Random Achaea Thoughts

Naverre

A lot of oversimplification of what makes things easy/hard to automate, that is, create an agent that can successfully compete/beat humans in Achaean combat, with no human intervention other than pushing the first button. And seeing as I'm bored and like to play around with AI, I thought I'd throw this in here. Spoiler tag, because it's long and some people just don't have the time/interest to read it all. This is the random thoughts thread.

There are 7 basic traits an environment dictate what type of agent you'll be building, and how hard it would be to do so. Don't take this all too seriously, it's pretty informal and I'm only an AI hobbyist.

Known/Unknown: Whether the agent (basically, the designer of the agent) knows all the of the rules (or laws of physics, etc) of the environment. Chess is easy: what can each piece do, when is a piece captured, when is a king in check, what can you do when you're in check, what are the rules for checkmate, what are the rules for en passant, etc.
If you're making an AI to fight for you in Achaea, we're going to assume you know all of the cures and what they do, the afflictions and what they do, the abilities and what they do, and the various commands for everything. If you don't, you shouldn't be making one. Known.

Fully observable/Partially observable: Whether or not all relevant information can be obtained at one single point of time; the AI can can knowledge of the entire state of the environment in a single instance. This is like the chess vs. poker scenario. In chess (and Go), you can see all the pieces on the board; all the information about the state of the game can be read in once instance. In poker, you can't see your opponents hand or the order of the cards in the deck.
Achaean is not observable. There will be times when your opponent has afflictions and you won't know what they are, or what that kelp they just ate cured if they have both clumsiness and asthma, nevermind tree tattoos or random heal stuff like rites/harmonics. You also do not know what herbs/vials they have in their inventory for certain. Probably a couple other things, too. Partially observable.

Deterministic/Stochastic: Whether an action will always have the same result  in the exact same situation. No random factors, i.e. rolling dice in a board game.
Unfortunately, Achaea fails this one as well. Herb cures will randomly heal a suitable affliction if you have more than one for that cure. Other abilities that heal at random, such as focus, shrugging, alleviate, and the appropriate rune/harmonic/rite also heal afflictions randomly. Given variance in ping times and the occasional lag spike, even normally deterministic actions can have non-deterministic balances. Also, correct me if I'm wrong but damage against adventurers is also a bit random. Stochastic.

Episodic/Sequential: a problem is episodic if the agent gains in all of the given information, performs and action, and the environment is then reset to its previous state. This way, past actions have no effect on the current action being decided, and the current action is not decided with the concept of planning for future actions.
Achaea is definitely not episodic. The idea of building up afflictions, abilities that require previous conditions, the building of shin/kai/arcane power/etc., breaking limbs... All violate this concept. Sequential.

Static/Dynamic: the way the environment acts (the rules of the environment) do not change over time. Basically, while an agent is thinking about its next move, the environment doesn't change. Games like chess, GO, solitaire, monopoly, etc. are all static. Chess with a clock is only partially static, because the time, part of the game, will change while deliberating, but the rest of the game won't. Even a single-agent problem can be dynamic, if the "dumb" factors of the environment change while you're trying to make a decision.
Achaea is not static. While sitting there thinking whether you should next inflict with paralysis or clumsiness, your opponent smacks you with stupidity, and that changes what you might want to do. Now, a computer can probably think fast enough to be able to treat it almost like it were static, but it's not. If you decide to check the next fifty possible states deep, that could take a while, and you could be hit in the meantime, causing you to have to start over. Dynamic.

Discrete/Continuous: can enumerate all possible situations. Chess and Go are both technically discrete, the problem is that there are so many, it's not a viable strategy to enumerate all of them. However, enumerating a small set of possible states from your current position is possible in discrete environment, and has been done for chess AI's: the next ten/twenty/thirty levels deep are enumerated, and the best path chosen from that, even if doesn't actually reach the bottom of the possibilities. This doesn't work too well with Go, though, since Go has a much wider tree of legal moves than chess.
Achaea is in that fuzzy area where you might be able to use both discrete and continuous approaches. Each "state" of the player technically gives way to a discrete solution. However, the role of time in the problem, differing balances for changing states and the fact that those balances are slightly randomized due to latency, mean that it would probably be better to treat Achaea as continuous. This is a debatable point, but I'm fairly confident in my view. Continuous.

