23 replies to this topic

[Goldbott]

About a year or two ago I set out to make a mmo-ish third/first person role playing game after my favorite game of that type (believe it or not) went down. I've gone though countless races and settings. My latest design basically takes place on an alien world and the main races are human, robotic, and an odd alien race I personally made up. I'm all out of ideas so does anyone else have a few? I don't plan to use any directly but I need a bit of help getting started on a better race set and basic setting. I have the programming tools, the engine, the modeling software, and the network library I want to use and I have about siz years of mmo specific knowledge under my belt. Plue I'm not aiming for a AAA game here. I just need an idea.

Thanks.

[rouncer]

How about an event system, where npc's are spawned with variables that make them behave in different ways.

That would make npc activity a little more interesting, other than just goomba like ai some games come with these days.

[alphadog]

I'm not sure I am clear on what you expect to get from forum posts.

I want a Glaxphrax bounty hunter. Glaxphraxites are stripped ecru and fushia, with long tongues and love tomatoes. Also, they exude gas (not deadly) and have a fetish for Earth-made snowglobes.

Does that help?

In a nutshell, you don't start from "a better race set and basic setting". You start from creating basically a blueprint or encyclopedia on a world, put in some history, some anthropology, some socio-political high-points, and then window it in time and place to find your game setting.

For example, post-apocalyptic has been done umpteen times, but why is it still somewhat fresh in Fallout Las Vegas? Partly because you are getting a new window into a larger story arc. It's not just x races, y settings and z factions. Rather, it's a "feels-real-man" world. Large, space-faring, warlike ugly creatures with bony foreheads are a dime a dozen, but why are Krogans from Mass Effect, or Klingons in Star Trek, memorable?

Google "world building" and you'll get lots of material that expands this topic.

[fireside]

Start with your world, rather than races, etc. Why is this world different from earth? What would that change? If there are different conditions, what type of people would live there? What kind of special abilities would benefit them? How would their societies trade and coexist? How did their societies get started? What power source do they use? How would they travel, etc? Ask a lot of what-if questions and then come up with your own unique answers.

[Goldbott]

alphadog said:

I'm not sure I am clear on what you expect to get from forum posts.

I want a Glaxphrax bounty hunter. Glaxphraxites are stripped ecru and fushia, with long tongues and love tomatoes. Also, they exude gas (not deadly) and have a fetish for Earth-made snowglobes.

Does that help? :)

In a nutshell, you don't start from "a better race set and basic setting". You start from creating basically a blueprint or encyclopedia on a world, put in some history, some anthropology, some socio-political high-points, and then window it in time and place to find your game setting.

For example, post-apocalyptic has been done umpteen times, but why is it still somewhat fresh in Fallout Las Vegas? Partly because you are getting a new window into a larger story arc. It's not just x races, y settings and z factions. Rather, it's a "feels-real-man" world. Large, space-faring, warlike ugly creatures with bony foreheads are a dime a dozen, but why are Krogans from Mass Effect, or Klingons in Star Trek, memorable?

Google "world building" and you'll get lots of material that expands this topic.

My friend I have gone through those steps many times, scrapped the idea and started over. I just want some inspiration - everything I come up with fits into the usual "magic class" "fighter class" ect. I think I have a new idea that might be a bit more fresh. NPC's with events was an interesting concept - maybe take it a bit further and create a yearly cycle that would effect each NPC such as events in certain seasons and such - but thats a matter to decide later and if I do go on that it'll set back the games production quite a bit as it would be a very complicated and intricate system. The AI is my main concern - the only mmo shooter I know of "endless ages" had one HUGE issue and that was the AI - it always followed and shot at the "foot" armature joint on the players - that was about as complicated as it gets - I think adding a system that takes the power of the player as a variable for the AI as well as dead AI causing agro toward the player who killed it ect. But I digress, My point in this thread was to ask if the players want more sci-fi, fantasy or something else.

