Page 2 of 2 FirstFirst 12
Results 11 to 18 of 18

Thread: Artificial Intelligence

  1. #11
    Welcome LDiCesare! It seems like we may have spoken somewhere before. I recognize that name Seriously though, thanks for joining us!

    OK, first WoM is probably going to be moddable, at least in the long run. We've already had players asking for editors for things like the factions, the races, the loot, etc. We've even had requests for an AI editor. At the moment I can't say what will and won't make it into the core game, but moddable is certainly in the game's future.

    So, how is that going to effect the AI? We'll start with what I feel is the meat of your question and that is changes in the power level of certain units and the addition (or alteration) of certain spells.

    This ties in to our using the D20 system for our combat mechanics and is one of the many reasons it's a good idea. (I don't know how much you know about D20, so forgive me if you already understand this explanation.) In D20 everything can be given a Challenge Rating (CR) that is a rough estimate of it's power level compared to other creatures. The CR attempts to quantify every aspect of a unit (or player or monster as the case may be) into a single number. (There's actually a lot to it, but I'm trying to keep it short here. If you want more explanation we can start a CR topic about it.) All the elements of a unit will be used to generate that unit's CR. The AI will consider the CRs of the different units in a battle to make decisions about what it can or can't defeat. It will also take into account specific things like Damage Resistance or Spell Resistance and update its tactics according to the units current stats, not the default stats that the unit came with in the game. Now, this is a complicated process and the base units that come with the game will probably be slightly more AI balanced than modded units. We will do our best to quantify as much of it as possible, but people who are determined to rook the system are going to be able to. Of course, as we find bugs, oversights, etc. we will update the AI.

    New or altered spells are a slightly different matter, but again they have a basic power per cost and effect. If you make a fire spell super cheap and super effective you can expect the AI to start using it against you all the time. To be frank, I'm not sure to what extent we're going to allow spell modding. We haven't discussed it much yet. Players may very well want it. In any event, it's all about quantifying it so the AI can tell what the current situation is and adapt.

    OK, diplomacy is a HUGE topic and we haven't really started talking nuts and bolts yet. In mind mind it could almost be considered a mini-game. So, forgive me for addressing your specific question. I realize it's just an example, but at the moment all I can do is offer you a counter-example. First, before you could make the trade agreement you used in your example you would have to get the AI in a good mood. For the moment we'll say there's a “good time happy feelings” meter. At the start of the game the AI wouldn't have enough GTHF (Good Time Happy Feelings) with you to make a deal like that. You would have to have had some other dealings with him first. Let's say for the sake of argument that you did that and have stung the AI with a successful con. Well, now he's mad. All his GTHF have become BYBOWAGCF (Beat Your Brains Out With A Gold Club Feelings), so he's going to hold a grudge. The next time you try to talk to him he demands tribute in the amount of double what you took from him. No matter what you want he's not going to talk until you give him the tribute. Even after that it's going to take a lot of work to make him trust you again. I know I'm speaking in broad general terms, but until we have more of the Diplomacy system set in stone it's all I can say. The idea I want you to take away from this is that we're aware of the delicacy of the AI Diplomacy dynamic and that we plan to give it in-depth consideration

    In response to that other post I would say this: We plan to update the AI as time goes by. We're going to put a lot of play testing in it before release and make it as good as we can. However, there are already plans for DLC and expanded content. Major changes in content are going to require AI updates. So, we plan to update them.

    Thanks again for coming by and jumping in. Feel free to ask any more questions you may have or join in on any of the other discussions!

  2. #12
    I'll re-post something I posted in a thread awhile back in relation to AI:


    Now, there's been something I've been thinking of that ties into the AI, diplomacy, and quest generation. And that is having a sub-plot \ motivation system that ties all these things together.

    Each AI (Be it a AI leader, a dragon, or a independent town), would have their own personality / motivation / sub-plot system.


    For example, lets say we have 2 AI's and a Red Dragon:

    AI Leader "Merlin"

    PERSONALITY:

    • Peaceful
    • Deceitful
    • Isolationist


    MOTIVATION (Generated From Personality Traits):

    1. Spell Research
    2. Mana Production
    3. City Defense



    AI Leader "Beowulf"

    PERSONALITY (Static):
    • Aggressive
    • Honorable
    • Expansionist


    MOTIVATION (Changes, but Generated From Personality Traits):
    1. Exploration
    2. Military Production
    3. City Settling



    Red Dragon

    PERSONALITY:
    • Greedy
    • Deceitful
    • Isolationist


    MOTIVATION (Changes, but Generated From Personality Traits):
    1. Accumulate Wealth
    2. Stay Hidden




