Relevance of the player

How relevant should the player be to the world at large? Are they just another mercenary going out and fighting monsters? Do they have the potential to be the grand hero who shapes the world around them?
«13

Comments

  • I would assume the player is exactly the same as any other NPC. Hopefully the AI will be good enough that the player will be decently challenged; this means that while you, as the player, could potentially become a grand savior of the world (or a wicked evil villain set on world domination), any other NPC has the potential to become this too - while only a select few would be able to succeed.

    Unfortunately, this also means that the player could "snowball" momentum - basically, once you get a really good player going, it would be very difficult to die. A good means of counteracting this would be for powerful NPCs (or the player) to have a price set on their head by the opposite alignment, and this price increases with time. This would see more and more higher-level NPCs coming to track you down. Then again, this could get tedious... it's a difficult problem, as I see it, but could be worked out in the end. It's going to need balancing no matter what you do, if you want it to be fun as a game.
  • Mystify said:

    How relevant should the player be to the world at large? Are they just another mercenary going out and fighting monsters? Do they have the potential to be the grand hero who shapes the world around them?

    You are only as relevant as you make your self to be. Just like real life, there is no guarantee to fame, fortune, or glory. The world is not player-centric. You play one role -- the exact same role that any other NPC can play. While the world might be pre-populated with powerful characters (i.e. kings and such), your starting role will always be modest so as to give you something to work up towards.

  • edited November 2014
    gavanw said:

    You are only as relevant as you make your self to be. Just like real life, there is no guarantee to fame, fortune, or glory. The world is not player-centric. You play one role -- the exact same role that any other NPC can play. While the world might be pre-populated with powerful characters (i.e. kings and such), your starting role will always be modest so as to give you something to work up towards.

    Sorry if you've talked about this elsewhere, but how does the player "advance"? There are plenty of options--completing quests, doing stuff for people, gaining experience in particular crafts or combat experience, maybe terrifying people into submission. You mentioned "Emergent stories/plots/quests" on the Kickstarter page. Will NPC characters be clever enough to take advantage of these and deliver what other people need, and will your relationship (or lack thereof) with a "quest-giver" affect your ability to provide what he or she needs?
  • VoxelQuest may allow people to do what few games allow- assassinate the king and see the changes in the world.
  • Or be the king who's being assassinated, if the NPCs decide they're tired of having you around :P
  • gavanw said:

    ..You are only as relevant as you make your self to be. Just like real life, there is no guarantee to fame, fortune, or glory. The world is not player-centric. You play one role -- the exact same role that any other NPC can play. While the world might be pre-populated with powerful characters (i.e. kings and such), your starting role will always be modest so as to give you something to work up towards.

    I like this.. I think there are enough 'sole hero saves the day' games out there already.
  • edited November 2014


    Sorry if you've talked about this elsewhere, but how does the player "advance"? There are plenty of options--completing quests, doing stuff for people, gaining experience in particular crafts or combat experience, maybe terrifying people into submission. You mentioned "Emergent stories/plots/quests" on the Kickstarter page. Will NPC characters be clever enough to take advantage of these and deliver what other people need, and will your relationship (or lack thereof) with a "quest-giver" affect your ability to provide what he or she needs?

    There are a number of ways you can advance, some of them are similar to how it would function in "real life." For example, better military rank: complete many quests (i.e. missions) given to you by your superior officers. As ranks become harder to get, maybe you only take over the rank when your superior dies (or is assassinated - heheheh). You could assume political rank perhaps from your bloodline, or clawing your way to the top of the political ladder (marrying into a royal bloodline, throwing a coupe, etc). These things are "built in" to the logic of the game, it does not have to discern what a king is, that is already an abstracted property/rank that users have or don't have. It knows what your bloodline is by any relatives you have. Etc.
  • edited November 2014
    gavanw said:

    There are a number of ways you can advance, some of them are similar to how it would function in "real life." For example, better military rank: complete many quests (i.e. missions) given to you by your superior officers. As ranks become harder to get, maybe you only take over the rank when your superior dies (or is assassinated - heheheh). You could assume political rank perhaps from your bloodline, or clawing your way to the top of the political ladder (marrying into a royal bloodline, throwing a coup, etc). These things are "built in" to the logic of the game, it does not have to discern what a king is, that is already an abstracted property/rank that users have or don't have. It knows what your bloodline is by any relatives you have. Etc.

    This makes it sound as though there is a well-defined social setting for the world of the base game: social/political/economic hierarchies and places demarcated by political boundaries (enforced by the local power structures).

    A story might be written (or assumed) that explains why these things will look the way they do when you start playing. But even without that, there's still a set of power structures.

