2D vs 3D

[Strangely]

I’ve been working on my DM clone for about 14 months now. I am writing it in C++ using DirectX 9.0. Last week I discovered Game Studio and was able to produce the same level of functionality in 3 days. So, I thought, “should I switch to Game Studio and throw my engine away?” 14 months vs. 3 days seems like a no-brainer. It’s not so clear cut though. As far as I can tell there are several pros and cons for each approach. If y’all don’t mind I’d like to do some ‘thinking out loud’ here and see if I can get some feedback. If you do mind then go read something else.

The issue comes down to fake 3D vs. real 3D. DM is fake 3D, Morrowind is real 3D. What are the pros and cons?

DM style:
Pro:
Simplified movement – can move with keys and control chars. with mouse.
Easier level design – my editor is basically a paint program. You just paint the squares onto the map.
Abstraction – more on this below…
Con:
Simpler levels – basically, it’s only really feasible to do dungeons.
Simpler story – very hard to present any kind of story. It’s just a dungeon hack.
Simpler graphics – if it’s step based, monsters must move that way too. It looks silly if they move smoothly but you jump. If it’s not step based (quickly scroll from square to square with each keypress) I feel sick in no time. I once nearly threw up when I ran down a twisty corridor whilst testing this mode.

Morrowind style:
Pro:
More interesting levels - can be as complex/simple as I can make them.
Better storytelling – much easier to present a story.
Better graphics – only limited by artists’ capabilities.
Con:
Level design – it takes _much_ longer to make a level.
Movement – you cannot have analog turning with a digital input device. It’s terrible. So, you have to have ‘mouse-look’. This means you can’t use the mouse to control the party unless you have two mouse modes. This is much more awkward and would get really annoying in a large combat.
Lack of abstraction – again, more on this below.

Conclusion:
I think all the cons of the 3D-style can be overcome (I’m sure I’ll come up with a solution to the control problem). All except one, that is. I think it comes down to the problem of abstraction. That’s why I wanted to talk about this some more.

I think the problem of abstraction is what brings most of us to this site. It’s at the heart of the nostalgia we feel for games like DM. In the old days, computers were very limited so games had to be abstracted. Without hundreds of megs of ram and flashy 3D cards developers were forced to focus on the only thing that was limitless – gameplay. They had to really whittle their ideas down to the simple question: what is the most fun? Morrowind is the archetypical example of this. It’s flashy, huge, detailed, and excruciatingly boring. The trend towards realism in games is leading us away from playability. I think what it comes down to is that if you want reality, go outside. It’s no good in a game. The amount of time I spent walking from place to place in Morrowind really browned me off. Another problem with the trend towards reality in games is the effect it has on the mindset of the game designers. As the levels get more real, they start to think about all the things that need to be added to a scene to add to the realism. So you get inhabitants of towns wandering around, like you would in a real town. The only trouble with this is that the player doesn’t know which inhabitants are story elements and which are ‘reality fillers’. So, you have to wander around clicking on each person and 99% of the time you just get the ‘peasant #1’ dialog back. It’s very annoying and incredibly boring. Then they add shops, but they’re not just in one place, they’re all over, like in a real town. So when you make a town run to sell your loot, you have to walk all over town to get to each shop. Again, boring. There are many more examples along these lines. Another, more insidious, effect is that no matter how much processing power you have, it is still a simulation. What this means is that you can be walking around in an almost photo-realistic environment and you will come across an engine limitation that can be quite jarring/annoying (people acting like automatons, simplified physics, etc.). This isn’t a problem in a DM game because you expect a certain paradigm. I found _this_ more immersive than the pseudo-realistic world of Morrowind because it was _internally consistant_. You never hit that ‘reality buffer’ where you realized the limitations of the engine. That’s why I nearly crapped my pants the first time I saw the dragon in DM, bet never got anything like that in any 3D game role-playing.

Anyway, I think I’ve waffled enough on this for y’all to understand what I’m talking about. It comes down to the question, “is it possible to have true 3D but keep it abstract?”

Any thoughts?

[PicturesInTheDark]

Well, the obvious answer that comes to my mind on that subject is "of course, if you don't try to fill a game with everything possible but keep it at a level that does not show the limitations of a game to the player".

So - if you cannot/do not want to have "default" sentences spoken by people/monsters in your dungeon, either program them one by one or find an intelligent database that assigns sentences according to a certain situation (I concede that might be a little hard ;o)) or simply don't allow talk. Same for shops etc... If you really want to be abstract, you'll have to abandon a lot of things anyway - but who says a graphically well-made game has to have full reality simulation in all other components as well? Why not concentrate on what you are able to to well and stick to that?