Single-agent/Multi-agent: is there more than one decision agent in the environment, and if so, will those agents change behavior based on the other agents' actions? Games are easily divided into two groups by this trait.
Achaea is definitely NOT single-agent. Your opponent makes his/her own decisions and will change those decisions based on what you have done, both offensively and defensively. Multi-agent.

Given the type of environment that Achaea has compared to Go (Observable, Deterministic, Static, and Discrete), comparisons between the two are unfair. They require different types of agents, and the agents for Achaea would be, in theory, more difficult to develop than one for Go. Specific algorithmic problems (like the ones with Go) might adjust the difficulty, and in all likelihood, we'd probably see a perfect Achaean combat AI before a perfect Go AI, but I highly doubt we'd see either soon.

As for whether we'd get automation that can beat humans, that could easily happen sometime soon. Achaea's interface is much better suited for a computer to interact with than a human, both in the format (text input) and timing (computers think/react faster than humans). These advantages alone mean that even a rudimentary AI would be able to outperform a human in fast-paced combat.

Ayami

Tael said:

Uncertainty does not preclude automation in any way. It isn't particularly hard to write an "AI" that takes into account uncertainty and Achaean combat is definitely amenable to full automation. Having Aetolia-style afflictions and third-person curing messages certainly makes it easier, but the uncertainty in Achaea's combat system is easy to quantify and easy to incorporate into an automated systems.

Having "psychologically random" things doesn't really change much with respect to automation either. Either people are changing their offence and defence reactively, which is plenty easy to write an AI to do, or you have a rock-paper-scissors scenario where the changes aren't in reaction to particular states and an AI transitioning randomly will perform equally as well as a human (or if you want to get fancy, tracking the statistics of the state transitions and using that, since people probably don't actually play rock-paper-scissors, or any kind of analogous game, by playing perfectly randomly).

Games that computers can't play well are pretty exclusively games with ridiculously many permutations and extremely chaotic systems (in the sense that small changes can lead to very large differences in outcomes) - Go is a good example since (1) there are far, far too many possible moves to brute-force a solution for the vast majority of the game and (2) a single move can, at almost any point, completely change the state of the game. The problem with Go is that a small move can have huge consequences several moves down the line, but no AI can evaluate enough moves deep to see those consequences and work back to what small moves to make. Put another way, computers are bad at Go because we don't really understand how humans manage to be good at Go. Humans don't have the computational resources to brute force the game either, so expert Go players must ultimately be employing some sort of more-local algorithm, we just don't know what it is. Perfect knowledge or lack thereof is not what makes the game (or any game) difficult to write AIs for - relative uncertainty is going to be the same for players as for automated systems. And that's the sort of uncertainty where the alternatives don't necessarily have equal chance of selection, whereas things like uncertainty about herb cures can be quantified using simple percentages. Crucially though, what matters is that Achaean combat is supremely analytic - one of the reasons a lot of people are drawn to it - people do things for generally clear reasons and not in any way the sort of reasons that are difficult to emulate with a computer program. Even if people are changing their strategy simply for the sake of changing it, not in reaction to a particular state, that's an easy thing to model and implement.

The reason we haven't seen more full automation is because the people who are most knowledgeable about combat either don't have the programming ability (not that it would require anything particularly fancy) or don't have the interest to develop fully automated systems, and the people with the ability and interest haven't historically been knowledgeable enough about combat to generate anything worthwhile. But it is absolutely, unquestionably possible.

You're right, uncertainty doesn't preclude automation. Neither does having an absurd number of permutations, though. I mentioned permutations passingly when I talked about no-limit vs limit holdem.  Both are simply things that will delay a 'perfect' automated system from being created when set against a) current computing power and b) the technical ability for people to create complex AIs.

In the long enough term, there's literally nothing that you can do to prevent automation of any game at all. But chaotic systems and extreme uncertainty are the biggest hurdles to current AI creation, so the best we can do is introduce enough of those two elements that a proper AI won't come around in a time period significant to us. You're completely right that Achaea, as it is, is easily automated for most classes and only slightly more challenging for others, though.

Sarapis

Great post, @Naverre.