    How malleable will those structures be? Are they hard-coded with a world-story? Randomly generated as part of each new world? Editable within the game? Editable through flat files or scripts? Editable through the C++ source?
  • The ranks, etc. seem like they'd be the kind of thing that would/should be editable via json
  • gavanw said:

    There are a number of ways you can advance, some of them are similar to how it would function in "real life." For example, better military rank: complete many quests (i.e. missions) given to you by your superior officers. As ranks become harder to get, maybe you only take over the rank when your superior dies (or is assassinated - heheheh). You could assume political rank perhaps from your bloodline, or clawing your way to the top of the political ladder (marrying into a royal bloodline, throwing a coup, etc). These things are "built in" to the logic of the game, it does not have to discern what a king is, that is already an abstracted property/rank that users have or don't have. It knows what your bloodline is by any relatives you have. Etc.

    This makes it sound as though there is a well-defined social setting for the world of the base game: social/political/economic hierarchies and places demarcated by political boundaries (enforced by the local power structures).

    A story might be written (or assumed) that explains why these things will look the way they do when you start playing. But even without that, there's still a set of power structures.

    How malleable will those structures be? Are they hard-coded with a world-story? Randomly generated as part of each new world? Editable within the game? Editable through flat files or scripts? Editable through the C++ source?
    In general, any given property is not hard-coded into the engine, but some functions might be, for example - a distance/proximity function. Most of this stuff will be laid out in JSON (or perhaps a superset of JSON if needed) and can be hotloaded in realtime. To some degree, it makes sense to define your own laws (for your province/town/whatever) in the game, so this could just as well be accompanied by an arbitrary power structure i.e.:
    morePowerful(King,Lord), morePowerful(Lord,Mayor) - by transitivity, a King is more powerful than a Mayor (although for the purposes of speed, rank could just as well be tied to a single number value).

  • Thanks, Gavan. I ask because this gets into several pretty deep areas related to core game design.

    One is that, while encoding the details as JSON will make them (somewhat) configurable, it still sounds like it will be the case that "hierarchical power structures" will be baked into the world of the game. I neither agree nor disagree with that; I'm just curious to see the consequences of that design decision on the mechanics/dynamics/aesthetics of the rest of the game. (I guess I have to add I'm also not commenting on "power structures" to score any real-world political points. There are interesting discussions to be had on what it means to import that stuff into games, but this almost certainly is not the place for those conversations. Just wanted to be clear: I'm talking about games here.)

    Secondly, if the forms of these power structures are static once they're loaded into the game at start-up time, does that imply that the entire world of the game operates under the original definition of a "class" system? (As in, if you start life as a farmer, a farmer is all you can ever be -- maybe an increasingly capable farmer, but never a guard or a merchant or a lord.) It sounds like the game's rules for power structures may cause it (again, please don't read any value judgments into my comments here) to produce a world that has a similar "you can't ever be anything than what you started out as" effect as the real-world class system. Questions about that would be: does this power structure design tie in to the RPG definition of "classes" as a mechanical device in games for enabling player progression? Or can the player switch between classes? Or can NPCs change classes? Or are you thinking of these more as titles, such that any character (or just the player's character?) can learn any individual abilities and can simultaneously be a Journeyman Baker, Master Ranger, Novitiate Priest, and Baron?

    Finally, what does the existence of hierarchical power structures imply about storytelling in the game? If you have a world, and the world has simulated people in it, and those people perform actions that are to some extent based on power relationships with each other, that generates questions: how did the world come to look the way it does, with those particular power structures? What kinds of stories are possible in a world where those structures cannot ever be changed (even if some of the people filling the available slots change)? What stories are possible if those structures can be changed? Do these structures enable a developer-defined Main Quest story even if each new world is randomly generated? Or can small stories be themselves randomly generated from the particular arrangements of power when a game world is created? Or is "story" going to be something players generate entirely in their own minds based on what happens to them in the game they play?

    This kind of stuff -- which you get the instant you decide to make a game that's a world with people in it -- is why so many developers throw up their hands and just make a Flappy Bird clone. ;)
  • Well, the class system is probably going to be skill-based (although as mentioned I'm supporting both). Nonetheless, jobs/ranks/etc are not always tied to the way you develop your class. You could be a hunter and decide to switch to a farmer, the main consequence being that you have spent points in being a hunter and will never be "quite as good" of a farmer, had you started out as a farmer. Your rank could be completely outside of your skillset. Imagine, for example, that you are in a tribe, and the only thing that qualifies you to be leader is that no one else has killed you (think: alpha wolf). If you go out and kill the leader, you are suddenly the new leader, regardless of how skilled you are, what your profession is, etc.

