Important vote: Single Player or Multiplayer?

[Leland]

I just stumbled into this site and probably haven't done enough digging, but I'll ask anyway: how will the turn based model be fitted to MP gameplay? Specificly, in massively multiplayer games you can't guarantee that everyone can attend at any given time. Some will drop out, others will just miss a few turns, some might think too long, their computers could crash... anything can happen. How are these situations handled with? If the civilizations just follow previous orders until they run out, competing civs can use this to their advantage. On the other hand, to make the day to day running of a civilization independent on the player requires making fairly good AI.

Perhaps I'm missing something....

[The Joker]

First we are using preplanned turns, which will mean that the amount of time having to wait for other players to make their turns will be the same whether you are playing one on one or with 50 other people.

Second a nice feature of a long mp game would be, to allow players to hop in and out of civs. A game going from 4000BC to 2000AD would last months and months.

This would make it likely that not all players are interested in playing a game all the way through. Therefore players could hop in and out as they wanted. If a player quits a new one could enter. That way.

But on the other hand you are absolutely right that we need to work on this, as on all other game aspects.

------------------
"The future is that mountain."
- Bret Easton Ellis

GGS Website

[amjayee]

I think this is indeed a very important issue, so I'm making quickly my point. Sorry I couldn't attend to the meeting!

As usually, I'm along the same lines as Joker. We will have to start making this game with multiplayer system, for all the reasons stated earlier; most of all, the ease of testing the game, and also the fact that without good multiplayer system this game will be nothing. But I know all too well that we can't get rid of AI entirely. If we want to reduce micromanagement, which we _do_ want, we will need AI to control the tasks the player doesn't want to be concerned with. So, the game has to be designed from the beginning with AI in mind; in the beginning, there will not be much of it, of course. When the multiplayer game is ready, we will have lots&lots of AI in it. Then the last task would be to create the "master AI" which would take care of the whole government, just like the human - which can prove quite tricky to do properly, so it's good to start to make a multiplayer game; computer will most propably be able to serve only as a sparring opponent for multiplayer games.

I think Joker is foreseeing the future quite well. This game is designed for the computer and internet situation of the year 2002 or 2003. Then, computer resources will not be any kind of problem for us, and also internet connections will be good. What we have been talking earlier is, that the game would have lots of small people, and some large empires; empires rise and fall. If there would be 100 people, we could have 100 players in the game. Players would join a game, and get a people to lead. They would play, and try to get most out of their situation. When they have played enough, they quit, and other people or AI would take over their people. When the game has ended, perhaps when a certain year has been reached, the game history will be reviewed. Players would receive points for their actions and achievements during the game, and some kind of system would be used to rank the players. The better the player does, the more bonuses he will get in the games.

All this needs careful planning, but I'm quite assured that this might be a really revolutionary idea of a multiplayer game. It needs work, and might be difficult for beginners like us, but in my life I have become to notice that hard and honest work will always recieve its price somehow. I'm certain that if we just try, and don't be let down by difficulty of our task, we will be making something incredible.

[The Joker]

As always Amjayee makes a lot of sence.

And on top of this he agrees with me!

Great.

------------------
"The future is that mountain."
- Bret Easton Ellis

GGS Website

[ElmoTheElk]

I also like to say that it's not an easy to play MP game. The turns take a long time, escpacially with a lot of players and expanded areas. I onces tried to play MP civ2 but I game up while I was waiting just too long. I know we have a turnbased system with simultaneous movement, but even than turns would take a long time playing the game. In SP this is normal, but in MP this could really be a difficulty.

[Leland]

In multiplayer game a time limit for each turn is a must. It could be configurable for each game, or maybe dynamically depend on the size of the civilizations. In any case, the length of a turn would have great influence in the gameplay:

1) use 5 minute turns or shorter, and quick wits are rewarded in wars and such. Long term planning is difficult because the AIs will mess up the civ during the player's absence.

2) set the turns to more than 15 minutes, and you get two different kinds of players: those who make their moves and wait until the turn passes, and those who keep giving more and more detailed orders to their underlings. In real time situations there is no difference between the two groups, but the civs with better instructions while the player is absent might have an edge over other AI-driven civs.

3) set the turns closer to hours and you get really long chess-like games. Only 24 years pass in the game world per each day, so a moderate game spanning from 1000BC to 2000AD would last about four months. The role of AI would be reduced.

All of these have their merits and as such should be possible by just tweaking the parameters for each multiplayer game. From game design point of view two things need to be taken into account,

a) players should be able to make big decisions quickly and
b) players must have means to elaborate their decisions as much as possible.

By b) I mean that there should be a way of giving orders like "achieve a pact with Vulcan and if the Klingons start boosting their military wipe them out" or "maintain status quo until explorers find a new continent; if new continent is uninhabitet or habitet by civs with lower technology, begin colonization".

