Spatial abstraction in cRPGs, worth it?

Since the beginning of the cRPG there has been little in the way of spatial abstraction. Going all the way back to the earliest cRPGs positioning was absolute whether tied to a grid or not. One could argue, perhaps, that the few monsters in IF that approach cRPG-like encounters were not gridded, like the axe-wielding troll in the basement of Zork I, but there the encounter was basically a one-square grid encompassing the room.

This is in contrast to many ttRPGs which are a lot looser with distance and positioning. Much of those games are played grid-less and it is enough for the GM to say "there are orcs positioned on a ridge on the far side of the room" or "two goblins fall upon you."

Would anything interesting come out of trying to bring more spatial abstraction to the cRPG? Imagine a very large cavern with an entrance, a pool in a corner, some stalagtites, and a ridge. Rather than it be a grid it would instead be a room consisting of four interesting locations and the characters and monsters would move from one to the other. Weapon ranges would be defined by these locations -- a melee weapon can strike at anything in the same location ("close"), throwing weapons could hit ones in adjacent locations (let's call those "near"), and bows could hit ones beyond ("far"). Targets could be self, an individual target, or anything in a given location. The spell fireball, for instance, would be a fire-based damage spell with range of far and target of the entire location. A healing aura would have a range of close and a target of the entire location. Heal would have a range of close and a target of an individual. Movement between locations consumes the mover's turn and would provoke an attack of opportunity from the location being left should there be hostiles there.

Why do something like this? Seems like it could be interesting in its own right, and perhaps hearken back to old school ttRPG play. I'm also wondering if something interesting could come out of it presentation-wise. By focusing on interesting locations a more artsy drawing (remember those nifty handouts in the old 1ed AD&D modules?) style could be used, and almost colorform-like rendering be done of the characters and monsters and objects done dynamically on top of them.

On the negative side, of course, is that all of this spatial abstraction does obviously work against establishing the dungeon as concrete places (a 100' x 50' cavern with a river snaking through it along the north edge, stalagtites clustered in the middle, with a 10' wide ridge along the east wall and an entrance in the middle of the western extent). Things like cover and line-of-sight are a bit interesting (though not unworkable) as well. It may also work against a party-based game too, as characters end up in different locations within a room (or even some outside of a room) -- the context switch may be difficult for the player.

Anyhow, I'm rambling -- anything like this been done? Is it worth pursuing (as a prototype)?

 

Sounds like a good idea worth exploring. The locations could be represented by nodes instead of a grid.

 

Doesn't the latest version of Warhammer Fantasy RP have a similar system? Might be useful for inspiration.

EDIT: Link to PDF rules

 

Yep, and as such standard pathfinding can be applied to it too for the AI, with weights (movement costs) calculated considering potential for attacks of opportunity (so a destination two locations away, the AI would favor the location that is empty over the one that is not all things being equal), contents of the location, etc.

Wow, I hadn't heard of the latest WHFRP doing this (I stopped following it with 2e) -- the tokens to indicate range are an interesting touch. For ttRPGs I've always thought of the location based focus as an entirely theater of the mind thing as opposed to a much more rigid grid-based, often mini-based, method of play. Didn't know any ttRPG had incorporated something in-between like this. Interesting, and definitely similar to what I was thinking.

 

I'd definitely like to see this idea explored more in cRPGs, and in fact I've been toying with similar ideas for my own project (a retro-style cRPG). I'm pretty sure Bard's Tale and Wasteland had some abstract notion of variable distances in combat, and the earliest Might & Magics (I and II) possibly too (my memory is hazy on that). But these didn't have discrete locations in the "room", as you describe.

When growing up playing AD&D we never had minis or a battle grid - it was all theater of the mind. And a lot of old-school tabletop players still play this way. It would be really interesting to figure out a way to translate that theater of the mind approach to a computer game. And I think you're on the right track so far.

I'll give it some more thought and come back to the thread later with any more ideas to add to it. But to answer your main question - yes it's definitely worth pursuing!

And thanks Copper for linking that Warhammer stuff. That is very interesting and something to consider for ideas.

 

I've played around with something like this, and have a prototype somewhere.

As J1M suggested, I had used nodes to represent a vague area on the battlefield, and they were laid out in a grid pattern, but there was space between each of them. At the time, any number of creatures could occupy the same node. Nodes were draw with an outline, and a node would expand or contract depending on the number of creatures in it, and the rest of the nodes would shift to accommodate the change. Creatures in the same node were all considered to be in melee range of each other, one node away was reach range, and outside that was ranged weapons range. AoE effects would affect everyone in the node. Certain locations could also have special qualities that affect those in it, like a hilltop that gives a bonus to ranged attacks.

 

I like that idea. Reminds me of Suikoden's tactical battle screen a bit (which was pretty simplistic, though; no doubt it could be much improved upon).

