So lately I have been delving into the AD&D books a lot and chatting with people online about it on forums. Unfortunately, I don’t have anyone to play it with and there are many like me on those forums. We got together and tried to think of ways to play online with eachother.

  Each person came up with some pretty good, but tedious ideas. When a member mentioned that we could actively play over a chat, the light-bulb went off in my head.

Well, in my excessive time playing Diablo II, I have ran into a lot of cool chat bots. Ones that monitor “runs”, ones that monitor duel games and record stats like deaths and deaths by whom, ones that organize that channel to organize a clan/guild and memberships [Non-clan members get auto-kicked and members get a ranking system], and most often, trivia bots.

That being said, I’ve seen chat bots do enough things to know that it’s possible for someone to create a campaign and role play bot. It could randomly generate damage, treasure, experience, etc. The players play by commands.

For instance, the chat bot could say

DM Bot: You stand on a precipice, something flies in the distance
DM Bot: *I will now give players a moment to submit commands*

Player A: /Atk @Target 1 w/ Bow

DM Bot: Player A has submitted an attack command [.Attack @target # /w weapon/spell]. Player A’s attack misses Target 1.

DM Bot: Upon being attacked, the flying creature, an Imp, descends upon you and attacks.

DM Bot: The creature attacks Player A and deals 12 damage to him/her.
Player A now only has 123 HP remaining.

DM Bot: Remember that at any time, you can type .stats to see your stats, .skills to see your skill list, .Gear to manage your equipment, and .Item to browse your items and or use them.

Is this possible and or does it exist in any form? If not, is there anyone willing to try it? I have absolutely no knowledge of programming, let alone chat bots, so I figured I would ask here.

Hi, Jeremy, ans welcome to chatbots.org!

Many, many years ago, I used to play Rifts® with a group of my friends, but sadly, we each went our separate ways. At the time, in an effort to keep the games alive, I began working on a way to allow us to play across the net, complete with a chat section, character sheets, and a “DM” function that allowed the current DM to see all necessary info on the other logged in characters. Unfortunately, due to the time it took me to get the project to a “marginally functional” state, the games died anyway, and I ended up abandoning the project. I’ve still got all of the project files, but not much in the way of time to continue working on it. If you know of someone who may be interested in taking it over, and is skilled in PHP/MySQL, I would be more than happy to zip the project up and turn it over to them. Bear in mind that it doesn’t have a working bot at this point in time (I wasn’t involved in chatbots at all when I was working on this), but it wouldn’t be difficult to integrate one into the project. I have to warn you, though. Building a bot of that nature from scratch (just the responses and conversation… NOT the interface) is a huge job, unless you want the bot to be nothing more than a random damage generator. In order for a bot of this sort to be a good DM (or a DM’s assistant), it needs to have extensive knowledge of the realm/region/locale that the game is set in, and needs to know where everyone is within that realm/region/locale, plus knowledge of creatures/enemies/NPC’s, etc. It’s certainly doable, and I don’t want to dissuade you from the notion, but ti’s going to be a lot of work; even if you had something like my Rifts® web app to work from. If you know of someone who would be willing to continue with the project, let me know, and I’ll get in touch with them to set things up.

Dave Morton - Oct 1, 2011:

Hi, Jeremy, ans welcome to chatbots.org!

Many, many years ago, I used to play Rifts® with a group of my friends, but sadly, we each went our separate ways. At the time, in an effort to keep the games alive, I began working on a way to allow us to play across the net, complete with a chat section, character sheets, and a “DM” function that allowed the current DM to see all necessary info on the other logged in characters. Unfortunately, due to the time it took me to get the project to a “marginally functional” state, the games died anyway, and I ended up abandoning the project. I’ve still got all of the project files, but not much in the way of time to continue working on it. If you know of someone who may be interested in taking it over, and is skilled in PHP/MySQL, I would be more than happy to zip the project up and turn it over to them. Bear in mind that it doesn’t have a working bot at this point in time (I wasn’t involved in chatbots at all when I was working on this), but it wouldn’t be difficult to integrate one into the project. I have to warn you, though. Building a bot of that nature from scratch (just the responses and conversation… NOT the interface) is a huge job, unless you want the bot to be nothing more than a random damage generator. In order for a bot of this sort to be a good DM (or a DM’s assistant), it needs to have extensive knowledge of the realm/region/locale that the game is set in, and needs to know where everyone is within that realm/region/locale, plus knowledge of creatures/enemies/NPC’s, etc. It’s certainly doable, and I don’t want to dissuade you from the notion, but ti’s going to be a lot of work; even if you had something like my Rifts® web app to work from. If you know of someone who would be willing to continue with the project, let me know, and I’ll get in touch with them to set things up.

The offer is appreciated, but I highly doubt I’ll ever find someone to take it on… but, if I ever do, Ill send you a PM.

I am thinking of some fundamental ways that can make the bot less of a project.
It should rely more of a “Cause and Effect” mechanic. A lot like how trivia bots work.