How the players can jump in and out of the game should be as invisible as possible from other players' perspective. There would basicly be three kinds of civs on the game: those controlled by human players in real time, those controlled by AI advisors for humans and the civs whose human player has resigned. The third ones can be used for new players who wish to enter the game, or alternatively the game could force a new nationality to be born from the population of an existing civilization. Of course, in order to make the game as versatile as possible, some games could be "closed" and not allow new players.

[chrispie]

Ok, it seems the game being made is -completely- different from the one I thought it was - fair enough, I'm flexible.

So, we going with MP with an AI which just 'looks after' nations when no human is in charge then? That's fine by me.

[Leland]

quote: Originally posted by Guildmaster on 12-28-2000 12:14 PM I see no reason not to have each game with a total of 128 slots open, some devoted to nations and peoples, some to economic tycoons, some to religious leaders, some to generals, etc.

I see plenty of reasons. For one, GGS is megalomaniacal enough as it is, and drawing in generals, tycoons and religious leaders would unncessarily bloat it. Secondly, there are plenty more positions for businessmen and generals than there are for supreme rulers... you'd have your game world filled with these obscure characters, most of them probably being played by AIs.

Thirdly, that's simply not the focus I see in this particular game. Other may disagree, but in my opinion GGS should be about rise and fall of civilizations, about the mechanisms how that happens and the about managing the factors involved therein (I probably read this somewhere...). Sure you could have a civ that draws its power from religion, for instance (think of vatican or the jewish or aztecs), but if the player can't cope with economy, population control, warfare, diplomacy, logistics, ... you name it, he loses. That's my vision, but of course I'm not a developer and probably would never qualify as such.

quote: When you decide to go to bed, and you leave the game, your reign comes to a close and you retire or die or whatever, and your position is replaced by either another AI or an awaiting player.

What if people want to play the same civ more than once? There should be a way to give the viceroy AIs plenty of instructions and guarantee that at least another human player won't be messing with your ideal civilization.

quote: My only question is, how do we do the turns? Do we want to make each turn last a set ammount of time? I know that I don't want to be sitting around forever waiting for some jackass in Montana to get out of the bathroom and hit "continue" if you know what I mean.

Configurable time limit. I'm not sure what the rest of you think, but my opinion on this is fairly well explained in my previous post.
[This message has been edited by Leland (edited December 28, 2000).]

[The Joker]

Leland:

Your idea about the max time limits are pretty much what I think we had in mind as well.

A:
This is completely true, and will be implemented. It shouldn't be too hard either.

B:
This is a more dificult thing indeed. I think since this will require immence AI that it is something that we will have to wait some time for. But when we get a good AI we should also try to make this possible. It would be a magnificant feature, that's for sure.

quote: What if people want to play the same civ more than once? There should be a way to give the viceroy AIs plenty of instructions and guarantee that at least another human player won't be messing with your ideal civilization.

Yes, it should be possible. Hard to create, though.

quote: Configurable time limit. I'm not sure what the rest of you think, but my opinion on this is fairly well explained in my previous post.

I agree.

Guildmaster:

Although I agree with you that mp should be the main thing, I think the "no AI at all" thing is just too odd to be possible. AI's in some forms are needed, if not for anything else then at least for making sure the units are going to end up where you want them to.

Besides, if we can make a good AI I see no need to not do so, and make it active and as much like the players as possible. Here I agree with that guy (can't find it just now) that said that a large mp game could include both AIs and players, and that the players should have as little knowledge about who was a player and who was an AI as possible.

The 128 slots is pretty much what I had in mind as well. But I think that we should not settle on this before we know what is possible and fun. So maybe 50 players will turn up to be the highest amount to keep the game and diplomacy fun, or maybe 200 players is possible. So whatever we find is the best amount should be what we pick.

The whole concept of players being able to play as religions or tycoons is one that I as well like. And since they should propably both be controlled by some sort of AI (at least they should) it would most likely be possible to put a player client in them in stead of an AI. But before we do that we should have a game that is fun to play as a civ leader. Most of the game should be settled before we can even think about making a game where the player is a religion possible. But I agree that it is a fun and interesting option. It should just not be anything that we spend all our time now thinking about.

I think the turn max times have been explained, so no more of that.

lol@the jackass in Montana, BTW.


Chris:

quote: So, we going with MP with an AI which just 'looks after' nations when no human is in charge then? That's fine by me.

I don't think we would have to be this extreme. AIs should be as active as possible, and as close to human players as possible.

------------------
"Life is a lesson. You learn it when you're through."
- Limp Bizkit

GGS Website

[Guildmaster]

Ok my vote:

Multiplayer all the way- No AI at all.

