to act intelligently by sampling a larger section of possibility-space.

That is fundamentally the problem that I find with RT vs TB games. It is extremely difficult to minimise losses in RTS and you fundamentally have to accept that you will lose units simply because you can't keep track of everything. TB, however, allows for precise placement of units that ensures maximum damage for minimum risk. On top of that, a grid system ensures that your placement is precise whereas a gridless system like FOT means that you can accidentally place the unit in the wrong place and end up with a dead unit. For example, you move the mouse a little too far and the bloody unit runs around the damned wall instead of running right up to it and stopping which means everyone and their mothers will open fire on the suddenly exposed unit. Opps! Although, I must admit that it is freaking hilarious when supermutants are involved as you can almost guarantee that you get multiple machineguns going off and rockets exploding everywhere.

 

[oldmanpaco]

This is the best insight I have found on this debate.

 

[deuxhero]

Triangles

https://youtu.be/HQoRXhS7vlU

Seriously though, what is the problem with right triangles if you don't have facing in your system?

 

[Norfleet]

Hexes tend to provide you with nothing but grief when you're dealing with ranged weaponry on units with a turning cost. Just TRY and advance a firing line down the crabwise direction.

the sane answer really is "gridless".

Clearly, the best answer is "gridless with a grid". Provide a grid so that the player has a means of keeping things sane and controlled, but do not force the player to adhere to the grid, much like off-grid Sims furniture placement. By and large, the game can be played without leaving the grid as the grid is sufficient to cover the majority of usage cases, but if you come up with something really esoteric, you can go off the grid.

 

[JarlFrank]

JarlFrank

Is the hex grid image you posted a copyright? I want to steal it for my tabletop.

Steal how? If you wanna print it out for personal use, be my guest.

If you're designing a tabletop system and wanna include it in the manual or something, it's copyrighted to us, yeah, but we could probably work something out.

 

[JarlFrank]

Imagine gridless chess. In realtime. With float-valued hitpoints for each piece. For me, it fundamentally changes the type of game from a tactical evaluation of distinct possibilities to something highly intuition-based, because it gets extremely complex. In RTwP, this often results in trying to slice up the problem into millisecond- pixel-sized fragments that are easier to understand.

I like a high degree of granularity. In space (grids), time (TB or phased), actor state (low scale of integer hitpoints and such) and actions (not too many similar spells, for example). Because it produces a feeling of control. And because the AI gets a chance to act intelligently by sampling a larger section of possibility-space.

That is fundamentally the problem that I find with RT vs TB games. It is extremely difficult to minimise losses in RTS and you fundamentally have to accept that you will lose units simply because you can't keep track of everything. TB, however, allows for precise placement of units that ensures maximum damage for minimum risk. On top of that, a grid system ensures that your placement is precise whereas a gridless system like FOT means that you can accidentally place the unit in the wrong place and end up with a dead unit. For example, you move the mouse a little too far and the bloody unit runs around the damned wall instead of running right up to it and stopping which means everyone and their mothers will open fire on the suddenly exposed unit. Opps! Although, I must admit that it is freaking hilarious when supermutants are involved as you can almost guarantee that you get multiple machineguns going off and rockets exploding everywhere.

Step up your micro game!

 

[Cael]

Imagine gridless chess. In realtime. With float-valued hitpoints for each piece. For me, it fundamentally changes the type of game from a tactical evaluation of distinct possibilities to something highly intuition-based, because it gets extremely complex. In RTwP, this often results in trying to slice up the problem into millisecond- pixel-sized fragments that are easier to understand.

I like a high degree of granularity. In space (grids), time (TB or phased), actor state (low scale of integer hitpoints and such) and actions (not too many similar spells, for example). Because it produces a feeling of control. And because the AI gets a chance to act intelligently by sampling a larger section of possibility-space.

That is fundamentally the problem that I find with RT vs TB games. It is extremely difficult to minimise losses in RTS and you fundamentally have to accept that you will lose units simply because you can't keep track of everything. TB, however, allows for precise placement of units that ensures maximum damage for minimum risk. On top of that, a grid system ensures that your placement is precise whereas a gridless system like FOT means that you can accidentally place the unit in the wrong place and end up with a dead unit. For example, you move the mouse a little too far and the bloody unit runs around the damned wall instead of running right up to it and stopping which means everyone and their mothers will open fire on the suddenly exposed unit. Opps! Although, I must admit that it is freaking hilarious when supermutants are involved as you can almost guarantee that you get multiple machineguns going off and rockets exploding everywhere.

Step up your micro game!

Hell, no. Just play Goths and have them streaming into the enemy base while you muck around reviving farms :D

 

[Zed]

hexes with 3 height/z-levels, no covers

 

[JarlFrank]

At the moment, only personal use. No plan for commercialization.

Then just print it and use it and tell your players to check out Realms Beyond

 

[Blaine]

How is this at four pages or even a question?

I'm not going to read this shit. I no longer want to know about others Codexers' shitty taste in everything; knowing just makes me miserable. Knowledge is extremely overrated. The RTwP freaks are depressing enough as-is.

 

[ASTRAL]

Hexes wont make a bad game much better and vice versa.Squares are good enough.

 

[Blaine]

A tiny minority of exceptionally rare games with shit thing were fantastic, therefore shit thing is equally good and just as valid as NO FUCK OFF, LA LA LA LA LA, I'M NOT LISTENING DOOP DEE DOOP DEE DOO

 

[Darkzone]

But you both do know, that depended on the on the dots (vertices) distribution and the dimension the result can be a square or a hex grid or something else? What you both correctly described is that you can use Voronoi partition growth to a higher dimensional room than R2 or R3. Btw. Truncated octahedron (you aim here at the bitruncated cubic space-filling tessellation) cells have two different sides and therefore are do not regular, but at least they are uniform.

That's true, but at least the "diagonal" movement distances here are not that different from the linear ones (sqrt(3)/2 vs 1) so they can be approximated as being the same (0.866 ≈ 1) for a game. Then you get 14 roughly equal neighbors instead of having 6 direct, 12 2d diagonal and 8 3d diagonal neighbors in a 3d cube grid. One disadvantage is mapping, however, you can't easily use a piece of grid paper anymore. Also, imagine moving in such a voronoi-blobber - how would be the control scheme?

For the A* this "3d cube grid" graph poses no problems and it might be more simple for human players than a bitruncated cubic honecomb, since the movement is more obvious in a square cube grid. But if it is not tested even once we don't know for sure the result.

 

[J1M]

Have any games done a 3d hex grid, or is it generally understood that hexes can't scale to a third dimension?

 

[vonAchdorf]

Chaos Chronicles and (I assume) Realms Beyond had different elevation levels. I don't know if two units could be in the same vertical space (1st floor / second floor). But technically it's possible. Blackguards two had different height levels too (but iirc disjointed).