A trivia bot asks a question out of many, there is no doubt you can make it ask these questions in a predetermined order. After the bot “Asks” these questions [Mostly will be *Players now have a chance to respond*], the players have options. I have seen trivia bots that have more than one correct answer, so no doubt, there can be many results recorded to these “Questions”... all with different commands.

/Atk = an attack
@Target # = an attack must have a target. The bot has every possible target recorded as Target 1, Target 2, Target 3 Etc, and will list what every target is if there are more than one [Just as if a trivia bot would list different categories].

/w Signifies what kind of attack you are using. This will be the hard part I suppose, as it would have to know what options you have and be able to record what adjustments there are to those options.

A bot also has to record creature stats and your stats. I am thinking a second bot just for that might be more practical, but I honestly don’t know, I am just using logic. The problem with that is that the two bots would need to communicate with eachother. Because the DM bot needs to know when the battle is over and continue to the next event [Question].

For that, I am thinking the second bot needs to have a database of all of the enemies you encounter, as well as the player stats. Hell, when the player says the attack command like /atk @Target #86 /w sword, it should be the second bot that responds to that with results. It compares both the players stats and the enemies stats and gives a result.

The second bot recognized Target #86 as an Imp and compares that vs the player.

A trivia bot can give players points [I have played with Diablo II trivia bots and have some all time streaks and scores] based on the result of their answers. This feature could be transformed into your experience and loot.

So yeah, it shouldn’t be that hard by my logic. The bot needs to be able to compare stats between opponents, that’s what is going to be truly difficult, but the role playing story part of it is just narrative with occasional “question-like” events that represent battles.

What would also be time consuming is creating a “Tree” of events. Often the bot needs to give players choices [Again, question like] and it will determine exactly which point in the story [Which order/place of questions/narrative[ the characters will head in.

For instance

DM Bot: You come to a crossroads, you can take the right [Command: Right] path where the fog builds or you can go left [Command: Left] where you hear roars in the distance.

Llorianas: /Right

DM Bot: You walk down the foggy path and lose your balance… suddenly the ground beneath you disappears and you realize you have fallen off of a cliff.
You have 2 choices: Grab [Command: Grab] anything you can on the cliffside, Fall [Command: Fall] and hope you land in water or on something that breaks your fall.

Llorianas: /Grab

DM Bot: *Rolling a dice for odds to grab something*
You did not manage to grab anything and keep falling, but luckily you land in a bog.
Unfortunately you are unconscious for now and get washed up on a nearby shore.

The entire story would have ended up MUCH different had he chosen the Left path. The bot needs to be able to skip around the “Chronological” list of events like that. Eventually, both paths end up at the same point [Some might actually be fatal instead though] though.

If’s it’s done as efficiently and practical as I am describing, I am sure it’s very doable, just time consuming because of all of the material needed to be programmed.

That was pretty much my point.

Having played several different “flavors” of AD&D type games, and also having created an AIML bot, I’m confident that this is something that can certainly be accomplished, but it will likely take a team of people, rather than any one individual, to make it happen. It’s been noted before, and I’ve experienced it myself, that those who have the skills to program something like this generally lack the story-telling skills to create the necessary geography, background history, characters, and monsters that would make the world “come alive”. Perhaps someone will read this discussion, and in so doing, will be inspired to bring this (or a similar) project into being.

Dave Morton - Oct 1, 2011:

That was pretty much my point.

Having played several different “flavors” of AD&D type games, and also having created an AIML bot, I’m confident that this is something that can certainly be accomplished, but it will likely take a team of people, rather than any one individual, to make it happen. It’s been noted before, and I’ve experienced it myself, that those who have the skills to program something like this generally lack the story-telling skills to create the necessary geography, background history, characters, and monsters that would make the world “come alive”. Perhaps someone will read this discussion, and in so doing, will be inspired to bring this (or a similar) project into being.

I was thinking the same exact thing, that the person who DOES decide to take this project on most likely wont have the story telling and creative spark to realize my vision and bring it truly to fruition.

A team WOULD be required, you’re right. If that team ever happens, I would be glad to be a part of the story and “blue-print” part of it, as I know exactly what needs to be done to make it special, I just lack the skills and time to do it.

I think the game play piece has been done before. Wizardry, Bard’s Tale, Ultima, HHGTTG or any of the other text-based 1st gen PC games essentially run a campaign. The benefit of a human DM is improvisation. It’s been 20 or more years since I played, but as a DM, I would ad lib to keep the game not only moving, but moving in the right direction. That sort of plasticity would require AI smarts on an unprecedented scale - A bot who could read his audience and play emotions.

Brien Malone - Oct 2, 2011:

I think the game play piece has been done before. Wizardry, Bard’s Tale, Ultima, HHGTTG or any of the other text-based 1st gen PC games essentially run a campaign. The benefit of a human DM is improvisation. It’s been 20 or more years since I played, but as a DM, I would ad lib to keep the game not only moving, but moving in the right direction. That sort of plasticity would require AI smarts on an unprecedented scale - A bot who could read his audience and play emotions.

Yeah, I understand that completely, which is why it does seem like a redundant endeavor.