The outline I was going to go with before is at this site (http://pastext.net/1060) I wrote that maybe four months ago. It doesn't seem to be good enough though so I'm working to revise it - thats why I wanted some ideas.

[rouncer]

Item and spell crafting, would be cool. :) (maybe something like how mine craft item creation works? or even better maybe, I havent really thought about it myself)
Remember minecraft has a lot of fans.
Like if you combine a gryphon feather with an axe, you get a returning tomahawk! That would be nifty, and people would be hunting around, selling and trading all these little things that are actually components in weapon creation.

Um..., some pvp thing, like castle holding and invading might be cool too, give the players something to do while they are in the game, to earn more money I guess. Not necessarily what I said, just some guild wars type thing.

Always remember, when you implement these things, think simply and dont overcomplicate them too much, but on the other hand, dont oversimplify it past a fun game too.

These 2 ideas I suggested would help enlongate how long a player could play the game still satisfied, my younger brother is an avid mmo player, and he says just monster mashing gets boring quickly, so if you want to make a best seller, maybe you do need other "activities".

If you implement enough activities, maybe the emphasis on combat would be lessened, and you needent concentrate on it as much either. (unless combat is extremely important to you.)

[alphadog]

Again, I am confused. (Yes, it happens to me often. )

You want suggestions of game mechanics/rules or suggestions on original content? When the discussion is about AI or spellmaking, that's the former. But when you feel you are in a rut wrt races, classes, etc, that's the latter.

[Goldbott]

I suppose I just want some feedback and some good ideas for improvements on the run-of-the-mill mmo mechanics. I'll have to come up with the main story and such but I'd still like to know how people think an MMO should go.

[SyntaxError]

My theory is it doesn't matter. This is the list I use for in order of importance for MMO design features. Note this is just my gut from having played MMOs.

1) World design. How big is the world? Is it laid out contiguously. Is it seamless. Are there lots of different interesting zones? This has to do directly with immersion which is the main point of an MMO, so therefore I put it #1.

2) Game play. How does combat mechanics work? Are there professions? Do you level or just have skills. The list is endless. After you get past being awed by the world this is what makes the game fun.

3) Graphics. What is the visual impact of the game? It may be surprising I put this at three. It's just that there have seen quite a few stunning looking MMOs that just don't cut it for various reasons. After you have a handle on #1 and #2 then you can refine #3.

4) Genre. This is of course what you are talking about. I swear you could do Anthropomorphic Mutant Rats on Crack! (AMRoC!) and it would be a hit if #1, #2 and #3 were all excellent. I suppose genre has some impact, I just think it has a lot less than most people image.

Again this is just my gut view after giving it some thought. It's certainly open to argument.

[Zeppelin21]

I want to see an mmo that is huge. That takes a large community but, I am only dreaming ..... anyways.

A huge mmo, where every race has their own world/planet. You can teleport to the other planets via portals that randomly spawn around each world/planet. Would love to see random black holes where when you enter there is a special elite mob that drops rare items. Once that mob is killed you loot and leave and the hole disappears. They are rare and are a one time use. Elite quests, where you travel to other worlds and gather, kill, and disrupt environments that is permanent. While doing these quests you have the ability to open rifts and your world can teleport through to help for a reward. Big alliance raids of enemy worlds with groups by ship. Large PvP map with ground attacking as well as ships/planes. Gathering areas for more rare materials in space (free world).

A large pvp system where you can kill other races and your own, if you kill your own race you are flagged and can lose exp (more than if killed by mob). If you kill your race too much you become wanted and you are wanted. Whomever kills the wanted player receives a reward and whatever the bounty is. Prices of bounty's are set by the amount of same race players you kill (aka) player kill points.Two types of player kill points, the same race points are bad the other race kills add up for rewards in end game. You can become wanted by other worlds/races if you are on their territory too long, or if spotted by enemies.

Just a few ideas from last night. Im sure some of them are dumb but I don't really care.

[Nerd_Skywalker]

Zeppelin21 said:

I want to see an mmo that is huge. That takes a large community but, I am only dreaming :)..... anyways.

