Imagine you are making a massive multiplayer RTS game, where players can build bases and units. Would you make it so people could attack only the stuff that belongs to other players currently online, or also allow to fight the AI that substitutes for offline players? it''s likely that the ratio between online players and offline players would be something like 1:24. So if you choose to allow players to fight AI''s that control stuff for offline players, then most of the fighting in the game would occur between real player and AI, even tho it''s a multiplayer game. But if you choose not to allow players attack stuff that belongs to offline players, then that can lead to all sorts of bad exploitations, like going offline when someone big attacks you.

I don''t think I would trust an AI with my base in an online RTS.;p

I think this sort of issue is one reason why the RTS paradigm would need a major overhaul before it was suitable to the ''massively-multiplayer'' approach. Not to mention the proliferation of units and resultant CPU death from all the AI.

[ MSVC Fixes | STL | SDL | Game AI | Sockets | C++ Faq Lite | Boost | Asking Questions | Organising code files | My stuff ]

yea that''s a tough one. Anyone got any ideas on how these problems can be solved?

I was thinking of a MMORTS that was really more like team based RTS. In it you have one player that plays the "Leader/King/Emperor/President/etc." and then you have other players fill the role of Generals/Admirals etc in charge of different parts of the armed forces. For example, you could split up your entire ground based army up in three parts, and then your naval assets to two players and your air force to two other players. You could also split it up in other ways...for example, just split up your entire combined arms forces between 4 players (for example, give each player one corps of ground troops, one fleet, and one air wing).

The trick in MMO games though is dealing with players who aren''t online. I''m not sure if there was really any way you could get around this, and server side CPU time would be tremendous. You could make the sides so huge that there would always be SOMEONE around. Another possibility is that the players are never really the top dogs...they are issued orders by the CPU "Leader" to tell them where they fight, and the battles are fought by the human players.

>Personally, I''m not too keen on the whole MMO paradigm. To me, it has no direction or substance or style really, but that''s just my take on it. I also have to wonder how many game designers also consider the fact that once you tack on the 12$+ monthly subscription fee, not that many gamers are going to invest in many MMO games. I think that even the most die-hard gamers will buy at most three MMO games...for even if they have deep pocket books, all people only have 24hours in a day. Basically MMO games will be the most cutthroat games on the market trying to vie for an audience that can only buy one or two of these styles of games.>

Although using AI to control offline player units could cause many problems, I think it is the right way for massively multiplayer games involving lots of fighting, because that completely solves the main problem - players logging off to avoid danger. However, in order to work this method would probably require:
1) Customizeable AI (for example, before logging off you can order you units to be passive or aggressive - or even to follow some attack or defense plan)
2) Player controlled to AI controlled unit interaction (for example, your allies can order your units to join their attack on your common enemy while you are offline - of course only in you have allowed that)
3) Battles must be long enough to allow all players to participate them online - that means at least few days - IMO the easiest way to achieve that would be to fight battles on huge terrains.

And of course that would require very good AI and very powerful computers - I''m not sure if this is possible at all at the present moment.

Think about it this way: don''t let every player have his own base. I mean, that''s not really feasible in a persistent or pseudo-persistent world, is it?
Many games nowadays do without base-building entirely. Both methods are fun, but they''re just different, neither way can be seen as better or worse.

The no-base paradigm obviously is suited well for MMO games, you can have a "drop-in" mode similar to FPSes, etc.
The base paradigm needs to be adapted, no doubt. A very vague idea of mine is as follows: don''t let users build bases at first. Have one or multiple "home" bases for each side, and give each player ("commander") a certain amount of resources/money, with which he can buy units. He then goes and fights a few battles, accumulates experience and depending on how well he fought (win/lose ratio, how big were enemy forces, what experience level was opponent, etc.) he''ll go up in the rankings, with the speed of (continuous) resources he is allocated increasing as his stats improve. At some point, you might want to allow the commander to gather resources (depends on your world of course) and leter even build buildings. The ultimate level would be to allow him to create a new base. This is of course very risky, so he''d usually want to cooperate with others (who are at the "build buildings" or "gather resources" level) to help him establish the base.
The idea behind it is that bases don''t really "belong" to anyone, and resources flow into a general pool that is then distributed to players.

This is obviously just a very vague idea, and it throws up some questions, the biggest being: "what if the commander is crap at constructing buildings?" (e.g. buildings block teammates)
Fights are reasonably easy to evaluate, but building a base is very complex, so we''d have to come up with rating ideas for that as well.

Thoughts?

- JQ
Full Speed Games. Period.

Here is another question:

Lets say that new players don''t get the ability to build bases right away, they first have to gain a rank thru combat. Just how much power and number of units should a new player get?

Do those units just pop out of nowhere when new people join? if no, who''ll be responcible for providing them?

And what to do with a new player who lost all his units? Start over with more units? I think that can be exploited

quote: Original post by berserk
Here is another question:

Lets say that new players don't get the ability to build bases right away, they first have to gain a rank thru combat. Just how much power and number of units should a new player get?

That would have to be balanced of course. Ideally, they wouldn't get units, but resources to buy units

quote: Do those units just pop out of nowhere when new people join? if no, who'll be responcible for providing them?

There would be bases to start with that provide basic facilities such as simple unit factories, barracks, etc. This would depend on the world of course.

quote: And what to do with a new player who lost all his units? Start over with more units? I think that can be exploited

All resources that are gathered and processed are distributed among players depending on their stats. If resource gathering units were built by a specific commander, he might get a bonus on resources collected by those units.
EDIT: this means that a player can build new units when he loses everything. Maybe he gets a small "advance" when he loses everything so he can be up and running sooner. This would have to be balanced through testing. (i.e. make it interesting to keep on playing using the same profile)

(those are just ideas of course, I bet you could do it a thousand other ways)

- JQ
Full Speed Games. Period.

[edited by - JonnyQuest on September 12, 2002 12:02:03 PM]

Another idea:
Have a "tactics board" where new tactics can be discussed among commanders of a team. So commanders would go into a sort of chat room with a map of the world that is dynamically updated (influence maps anyone?) and various elements can be applied to the map, like "commander X attacks here to distract, while commander Y takes out this crucial enemy pioneer base". Such combo missions would be scored collectively, so the commander who distracts the enemy would obviously lose a lot more units than the one who takes out the outpost, but altogether it is still counted as a success for both players.

Imagine playing such a game in a LAN for a weekend. Say, 16 people on each side, each side in a seperate room, with huge printed maps on the walls, and people making plans and all offline and then going back to the PC to embark on missions together. It all takes part on an entire planet or so. You could even bring politics into it by having 3, 4 or even more sides instead of only two! That would be the best game EVER!!!

- JQ
Full Speed Games. Period.

[edited by - JonnyQuest on September 12, 2002 12:11:57 PM]