Ok exactly when ARE we going to be finished with this game? I can see it done in a year maybe two years from now. Most people I know already have high-speed internet access, and in another two years wether to be connected or not won't be much of an issue, especially since the most popular games even now and for the last few years have been internet-exclusive games.
I see no reason not to have each game with a total of 128 slots open, some devoted to nations and peoples, some to economic tycoons, some to religious leaders, some to generals, etc. You have each slot that is not occupied by an active player be an AI (ok some AI is required) but that AI will be basically weak, easily bullied around and not much of anything to anyone. Essentially an inactive player.
Once you join the game you look at the open slots and pick one and you are the newly "elected" leader of that particular position. When you decide to go to bed, and you leave the game, your reign comes to a close and you retire or die or whatever, and your position is replaced by either another AI or an awaiting player.

My only question is, how do we do the turns? Do we want to make each turn last a set ammount of time? I know that I don't want to be sitting around forever waiting for some jackass in Montana to get out of the bathroom and hit "continue" if you know what I mean.

------------------
He's spreading funk throughout the nations
And for you he will play
Electronic Super-Soul vibrations
He's come to save the day
- Lenny Kravitz

[cavebear]

It seems that you are planning both large MP play *and* a sophisticated AI in the recent posts, and most everyone was recommending against that earlier. The question of what to do with a civ when a player leaves seems important.

1. If civs are allowed to die...

If civs die when players leave, then there will never be much advancement in the game unless you have a way to create new civs at advanced levels after the game has started. Otherwise, new civs would be constantly created at the "start" level, and would be hopelessly behind the current human players.

What if new civs (after the start of the game) were allowed to commence at some "average" level of existing civs? Or perhaps equal to the worst human civ in play at the time (otherwise, players doing very badly would keep quitting and rejoining).

2. AI

Or, you could have abandoned civs continue with an AI to guide them. We know how poorly Civ2 AI's operate, so that might be difficult.

Or, again, perhaps some "average of existing human civs" trick could be used. Not quite an AI, but just a way to keep computer civs viable for new players to take over.

It might even be worthwhile to consider copying an existing civ when a new player enters.

3. The map is finite.

In most MP games, some civs advance and some are wiped out. It doesn't take too long for most of the map to be controlled by only a few players. You may be planning to allow 100 players, and new civs to be created as new players enter the game, but where will they go? Will you steal space from existing human civs?

I would expect to find that a game starting with 20 players would quickly be reduced to a dozen or fewer that controlled the whole map.

Perhaps a game like that should be locked until an existing civ becomes available (by a human player quitting). And if so, then you need to allow for multiple games.

[Leland]

The impression I got was that the "fall and rise" model of civilizations would be inherent in the game mechanics. Big, old empires would either break down in face of external threats (small civs looking for lebensraum) or stagnate to death. Kind of spenglerian, but I like it.

If the server was really smart, it could find evaluate the status of current players and find suitable revolutionary spots where new players could start. This wouldn't bother the esisting players much since they had problems with those troublemakers anyway; and a new neighbour migth prove to be a valuable ally.

But what do I know.

[The Joker]

Cavebear:

Again I don't think a sophisticated AI and mp would be mutually exclusive. We just have to figure out how many civs we want to have in the game, and make the game work with that amount. That is the key issue.

1:
Think history. First, if a player leaves his civ should be controlled by an AI, untill another player takes it over. Second new civs are going to get techs really fast, due to tech diffusion. And the rise and fall concept will make sure that civs, once strong, will be surpassed by other civs at one point.

3:
The map will be huge. Most likely around 1,000,000 hexes, or 1000x1000 hexes. This will mean that untill modern times there will always be free space for a new civ to get. And a completely new civ may just get one hex to start with. Another option will be to use nationality to let a new player lead a nationality conquored by another civ, or a new nationality just arisen, and try to lead that to independance.


Like Leland wrote, pretty much.

------------------
"Life is a lesson. You learn it when you're through."
- Limp Bizkit

GGS Website

[ElmoTheElk]

1. AI

Whatever we choose for our game, we NEED and AI. Either for not human player and abandoned civs. If a MP player leaves his game, the civ won't disapear instantly. (Or at least I think it shouldn't.) The AI needs to handle this civ. Then, if you really have a standard non-variable number of slots, thre should always be AI civs that probably won't be ever run by human players. Espacially with so much slots as 128, I don't think we see this large games totally fulled up with human player.
Now, if we NEED an AI, we can also implent single player. Maybe I'm to simple minded, but we need a decend to good AI anyway...

2. MP
I am convinced we should make MP the main thing in our game (so not exclude SP). However the max you set (128) is pretty much. Even if we have maps with 1,000,000 hexes (so about 25 km hexes) it may not be reliable to play on such map with 128 players. This is thought that every player runs 1 civ.
A good thing about a fixed number of players is of course that there will always be a place for new player to enter the game...

Basicicly I like Lelands ideas.