A huge mmo, where every race has their own world/planet. You can teleport to the other planets via portals that randomly spawn around each world/planet. Would love to see random black holes where when you enter there is a special elite mob that drops rare items. Once that mob is killed you loot and leave and the hole disappears. They are rare and are a one time use. Elite quests, where you travel to other worlds and gather, kill, and disrupt environments that is permanent. While doing these quests you have the ability to open rifts and your world can teleport through to help for a reward. Big alliance raids of enemy worlds with groups by ship. Large PvP map with ground attacking as well as ships/planes. Gathering areas for more rare materials in space (free world).

A large pvp system where you can kill other races and your own, if you kill your own race you are flagged and can lose exp (more than if killed by mob). If you kill your race too much you become wanted and you are wanted. Whomever kills the wanted player receives a reward and whatever the bounty is. Prices of bounty's are set by the amount of same race players you kill (aka) player kill points.Two types of player kill points, the same race points are bad the other race kills add up for rewards in end game. You can become wanted by other worlds/races if you are on their territory too long, or if spotted by enemies.

Just a few ideas from last night. Im sure some of them are dumb but I don't really care.

That sounds kind of like Eve Online.

[Zeppelin21]

Nerd_Skywalker said:

That sounds kind of like Eve Online.

im sure it sounds like a lot of things.

[SyntaxError]

Zeppelin21 said:

I want to see an mmo that is huge. That takes a large community but, I am only dreaming :)..... anyways.

This is exactly what I'm working on; or at least the technology for it. The actual game play aspects I'll leave to others or at least get some advice. The problem with having a huge world is that you need a lot of data on disk. The data required to build even a small sized planet at reasonable resolution is mind boggling. However with fractals you can store a lot of terrain in a small set of functions. This gives you the advantage that (a) You have a lot of space to run around it. duhhhh! (b) everyone is on the same shard and can all interact and © you can do this with no instancing, or at least instancing only where you want it for game play reasons (such as possibly in dungeons)

If you Google or search youtube for procedural planet, fractal planet etc, you can find a lot of examples. The problem is getting it all together in a game engine. For instance collision is a bit of a trick since you have to generate and refine your data as you go but at the same time you have to make sure you don't move fast enough such that the terrain doesn't refine underneath you and you get stuck. There are also a lot of issues with precision especially on the graphics card.

In my view the first one to develop a game like this will crush every other MMO (assuming the more generic aspects of it are not substandard). I'm sure there are a lot of people working on it. I know of at least one company that went belly up trying ot do it.

As for EvE, They get away with one shard because they don't have a lot of terrain data. The down side is you really aren’t in first person, more like first ship. Obviously there is some segment of MMO players that's Ok with this but I personally think it's limiting.

[alphadog]

Here's the paradox: manual content is expensive but usually gives users a better experience, whereas procedural content is cheap but usually an unfocused, undirected and thus unexciting experience.

[SyntaxError]

alphadog said:

Here's the paradox: manual content is expensive but usually gives users a better experience, whereas procedural content is cheap but usually an unfocused, undirected and thus unexciting experience.

I think it depends on your procedures. I would venture to say the procedural terrain often looks better. One way to solve this problem partially is to let the procedural routines select from a set of manually drawn components. I'm using drawn trees but the placement and selection is fractal. You can do the same thing with architecture components. The down side of manually drawing everything is you have to use a lot of identical assets. WoW does this and it's OK but I think WoW would be even better if each inn was different.

Edit:

One more thing I wanted to add. I think procedural content generation hasn't been explored much yet. The notion that it's unexciting may not necessarily be true. The trick is the have innovative algorithms. As it stands now can generate a planet and even as the developer I really don't know what's on the other side of that hill over there. I need to go climb that mountain and explore. This is with my relatively simple algorithms.

I can imagine a fantastic world with a high degree of variation. Also I think the size of the world is a big plus. You can make makes something large enough that your river boat journey might actually take a few days to complete and go though many environment changes. To me that's adventure.