 

Is the key question here, as with pretty much all abstractions.

You give up:
-ability to automatically tell how many characters can fit in a room
-ability to have locations of vastly different sizes
-ability to implement chokepoints and many other spatial features crucial for tactics
-ability to integrate presentation and mechanics
-ability to integrate development of locations in terms of presentation and mechanics
-ability to let the game infer a lot of spatial relations we take for granted

for what?

Because "seems like it might be interesting" and "feels a bit like tabletop" don't quite cut it when it comes to dumping huge amount of interesting game logic and doubling or tripling your development effort.
I guess it might work for a tile-based blobber with premade backgrounds if you dropped stuff like unlimited location capacity, but other than that?
It seems like pure cargo cult design to me.

And this is actually point worth considering. I'm picturing a hierarchical model with graph of areas with, with smaller scale graphs nested in its nodes, up until the area can no longer be divided in a meaningful manner, but can be treated as contiguous space that can be traversed in easy way.
It still doesn't require such system being used for portrayal of game space in general, just partitioning it for the AI to digest.

 

Many small locations/nodes?
This concept (sans combat) reminds me of the adventure games. And don't forget the forest maps of Obitus.

 

Yeah, early games had some concept of range but there weren't discrete locations. Locations could have different attributes like "high ground" offering a bonus to hit on other locations (perhaps at a cost to the ability to conceal), "bloody" making footing treacherous after a couple of rounds of combat, that sort of thing.

That sounds a lot like what I'm thinking of. How did the prototype turn out? Was it fun? What advantages and disadvantages did it have versus the typical cRPG grid-based combat? One of the appeals of going to a more location-based combat is the potential for a different type of presentation, like an Erol Otus drawing. The abstraction makes something like that more possible, it isn't necessary to present exactly where the nearby critters are, but that they are there at all. In doing so, however, is one of the greatest strengths of having RPGs on the computer -- number crunching and quick situational rendering -- being forsaken? And of course one problem with going towards a more "artistic" rendering is that the development cost balloons greatly -- I'm no great artist, that's for sure!

Do the small dots between the larger dots denote distance in any way (similar to the tokens in the WHFRPG link above)? Interesting, I have never played Suikoden.

Yup, it worked really nicely in Adventure games -- just not sure it works for cRPGs.

 

Your list of what is given up is helpful. #2, 3, and 6 are particularly worrisome. I'm not sure what you mean by #4 (ability to integrate presentation and mechanics) and #5 (ability to integrate development of locations in terms of presentation and mechanics), however.

I'm a business programmer by day so "seems like it might be interesting" is a driving force behind my hobby programming. The fact that this idea has been done a few times is actually liberating -- there are a lot of other ideas I'd like to tackle and if this has been done then it may be best to move on. Something I'd really love is to create a game that I can continue to work on in my spare time that others find interesting, but can continually evolve. I was working on a Seven Cities like exploration game (http://www.bigweathergames.com/?tag=into-the-new-world) but as I started to delve into the history I frankly got depressed and my interest waned. =/

I'm guilty of being swayed by interesting games -- computer and otherwise -- in driving what I'd like to code. Whether it's Spelunky making me want to do a top-down real-time procedurally generated dungeon crawler with great environmental interaction or Dungeon World making me think about location-based combat, placing tags as behavior drivers on monsters, and no initiative combat or Wizard's Crown making me want to do an uber-tactical turn-based party-based game. Sometimes I want to do a single-character but with henchmen that can be commanded type game, likely turn-based.

Firm things on what I'd like to work on: procedurally generated dungeon crawler (I have no drive in me to tell a story, or create some massive story arc); traps as varied, interesting, and deadly as critters; interesting dungeon ecology; punishing dungeon environment including limited food, water, confined spaces, stale air, cave-ins, etc.; very mundane loot (I've never been a fan of "sword +X of slaying"); not just light, but smell and sound and touch mattering; brutal injuries with real effects -- broken legs and not just -9 hitpoints.

The rest is up in the air. Turn-based or real-time (very heavily leaning turn-based)? Top-down or Ultima-style (3D, isometric, etc. aren't in the cards due to not wanting to invest the art effort)? Single-character or party?

I hadn't heard the term cargo cult design before, but after reading it you may very well be right. Just because something may be interesting doesn't mean that it serves the ultimate point of the game -- the gameplay being enjoyable. Implementing it without an understanding of what it brings to -- and removes from -- the table would be a mistake.

 

It's a visual representation of distance, yeah, but I don't think it has any actual gameplay revelance. Would be cool if it did though.

 

Turn-based, party, ultima style perspective would be welcome. Maybe look to the SSI Gold Box games (battle maps) for inspiration, although you already mentioned Wizard's Crown.