Context: A 2D MMORPG is focused on PvE aspects. The game implements a quest system. Each quest has different difficulty levels, and each difficulty level is associated with a minimum and a maximum number of players, and with a minimum and a maximum experience level. When they want to participate to a quest, players have to group so that their number complies with minimum and maximum number of players, and character of each player must comply with minimum and maximum experience level. Problem: Some players fear to have difficulties in finding partners to participate to the quests. A possible solution: The solution is based on the possibility to use BOTs (like in FPS for example) as partners. Players have the capacity to create several characters. When they connect to the game, they choose a character as main character that they will control during the game session. They can also to enable their other characters to be controlled by a BOT program. This program will control other characters so that they follow the main character, and can attack automatically when enemies come close. During the game session, player can send orders to the BOTs associated with his main character using the chat system. A character is not always associated with a role (main character or BOT). Each time that player starts a game session, he can choose which of his characters is the main character and which of his characters are used as BOTs. What do you think about this solution ? dstar

Assuming most MMORPGs are forcing players to group up to play together, the introduction of bots to quest together simply defeats the purpose of grouping with other players. Unless there are other gameplay elements that takes advantage of the social aspect of the game, e.g. crafting, trading ... etc or partying with human players offer something more then bots, e.g options or unique abilities etc. Why not just make a single player game instead?

If the bot is poorly written, it will be a source of immense frustration. In fact, it may be nearly impossible to write a bot that can fill the role of a human player in a co-op campaign.

I remember Half life 2 on the PS2. There was a co-op campaign which my brother and I enjoyed alot. However, you could also play the co-op campaign as a single player, using the select button to switch between the two characters.

The character which was under CPU control would not move around and perform objectives, but would stand stationary and fight with reasonable (i.e. equiv. to an average player) skill. Perhaps this is the option to consider?

Just playing devils advocate: are you qualified to write an MMO?

Guild Wars does this. The problem comes that bots are invariably worse than people. They don't target smartly, they don't have subtlety, they don't answer to commands... It also limits what sort of quests you can provide. AI can't do 'sneak around this corner'.

Are they better than not being able to play because nobody is online? Sure. Are they (imo) always worse than a turn based game where you can control many characters? Yeah.

Wouldn't that be just a step further into turning the game in an essentially single-player experience?

If there are problems with parties and grouping, then maybe you should investigate the social and economic aspect of the game and fix the actual problem. Or alternatively, make single-player games.

Maybe I have to precise the context.

Typical situation is the following:

A quest requests that at least 8 characters have to group to participate. A group of only 6 players wants to participate.
Interest of the solution is the capacity of the group to use 2 BOTs to complete the group, even if they do not play as smartly as real players, they enable to upgrade the strength of the group as if they were 8.

It is not expected that a group of one player using one main character and 7 BOTs participate to the quest. Even if it is technically possible, this group won't have the same capacity to adapt to the different situations proposed by the quests as it will have been composed of real players.

Quote:

Just playing devils advocate: are you qualified to write an MMO?

I hope so ;) Game has been in development for more than one year, and is now well advanced. Graphic aspects are very poor, development efforts are mainly focused on the content.

dstar

You should definitely check out Guild Wars where players go through various levels in teams up to 8 players/bots. The effectiveness of your bot solution depends on how well their AI is written. Poor AI will create a frustrating game experience - then again, if your focus is to persuade people to join up and play with other human players rather than go soloing with bot parties, having a very basic/poor AI could do just that.

[Edited by - Tangireon on November 3, 2008 10:15:56 AM]

I'd wager that the inefficiency in trying to put together a group is by design. With the hours spent in WoW (as an example) waiting for groups, well, those are hours not spent getting through content. That means I have to keep playing the game longer (i.e. sending more money to Blizzard) to get through the content. Streamline all of that, and people run out of a reason to play the game sooner.

The bots behaviors are usually a hair smarter than the average rock and games previously had them be more like 'pets' which required constant direction.

One aspect is if they are to be smarter it will take more CPU than the game company wants to load their servers down with. So you likely would have to run the BOT on the client side (validated on the server side of course). A limitation would have to be made to limit how far away from the player the bot could be so as not to be a cheat (since te bot has to get the situational data around itself similar to what is sent the player for display.

Actually opening up the bot programming to the community could generate behavioral code better than what the company could do (example what happened in Neverwinter Nights with the combat AI).


I actually was thinking of starting up a thread about how big its practical to make player parties (quest requirements) because of the difficulty of finding other players doing the same thing at the same time (and getting past the 'hurryup and wait' thats typical player behavior). Pickup groups for difficult quests leads to inconsistancy and failure (and players eventually quitting).

Too many people required and its just not workable for most players and too few and you lose the interesting player interactions. Bots might at least bulk up a party (and the bot recruiting can become an addon aspect of the game (subquests)).

That is of course assuming the bots are capable of more than just getting in your way.

I'd also have to agree that you should check out GuildWars, most standard party bots are stone dumb, making some quests virtually impossible for all but the most skilled/tricked out players. However Hero bots (an upgraded version) seem to be better, largely because players have more control over them in term's of what skills they use, items/armor, actions, etc. Too bad you have to buy the later expansions to get access to them. :/