Regards, PitD

[andyboy_uk]

What about the levels being made in a moddeler like 3d studio for complexity and cool looking-ness but they abide by a certain rule that allows the movement to remain like dungeon master. Giving you smooth transitions, nice looking levels but keeps that dungeon master gameplay.

Then you could have the 3d map overlay and your square map editor as well to define where the player walks.

THink stonekeep in real 3d?

I have no idea if that made sense or not Just an ideas

[linflas]

How exciting ! this is exactly the topic i wanted to open.
@strangely : if you haven't read my posts in DM Clones forum, please give your opinion on them :
http://www.dungeon-master.com/forum/vie ... 8&start=72

you can forget the "SBovis story" that follows

[Gambit37]

Hi Strangely! I was very interested in your post, in particular your views on abstraction. I agree wholeheartedly.

These days, computer games are caught in an arms race of eye candy over playability. I still long for the days of yore when -- as you rightly point out -- playability was much more key. Recently I have played what some consider to be 'realistic' games and found that the experience was anything but enjoyable:

Unreal 2: Very, very pretty graphics but excrutiatingly painful to play. The main problem is that the main character has been made to walk 'realistically' -- that is, he's wearing lots of heavy armour and can't move very fast -- and it kills the game stone dead.

Tomb Raider: Angel of Darkness: I've always liked the TR games for their puzzles/exploration. You can really feel a sense of achievement in the games when you solve something tricky. The new game however tried too hard to be clever/realistic and ended up being contrived. In some parts, Lara has to push a box or climb a wall, just to be come strong enough to complete a particular jump. She even says "I'm not strong enough yet" then when you become stronger says "I feel stronger now". It's the most moronic, contrived 'feature' I have ever seen in a game and is really only a simple variation on the 'find the key, open the door'... what a sacrifice of playability for 'realism' -- might as well have just got the player to look for a switch! (I'll ignore the fact the game was clearly unfinished and rushed out to cash in before the end of Eidos' financial year, but I digress...)

As for your question about 2D/3D, this might help, although it doesn't answer your question directly: In an interview after Dungeon Master 2 came out, Wayne Holder (president, FTL) was asked:

"Why does Skullkeep use a similar graphics engine to the original Dungeon Master, instead of a free movement design such as Doom or Ultima Underworld?"

To which he answered:

"We considered it for Skullkeep -- it's not that hard to do -- but we prefer puzzle-oriented game design. When you introduce free movement, puzzles become much more complicated to design. For example, with free movement, you can step around pits. We don't see our game being in that action niche anyway. We want something with depth. A lot of free movement games are tedious to play because you spend so much time bouncing off the walls. Personally, it gets very tiring."

So there you go. Wayne wants depth. I too want depth. I would much prefer depth and playability at the expense of realism. Whatever approach you take, as long as there is a good balance between playability and believability (NOT realism), I think you will do well. But don't waste time trying to be realistic and taking all the fun out of it...

[Strangely]

Thanks for your input, people. I'll address them one at a time:

PicturesInTheDark

I didn't mean that the problem is the sentences spoken by the generic NPCs but that they do not advance the story/current quest in any way. No matter how clever one gets with the 'generate dialog' code, they are still just filler. In fact, if you got really clever with the code, it would be confusing to the player because they would think it was useful rather than random filler. Too many red-herrings is frustrating in a story.

I think you got this though with your further comments. I was thinking of making places like towns totally abstract. You enter a town and go to a fixed screen. Sort of like a town in Baldur's gate but every shop on one screen and you click on it to enter. Then you go to the shop screen etc.

andyboy_uk

I tried that in my engine. Step-based movement but with smooth-fast transitions between squares. Made me want to puke. Very bad motion sickness and I'm normally immune to that in games.

linflas

I did look there. It's what gave me the impetus to post these ideas. The free-look with the mouse and a 'lock-on' key is a good idea. I still think it would be a little awkward in a tight spot. I think I've got a solution for this. I'm thinking of having smooth movement, but guided by the engine. What I mean is, when walking down a corridor, the engine keeps the player on a 'rail' down the middle. That way there's no bouncing off the walls etc. I'd also have to guide the turns to home in on the next creature/item/corridor in that direction. Not sure if this will work, I'll have to try it and see.