    In this "ecosystem" you'd ideally the game would generate sub-plots / quests based on these various factors, for example:


    Merlin controls most of the island he is on, with the exception of a single town that Beowulf settled, thereby encroaching on his land. Even though Merlin is peaceful, his deceit and need for isolation drive him to want the town gone or in his control. However, his need for peace makes him want to do this under the table so that he doesn't risk war with Beowulf. Due to this, a subplot might be generated a number of ways:


    Player and Merlin on Good Terms: Merlin presents a quest for the player to get Beowulf's town for him under any circumstances. His reward might be good research, a shipment of mana crystals, etc. Player could do it through diplomacy, force, or even by creating quests for others (The Red Dragon) so that Beowulf doesn't know they were involved.

    Merlin met Red Dragon: Merlin presents a quest for the dragon, in trade for wealth, the Dragon will eliminate the town for him. In above scenario, player might use the dragon in the same way. If the Dragon sees that Beowulf has much more money than Merlin/the player, then it might actually betray Merlin/the player, and tell Beowulf of his plot in return for a larger fee. The dragon might even just demand the money from Merlin/the player up front, and then take the money and run due to it's deceitfulness and need for wealth and isolation.




    As you can see, there is a lot of potential for such a system. And since it would be generating some really cool events, you'd want the player to have some way of getting a glimpse at these things happening between the AIs, possibly due to a spy network, or maybe an advisor or news feed.


    The system would be for all AI's (Dragons, independent towns, roaming monsters, lairs, merchants, etc).

  3. #13
    Wishing you the best of luck! I've pledged at the "mage" level on kickstart...

    Let me ask about my biggest fear about the game...

    We all know that most games have poor AI. Actually, a lot of older games just had CP (Computer Players) that didn't use any techniques from AI.

    I'm a pretty good programmer. And, I have a degree in computer science. And yet, I know almost nothing about building AIs. Sure, I've played with using things like minimax and A* in toy projects and I've read up on various algorithms and subjects. But, without probably years of research and study, I doubt I could build a competemt AI for a 4X game. I suspect most programmers, even the really good ones are in the same boat. In short, being a good programmer doesn't translate into being able to write an AI. Just like you'd need to know a lot of geology in order to write a seismic processing application, you need to know a lot of the AI field in order to write a good AI.

    This isn't your first game, so obviously you can do some AIs. But, a 4X seems to be a bigger beast than a wargame.

    Competent AIs seem to be the exception rather than the rule (especially for complex 4X games), The same folks that were praised for GalCiv stumbled badly on the first version of Elemental.

    How do you plan to avoid the all-to-common fate of a lousy AI?

    P.S.
    I'm personally uninterested in multi-player, but what about an API similar to what would be used for MP that could be targeted by external AIs? With luck, you'd one day see a google contest or other contest to improve the AIs.

  4. #14
    Relative newbie, I also came here due to the kickstarter. I've toyed with 4x games a lot..and while I generally have issues with long term interest in such game types. I still enjoy the chance to sit back and play them..or dream about playing them or whatever. Basically I'm saying I can definitely appreciate the game type, and I've already signed up for kickstarter in order to help fund the game as i've been looking for a decent version of MoM that'll run on modern computers...without a chance at a remake, this game seems to carry over much of what I want. But lets get to the point of the thread.



    I think part of the idea of making an AI, and as mentioned, make it seem like its an intelligence with a thought process..over just a computer. Is the fact that you need to properly emulate how a player will do things. I think beyond that you need to convince the computer to 'have fun', that is it also needs to be programmed to do things...just because its a 'fun thing' to do, over it being a best strategy. Not all need this, heck some people think the 'most fun' thing to do is play to the best of ones ability. But sometimes the computer needs to be willing to not pick the optimal strategy because it finds that strategy 'boring'. Beyond this...beyond the whole 'everything must be an even number' beyond all of this. There is one thing I find needs to be importantly done. The computer needs to communicate such intentions with the player from time to time. The computer needs to admit to the player 'no this isn't optimal..but I don't feel like being optimal right now'. Not all the time but from time to time.

    Which brings me to a general idea. That I came up with due to this line of thinking..of 'communicating' with the AI. There needs to be an ability to attach a short worded message to an action. Oh I'm attacking this city? Well I click the option to attack, and while its confirming if I wish to do this, it also has an option to attach a short (10 to 20 char) message. So if I'm playing against another player, when they get the save...and they watch the last turns replay, when it shows them the part where I attack their city (which how will combat like that play out over multiplayer....will defenders always be computer controlled?), it'll also show my message attached to the action.