    So, in shorter words, there probably won't be many artificial restrictions - only some restrictions to prevent uber-classing (i.e. limited amount of skill points to distribute between a larger pool of skills).
  • gavanw said:

    Well, the class system is probably going to be skill-based (although as mentioned I'm supporting both)...You could be a hunter and decide to switch to a farmer, the main consequence being that you have spent points in being a hunter and will never be "quite as good" of a farmer, had you started out as a farmer.

    I like how this encourages you to team up with NPCs whose skills complement your own, especially if you're going to implement a party system.

    Speaking of parties: will you control all of the characters in a party, or will NPCs just have a tendency to band together (with you, occasionally) and you're forced to balance their goals with reaching your own? The latter sounds like a cool thing to do with the AI.

  • Speaking of parties: will you control all of the characters in a party, or will NPCs just have a tendency to band together (with you, occasionally) and you're forced to balance their goals with reaching your own? The latter sounds like a cool thing to do with the AI.

    I am making it to be either way, based on player settings. Some people will prefer to micromanage their party, others might prefer to make it feel like they are controlling just one person.

  • I rather like that... I'm going to make a thread on parties in either case, I think. If they're guaranteed for addition, I'm sure the community would be able to give some interesting feedback on it, and perhaps a few insights you might not have thought of yourself. :)
  • gavanw said:


    I am making it to be either way, based on player settings. Some people will prefer to micromanage their party, others might prefer to make it feel like they are controlling just one person.

    I feel like the person micromanaging will have a huge advantage compared to the latter person, which would make balancing tricky.
  • Not really. Aim to balance for the person who micromanages. If they don't want to micromanage, that's great - they'll be at a disadvantage, though the gameplay might be faster-paced and more exciting for them. That's the sort of thing they'll like if they don't want to micromanage anyway.
  • I strongly disagree with that design. They shouldn't be put at a disadvantage for liking faster-paced gameplay. Gaven said there is only going to be one difficulty, so there isn't going to be an easier mode where you can get by with AI controlled allies and a harder mode that will require all of your skill and tactics coordinated together to win.
  • edited November 2014
    Micromanagement, as a rule, makes gameplay easier. The reason for this is that you're putting more effort into it - putting in extra effort yields greater rewards. It's a sort of balance in itself. If you're not willing to put in the extra effort, your efficacy is of course going to suffer.

    Micromanagement <-------> Fast-paced gameplay
    Effort <------------------> Action
    Efficacy <----------------> Speed
    Thought <----------------> Instinct
    Complexity <-------------> Simplicity
    Difficulty through thought <-> Difficulty through gameplay

    Although they aren't mirrored, it provides two sets of playstyles, each being fun for the associated type of player. It might be difficult to design a system that permits the faster, action-oriented players to still enjoy themselves; this could be accomplished by making micromanagement helpful, but not necessary. The split also gives rise to a much higher variety of decisions for the player to make, increasing their experience while allowing them to keep the game from getting more complex than they're willing to handle.


    I'd be interested to hear any alternatives you might suggest... I can't think of any myself.
  • Its more about the gulf between "AI does whatever it wants" and "full micromanagement". Imagine trying to play starcraft where you control one hero unit and all the other units do what they feel like against someone with full command over everything. Not only are the specifics of what they do out of your hands, but your entire ability to apply tactics and strategy is gutted. The only way I see it being feasible is if the AI's own grasp of tactics is advanced enough that a micro-managing player isn't going to be drastically better, and even then your own character needs to act with the team.
    Either style is fine in and of itself, its when they have to act inside the same world and both be functional that difficulties arise.
  • edited November 2014
    There are ways around this. Different tactics and strategies can be clumped loosely into different "types", which can then be selected, in segments, from a list. For example, one list might contain (partial lists, to be clear):

    Attack
    Counterattack
    Defend
    Flee

    Another might contain:

    Stand ground
    Seek Cover
    Charge
    Flee

    Yet another might contain:

    Magic
    Melee
    Archery
    Magic/Melee

    The exact design of the system might need some work, but the object isn't to remove micromanagement - the object is to permit its simplification so that the player that abhors micromanagement has to do as little as possible.

    I do think that the AI should hopefully be intelligent enough to stand on its own, though, and for what I see as a good reason: You're not the driving force in the world. If the AI can't handle itself in the wilderness, in a game where the NPCs are supposed to be the player's equals, something isn't right.
  • The game should have a deep and rich tactical side and it will be hard to make a system where the AI can take full advantage if it, the player can take more advantage of it but not drastically so, and a single character is at all a valid tactic against groups of enemies
  • edited November 2014
    Not really, I've described it in the AI thread. You can actually design an AI that can teach itself from scratch with relative ease, if you design the system with that in mind. If you give it enough options, it can eventually become as smart as the player, and will have an inherent ability to learn.