Another idea I had was starting players of different nationalities in different areas of the virtual world. Imagine you travel a long way and suddenly all the players are speaking Chinese because they are in fact Chinese :o Maybe the NPCs speak Chinese too. So you have to find a translator and/or get a local to show you around. Now you have more opportunity for adventure.

[alphadog]

SyntaxError said:

I think procedural content generation hasn't been explored much yet. The notion that it's unexciting may not necessarily be true. The trick is the have innovative algorithms.

I've dabbled in procedural stuff.

Procedural content is exciting in the detail. The way a random landscape can be generated fractally is cool, and the level of detail that can be achieved is incredible.

The problem with all the procedural stuff I see right now is that all that algorithm-driven lacks a unifying, top-level layer.

For example: There's usually a sometimes weird reason why a city may be located on some part of a river in Real Life. If you take that and apply something similar into your setting, it adds character that draws players in. That kind of personalization and dimensionality is hard to get if one obsesses too much on generating everything by algos.

[alphadog]

PS: SyntaxError, have you been to this site (probably so, but just in case)

http://pcg.wikidot.com/

[rouncer]

Really, I just want to play an extended "diablo 1" that game rocked!

[SyntaxError]

alphadog said:

I've dabbled in procedural stuff.

Procedural content is exciting in the detail. The way a random landscape can be generated fractally is cool, and the level of detail that can be achieved is incredible.

The problem with all the procedural stuff I see right now is that all that algorithm-driven lacks a unifying, top-level layer.

For example: There's usually a sometimes weird reason why a city may be located on some part of a river in Real Life. If you take that and apply something similar into your setting, it adds character that draws players in. That kind of personalization and dimensionality is hard to get if one obsesses too much on generating everything by algos.

I agree, I actually don't care if I generate everything procedurally. My main goal is to reduce disk and memory usages so I can support an uber large world. Placement of cities for instance could be done by hand. Even if you have 1000 cities it's not that big a deal. Generation of cities themselves may need a more procedural approach however assets still may be hand drawn and combined procedurally. There is also the kind for fractal data compression which I haven't really gotten into yet. That may have some potential for storing hand drawn assets with minimal storage.

There are other issues with procedural generation. One MAJOR issue is rivers and roads. Rivers must flow downhill and take realistic and interesting paths and roads need to wind though passes and can't be too steep. This could obviously be done by brute force approaches such as moving fronts algorithms and rainfall simulation. The problem with these approaches is that you typically need to store your rivers and roads in a more conventional manner and they tend to take up a lot of storage. This starts to defeat the purpose of real time fractal generation. I had some ideas for generating terrain with the rivers and roads integrated into it. That way the whole thing can be stored as functions and not raw data.

The other thing I notice is a lot of guys are mainly working on the graphics which to me is maybe 10% of the problem. For an MMO you need basic physics (collision, gravity at least). You need to keep your avatar from getting stuck in terrain because it gets generated under his feet. It took me a long time to get this working properly. I wanted an algorithm that would guarantee this would never happen no matter what kind of delays occurred. You also need mob pathing which means for large terrain you need a server model which lets you generate parts or you terrain on the fly and populate it with mobs. Then you need to despawn that terrain when not in use. This has to be invisible to the player and seamless. Then almost everything with maybe the exception of local foliage billboards has to be LODed and of course you want the highest detail graphics models to match the collision model as much as possible.

At this point I'm just trying to get the basic stuff in place. There is no possible way I can do a whole commercial quality MMO from scratch by myself but I can do the basic engine and a demo and hopefully generate some interest. I still think some game company is going to get something like this to work and it will be huge when they do because the potential for a giant world online community is there. It's just a matter of time.

[fireside]

I think procedural modeling and texturing will find it's niche, but that's all it will be. Most games need to be populated with well thought out goals and that means conscious placement of game objects. Try drifting around in a simulation for a while and it will get old in a hurry. That's not to say there are manipulative games that don't work that way such as Minecraft or some simulations.