    In the same respect a computer will have a short message list..or generation system. For instance, at one point the 'puter might attack my high level city with a lower level unit. And its doing this because its 'I'm just having fun' algorithm just popped up and said 'lets be wacky and do this' maybe its algorithm actually came up telling it to do this in hopes of provoking a fight with the player, because its 'bored of peace'. At that point the computer randomly decides if it wants to attach a message..it comes up yes, and suddenly while the game shows the computer doing this attack the player suddenly sees the computer go "YOLO" as its message during the attack.

    Little things like this will help breath life into the computer. And for other thoughts..and yes I know such things like this might take awhile to be added....but hey on the backburner....

    Such things like this should probably be a slider...as in how often the computer is willing to do something because 'its fun'. Heck you can name it the YOLO slider if you like.

    But just like you have a difficulty level for AIs. The AIs should also have a 'just having fun' level as well. The lower this is, the less likely it is to make such choices. Or more likely, the less 'oddities' it should have...such as the need for an even number of things (or it controls a percentage of the chance of such things showing up). So someone who wants a tough game without quirks can slide it all the way down. While someone who wants a bit of levity and to play against a more human opponent might turn this slider up.

  5. #15
    Arcane Candidate
    Join Date
    Jul 2013
    Posts
    1

    Lightbulb A Humble and Hopeful Request

    Relating to Just having fun AI, I just wanted to expand on that in that I really hope you do have that. In original MoM I would have a bad day and just want to come home and completely demolish something. Taking out my frustrations by building a huge army of Drakes and Wyrms and Paladins and simply crushing Merlin was a great way to do that. I would often set it to Easy level, kill all the dungeons around Merlin, populate Arcanus and Myrror with dozens of my cities, summon the best creatures I could, surround his cities, blast them with Chaos city-destroying magic, and then finally put him out of his misery. Yes I played the harder levels, but sometimes you really just want to create without having to worry about an AI with intelligence like mine coming after me. I absolutely adore that about MoM (which is why I still play it from time to time care of my DosBox) because it really was just a perfect game no matter your mood. I really really hope you preserve that aspect of it. I'm not programmer savvy so I don't know how, I just trust in your MoM loving ways to make the magic happen and then promise to buy the heck out of it the second it's available.

    So in summary, please preserve that ever-important "Easy" and "Intro" levels because while I am a frequent gamer (we don't even have a TV b/c we're on our computers playing games so much), I also get easily frustrated with a game set at the "easy" level that is by no stretch of the imagination actually easy. I would be forever in your debt. I really just wish you could copy and paste the AI code directly from MoM b/c I loved their set up and my dad loved the challenge of the Impossible level (though the computer struggled to handle all the "thinking" required).

    In great appreciation,
    sussetorte

  6. #16
    Mage of the Inner Tower Endless Rain's Avatar
    Join Date
    Apr 2013
    Location
    RPG.net
    Posts
    532
    Quote Originally Posted by sussetorte View Post
    Relating to Just having fun AI, I just wanted to expand on that in that I really hope you do have that. In original MoM I would have a bad day and just want to come home and completely demolish something. Taking out my frustrations by building a huge army of Drakes and Wyrms and Paladins and simply crushing Merlin was a great way to do that. I would often set it to Easy level, kill all the dungeons around Merlin, populate Arcanus and Myrror with dozens of my cities, summon the best creatures I could, surround his cities, blast them with Chaos city-destroying magic, and then finally put him out of his misery. Yes I played the harder levels, but sometimes you really just want to create without having to worry about an AI with intelligence like mine coming after me. I absolutely adore that about MoM (which is why I still play it from time to time care of my DosBox) because it really was just a perfect game no matter your mood. I really really hope you preserve that aspect of it. I'm not programmer savvy so I don't know how, I just trust in your MoM loving ways to make the magic happen and then promise to buy the heck out of it the second it's available.

    So in summary, please preserve that ever-important "Easy" and "Intro" levels because while I am a frequent gamer (we don't even have a TV b/c we're on our computers playing games so much), I also get easily frustrated with a game set at the "easy" level that is by no stretch of the imagination actually easy. I would be forever in your debt. I really just wish you could copy and paste the AI code directly from MoM b/c I loved their set up and my dad loved the challenge of the Impossible level (though the computer struggled to handle all the "thinking" required).

    In great appreciation,
    sussetorte
    +10000. I am not a good strategy player, so having easy enough difficulty levels is very, very important to me.

    ---------- Post added at 09:42 AM ---------- Previous post was at 09:04 AM ----------

    I finally made it into the tower!
    I am waiting for the relaunch before playing Worlds of Magic, so I don't check the forums as often as I used to. In the meantime, the main forum I post at is RPG.net, if anyone here needs to contact me.