    edit: Nope, I just linked to it, it's on the LT forums. http://forums.ltheory.com/viewtopic.php?f=2&t=2990&start=15#p50171 I should write up a version specifically designed for VQ, but the concept is the same - and simpler, in fact, for VQ, because it's not a real-time affair.
    http://voxelquest.vanillaforums.com/discussion/comment/656/#Comment_656 Here's where I came closest to describing it.
  • As good as the player? I doubt that. Even defining all of the relevant factors to consider would be hard- namely the enviromental and positioning factors. You have factors like "I can block this doorway" "standing here will prevent the enemy from reaching my ally in one turn" "if I shove the enemy 2 squares, he will get hit by the mage's next fireball" and similarly "I should move to the side to create an opening for the fireball". "If I move here I can get a clear line of fire, and it will be hard for the opponents to reach me, but they can move to the side to avoid me" "If I dance around the pillars in this room, I can get strikes at the opponent while using my additional speed to get away" "If I move over here, the enemy will chase me, putting him in line with this other enemy, so I can fire a lightning bolt that will hit both of them, rebound off the far wall, then hit them both again" "If I push this enemy, I can give my ally an opening to retreat"

  • @Talvieno
    That was a great read, thanks! I do wonder how complex strategies could get, as the system that you describe seems to be geared towards choosing whether or not to do damage and choosing positions. Would that transfer well to an ability-based combat system, and do you think that all AI characters would tend towards a single fighting style?

    @Mystify
    I share many of your same concerns, especially because if the combat system isn't well-balanced to allow for many playstyles, fighting will devolve into a set routine. If it comes down to simple numbers and constant risk-reward calculation, the computer will win every time. Perhaps the AI shouldn't know how the player's abilities outright and must learn their extent over the course of the encounter.

    Another alternative to the combat system that Talvieno discussed: treat combat like a chess engine treats a chess game. Each AI character knows their own abilities, damage output, and movement ability and can thus predict how much damage they can do per second, and more importantly, how close they need to be to another character to inflict that damage. Calculations would be much more complicated than chess moves due to the ability system (like you said, @Mystify‌, characters can toss fireballs and have to navigate a non-homogenous environment), but the basic principle remains. The AI can calculate the the relative value of each move in terms of how much damage they can inflict and how much they're likely to receive, as well as the relative merit of future moves after each immediate move. Smarter enemies can have more time to calculate their options, while weaker enemies are more prone to make mistakes.

    Granted, this sort of balance is probably hard to achieve from the outset (not to mention there are many different ways of implementing this sort of decision-making). We have a community of programmers and designers and people with great ideas--why not put them to use? We could create a standard arena map and hold contests between different AIs to see which AI behaviors work best for different character classes and monster types. We get information about how different classes match up and how abilities affect gameplay, and we can use that data to tweak the current available ability set. We can run many, many simulations to gather these statistics and improve the AI and thus the gameplay for VQ's final release.

    What do you all think?
  • A chess-like approach seems more feasible, though it would be interesting to see how well it works in practice, there would likely need to be heavy pruning on what sensible options are at any given timestep. Having more than 2 players also complicates it, but the hp gives a much clearer advantage metric to work with.

    An AI contest would be fun. I used to compete in those in college.
  • Mystify said:

    A chess-like approach seems more feasible, though it would be interesting to see how well it works in practice, there would likely need to be heavy pruning on what sensible options are at any given timestep. Having more than 2 players also complicates it, but the hp gives a much clearer advantage metric to work with.

    Yeah, the addition of multiple players does seem it would add to CPU load. And I wonder how difficult it would be to apply this sort of system to each and every NPC/animal/monster/etc that currently exists within the game world.
  • Well, you only need to apply such an AI if there is combat going on, so every creature in the world is not necessary. If the player is not present, it is likely unnecessary to simulate the battle in detail, a quick and dirty simulation can probably uphold the richness of the world without worrying about the precise tactical details of a battle on the other side of the world.
  • Mystify said:

    A chess-like approach seems more feasible, though it would be interesting to see how well it works in practice, there would likely need to be heavy pruning on what sensible options are at any given timestep. Having more than 2 players also complicates it, but the hp gives a much clearer advantage metric to work with.

    Yeah, the addition of multiple players does seem it would add to CPU load. And I wonder how difficult it would be to apply this sort of system to each and every NPC/animal/monster/etc that currently exists within the game world.
    Yep, there has to be some creative pruning at times. As mentioned elsewhere, I'm going to probably brute-force it at first with as small a pool of NPCs as possible, but eventually AI will be pruned by proximity and eventually by relevance.

  • I strongly agree with this approach - it's good to start with a small, manageable group, make that work naturally, and then spread outwards from there. :) This goes for any aspect of game design. I'd say we're on the right track.
Sign In or Register to comment.