Gambit37

Wayne Holder's comments on puzzle-oriented design is exactly what I'm thinking of. That comes from the editor, I think, rather than the engine (you could make pits fill the corridor and, therefore, unavoidable). The easier it is for the designers to use the editor, the more thought they can give to puzzles rather than how to make the editor work. I like the distinction you made between realism and believability. This is key. You can do away with realism as long as the setting is internally consistent. This is easier to do in DM than Morrowind.

General comments:

I think 3D would work as long as I could solve the movement problem. I'll try the guided method and see what happens. Oh, hang on, I just thought of a problem with it. If you're in a large room, and you turn to face a corridor to exit the room, you could easily be off-center relative to the corridor. Hmm. Not good. More thought required... Ooh, I know. If turning locks you on to the nearest 'feature' it would point you to the center of the corridor. Then, as you moved into the corridor, it would correct your facing to point along the center. I'll have to try it. It might be horrible in practice. It has to be something like this because you have to free the mouse for character control.

I'll run some tests and let y'all know how it looks.

Thanks again for everyone's comments. I'm sure this is a solvable problem. If we keep throwing the idea around and chewing on it someone will hit on the solution.

[PicturesInTheDark]

Hello Strangely, yes I think we are on the same wavelength there (Gambit put it much more precise though), I simply tried to explain by giving "unrealistic" examples (such as programming thousands of individual dialogues) that I don't favor realism over atmosphere or, as the top word currently seems to be, depth. Not quite working irony approach, it seems. Will try to write it differently next time ;o)

Regards, PitD

[sucinum]

couldn't you have added a vote? i feel bad only posting:
2d > 3d.
but i still do.

[Strangely]

I didn't have a vote because I want to see people's reasoning rather than just the conclusion. That way I'll be able to get to answer the following:

Is 2D better than 3D because that's the only way to do it (games as fun as DM) or is it just a consequence of 3D shifting the focus of development for the designers? i.e. metaphysical or epistemological?

[linflas]

for me it's very simple, DM with its actual features can only be 2D.
So what game do you want to make ?
a DM enhanced clone ? Make it in 2D, with awesome graphics and add your own enhancements to it (cities, shops, non playing characters, dialogs...).
If not, you can try 3D to make something very different but with some DM inspiration i.e. world, heroes, weapons and the way they're used, runes, spells and invoke method, monsters, etc..
If I was a game developer, i'd prefer the second choice because you're completely free, you can even imagine a new game concept.

[Strangely]

The second option is the way I want to go. Long way off, but I'll get there eventually.

[MadMunky]

I havnt really read all the thread yet but thought i'll add my 2 pence, I dont think it makes a difference if the game is 2D or 3D, Dungeon Maker is 3D but its dont in a way to keep the Dungeon Master feel but enhance it with better effects plus thought its done in 3D we still use 2D for some objects and stuff which on most things you cant really tell the difference.

[Strangely]

Yes, I'm coming to that conclusion. I think that as long as the level designers (well, me) keep playability as the top goal and don't go down the "let's put this in to look cool" or "this is more realistic..." roads, everything should be ok.

[linflas]

then Strangely, if you're making a 2nd "Dungeon Maker", you should collaborate with madMunky in a unique project don't you think ?

[andyboy_uk]

Im gonna put in a little moan to say when can we have a demo of DMaker

[Strangely]

Re: linflas post on collaboration.

No, at least not yet.

I don't expect the first version of my game to be the one I really want. I'm just trying to get a base-line to work with. Then I can post it here and get lots of feedback. That way, instead of trying to decide if option A is better or worse than option B for everything, I'll actually have something to try.

I like the idea of there being competition. I'm sure there are lots of cool ideas MadMunky has thought of that I haven't and vice versa.

Then, at the end of mine and MadMunky's first iteration we'll either have:

1. - 2 great starts that will lead to 2 great finished games.
2. - 2 crap starts.
3. - 1 good and one crap.

In the first case, we all get more.
In the second we can either go do something more useful or try to collaborate and figure out what we did wrong.
In the third case, maybe the better one could absorb the resources/ideas/people from the bad one.

So, until I've got something to show to prove that I'm not just all talk, I need to keep going with my own project.

That doesn't mean I don't want help though. I'd love to have a team working on my version. It's bloody hard work doing it all by myself. Not a problem in itself, it just means it's going to take a long time. You see, there's the small matter of providing food and shelter for myself

So, anyone want to help with code (mostly script in Game Studio), modelling, graphics, ideas, spare buckets of cash?

[MadMunky]

I put up some videos on DM what more do you want!

Anyways the demo release is upto wishbone so send all your requests over to wishbone2003@hotmail.com

Put i think hes planning on a release before christmas I dunno if it will have any playable maps only test ones.