  7. #17
    Acolyte
    Join Date
    Jun 2013
    Location
    Texas
    Posts
    17
    Quote Originally Posted by sussetorte View Post
    So in summary, please preserve that ever-important "Easy" and "Intro" levels because while I am a frequent gamer (we don't even have a TV b/c we're on our computers playing games so much), I also get easily frustrated with a game set at the "easy" level that is by no stretch of the imagination actually easy. I would be forever in your debt. I really just wish you could copy and paste the AI code directly from MoM b/c I loved their set up and my dad loved the challenge of the Impossible level (though the computer struggled to handle all the "thinking" required).

    I would agree, but for a different reason. If a game is somewhat complex with a steep learning curve, you need an easy/novice level to allow the player to learn the ins and outs of the game to better enjoy it, and then to advance to a more difficult level as the players learning curve increases.

    One of the recent 4x space games, Endless Space, has a steep learning curve and even at the easy difficulty, many players complained of the AI stomping them before they can even get a chance to explore the universe. Then the players disparage the game and don't recommend it to others to buy and play. As an old hand at strategy games and still having my original copies of MOM, MOO, and MOO2, amongst others, I had no problem at medium difficulty levels. I would tell these new players, to play on the easiest mode and turn off pirates to reduce the frustration and dumb down the AI so they can learn the mechanics of the game first.

  8. #18
    Quote Originally Posted by Aaron View Post
    Alright, one of the things people have asked about a few times is the AI. How is it going to work in WoM? Are we ready to handle something of such grand complexity? Well, this post is to give you a little more information (albeit in the loosest philosophically terms).

    I want to start by addressing what Artificial Intelligence is. I know that most of you know and that it's fairly obvious, but I have to start somewhere and the beginning seems like a logical place. Artificial Intelligence is Intelligence that is (wait for it) Artificial. It's job (in WoM anyway) is to do what a player would do without a player being there. When you play against the computer in WoM you will be playing against a number of AI opponents. Each of them should have a unique play style. As you play against them you should be able to recognize their patterns of behavior and get to know them just as you would a player. It should leave you with the feeling “That's just how George plays.” not “The computer gets confused if you BLANK.”

    So far, that's pretty straight forward. How does it work? On the most basic level the coder who wrote the AI looks at the current information and makes decisions on what the AI forces should do. A lot of people don't seem to get that when you play against an AI you are actually playing against another person (or team of people). The computer doesn't have a goal, it doesn't want to win, it doesn't want you to enjoy the game, it just computes. This may not seem to matter, but it explains why a lot of AIs fail. The coder who is writing the AI doesn't know how to play the game. That's a problem. It leads to things like just dumping loads of resources into the AI's lap in order to give them a chance to beat a player that can actually PLAY.

    Now, that's one of the advantages our team has. We can play. Not only that, a lot of you can play. So, when we start banging out AI code we're going to be putting together computer opponents that have strategy arcs that span the entire game. As we get feedback from play testers explaining some weakness in a certain opponent's strategy we'll be able to work on that point and strengthen it. Plus, of course, the computer never forgets to check anything. It doesn't make a rough guess at the odds, it knows exactly what they are. It doesn't get tired of checking every city every turn. It doesn't feel any sense of loss when it sacrifices a hero to win the game. All this added together means that we can make opponents that are truly almost impossible to defeat.

    We plan to go beyond the basic “Make the AI play like we would” and attempt to quantify certain strategic concepts like “Expansionist”. The idea is to boil down certain elements of strategy and use them like legos to build a completely new and unique AI that has a solid play style pieced together out of a number of viable play strategies. Now, that is going to be the real trick. It's relatively easy to hand craft a sorcerer with a certain set of skills to play a certain way. It's another thing entirely to be given a set of skills and strategic “methods” and try to dynamically make something that can actually offer a challenge. However, it's something we're aiming for.

    We want the AI in WoM to feel like artificial intelligence not artificial intelligence. It's going to take a lot of work to make it happen, but it's work we're going to love. As with all things WoM related, play testing is going to be a big element. In fact, it's going to matter more with the AI than with any other single element.

    I could go into more detail here, but this post is almost a page long. As you guys know I try to cut it there. (Also it's my sixteenth anniversary, so I have to take the wife out for a bite to eat here in a minute, lol). Anyways, feedback and suggestions. You know the drill.
    Happy Anniversary to you! Seems like you know a lot of stuff about Artificial Intelligence. Great to meet you.

Page 2 of 2 FirstFirst 12

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •  
footer