Jukilian

IMO make a separate thread for the automation discussion.

Tael

Uncertainty and the number of possible permutations raise very different problems - uncertainty is something you can simply quantify. A computer can easily do the same thing with the uncertainty that a person can, and there's nothing computationally expensive or particularly complicated about it. There is a theoretical computer that can brute-force Go in reasonable time, but I'm not even sure such a computer is possible to construct given fundamental physical limitations - the number of possible games of Go is literally astronomical (and the problem is that we don't know what strategies humans are using to reduce that problem space - we absolutely know how to reduce that space in Achaea to the point where it is very, very tractable in real-time). But any computer can quantify uncertainty - that's just a question of whether a programmer is familiar enough with the relevant mathematics and, in terms of the kind of uncertainty we have in Achaean combat, the relevant mathematics are pretty trivial. Making a one-button offense that you can press and then go make some popcorn, an offense like that that can outperform human players at a similar level of combat knowledge as its author, would not be difficult. That can be done right now. Adding more uncertainty wouldn't make it any harder either since any additional uncertainty affects players and AIs equally. If there is a 25% chance that the opponent has cured claustrophobia, there isn't really anything a human can do with that knowledge that an AI can't - and reducing it to a 15% chance doesn't change that.

Put another way: combatants right now are not playing "intuitively". There aren't typically any circumstances in Achaea where a combatant can't explain in very specific terms why they did something. When they lose and go look at a log, they can typically figure out what they should have done differently, the point at which they should have made a different decision given the state at that point - and it's easy to script that modified reaction. While there are a ton of possible states and transitions between them at just about any point, players have already done the work of identifying the important transitions.

A lot of the discussion here has been aimed at what writing a naive AI for Achaea would be like, but that isn't the relevant problem - all an AI for Achaea has to do to be better than human players is implement the same strategies they're already using, only more reliably. AIs can easily beat humans without solving or brute forcing a problem. The problem with games like Go is that no one knows what strategies expert humans players use in concrete enough terms to implement them. That just isn't the case in Achaea.

It's true that a system that can develop novel, more effective strategies in Achaea would probably be quite hard to write, and is probably beyond the capabilities of any players, for many of the reasons you mention, but the strategies human players already use in Achaea are, computationally, a lot simpler and a lot better defined than in many other games. If he wanted to and had a lot of spare time on his hands, there is no particular reason Santar couldn't automate his entire thought process to fire off of a single button and simply make his existing strategy fully automated and more reliable. Uncertainty and stochasticity make the problem slightly harder, but the kinds and amount of uncertainty and stochasticity you have to deal with for Achaea are minimal - they're easy to quantify and very tractable to model in real-time. Solving Achaean combat is probably very difficult, but automating the entirity of the strategy a given person already employs, making it approximately as effective and far more reliable than that human, would not be very difficult (though it would probably be pretty tedious).

As an actual random Achaea thought: Tutors have gotten kind of weird since autoclass, and even weirder since the move to city-based tutors over house-based tutors. Maric has the knowledge and skills of some sort of God.

Tharvis

soon, @Tael, Soon the city tutors will take over the world!

Ayami

Tael said:

Uncertainty and the number of possible permutations raise very different problems - uncertainty is something you can simply quantify. A computer can easily do the same thing with the uncertainty that a person can, and there's nothing computationally expensive or particularly complicated about it. There is a theoretical computer that can brute-force Go in reasonable time, but I'm not even sure such a computer is possible to construct given fundamental physical limitations - the number of possible games of Go is literally astronomical (and the problem is that we don't know what strategies humans are using to reduce that problem space - we absolutely know how to reduce that space in Achaea to the point where it is very, very tractable in real-time). But any computer can quantify uncertainty - that's just a question of whether a programmer is familiar enough with the relevant mathematics and, in terms of the kind of uncertainty we have in Achaean combat, the relevant mathematics are pretty trivial. Making a one-button offense that you can press and then go make some popcorn, an offense like that that can outperform human players at a similar level of combat knowledge as its author, would not be difficult. That can be done right now. Adding more uncertainty wouldn't make it any harder either since any additional uncertainty affects players and AIs equally. If there is a 25% chance that the opponent has cured claustrophobia, there isn't really anything a human can do with that knowledge that an AI can't - and reducing it to a 15% chance doesn't change that.

Put another way: combatants right now are not playing "intuitively". There aren't typically any circumstances in Achaea where a combatant can't explain in very specific terms why they did something. When they lose and go look at a log, they can typically figure out what they should have done differently, the point at which they should have made a different decision given the state at that point - and it's easy to script that modified reaction. While there are a ton of possible states and transitions between them at just about any point, players have already done the work of identifying the important transitions.

A lot of the discussion here has been aimed at what writing a naive AI for Achaea would be like, but that isn't the relevant problem - all an AI for Achaea has to do to be better than human players is implement the same strategies they're already using, only more reliably. AIs can easily beat humans without solving or brute forcing a problem. The problem with games like Go is that no one knows what strategies expert humans players use in concrete enough terms to implement them. That just isn't the case in Achaea.

It's true that a system that can develop novel, more effective strategies in Achaea would probably be quite hard to write, and is probably beyond the capabilities of any players, for many of the reasons you mention, but the strategies human players already use in Achaea are, computationally, a lot simpler and a lot better defined than in many other games. If he wanted to and had a lot of spare time on his hands, there is no particular reason Santar couldn't automate his entire thought process to fire off of a single button and simply make his existing strategy fully automated and more reliable. Uncertainty and stochasticity make the problem slightly harder, but the kinds and amount of uncertainty and stochasticity you have to deal with for Achaea are minimal - they're easy to quantify and very tractable to model in real-time. Solving Achaean combat is probably very difficult, but automating the entirity of the strategy a given person already employs, making it approximately as effective and far more reliable than that human, would not be very difficult (though it would probably be pretty tedious).

As an actual random Achaea thought: Tutors have gotten kind of weird since autoclass, and even weirder since the move to city-based tutors over house-based tutors. Maric has the knowledge and skills of some sort of God.

Naverre covered what I'm talking about better than I could in his fully/partially observable part. We essentially agree though, Achaea as  it is right now is trivial to write a full automated offense for. I was under no impression that Achaea's current uncertainty/etc is at a level where it'd be difficult to automate, I was simply saying increasing the complexity an AI would need would be a good idea.

Exelethril

It's pretty annoying to fight against someone that has the capacity to talk in full sentences while their "flawless" offense goes ham on you tbh. I just want auto-occie to go away.

Jules

Ayami said:

Naverre covered what I'm talking about better than I could in his fully/partially observable part. We essentially agree though, Achaea as  it is right now is trivial to write a full automated offense for. I was under no impression that Achaea's current uncertainty/etc is at a level where it'd be difficult to automate, I was simply saying increasing the complexity an AI would need would be a good idea.

That's what has been so weird about the past page and a half or so.  Everyone seems to agree with the basic idea you brought up, which was "there are a few things that (some) humans still do better than computers (with certain types of poker exemplifying those things), so if you incorporate more of those things into the game, you help (some) humans have an advantage over things that are automated, so it encourages people to back off automation somewhat", but even with everyone seeming to understand and agree on that basic point... from there it all seems to unravel a bit with people partly misunderstanding and partly actually disagreeing with each other - often in fairly long posts (probably somewhat unavoidable, but the forums have a pretty low tolerance for long posts, and there's also a very normal tendency to not read multiple lengthy posts as carefully). 

Anyway, the basic idea you brought up makes sense, asking the question "what things are (some) humans still better at than any computer program"?  That said, there are fewer and fewer things that humans are better at than a well-programmed application, and the subset of people who truly excel at those things is small, too. 

Antonius

Is the problem really that it's possible to automate things or that it's possible to automate things that (most) players can't do manually? Would people, such as @Herenicus, have such an issue with losing to somebody using a script if it was just as easy - or easier, depending on your level of coding ability - to manually do what the script does as it was to write the script?

The "need" for automation may be a barrier to entry, and may discourage new players, but I'm not convinced that it discourages them any more than learning that they'd need to put in months, if not years, of practice to become a truly competent fighter. Most players may not be capable of creating anything more complex than a trigger, and a lot of them may not be willing to learn how to do more than that, but I think it's been continuously demonstrated that the "haves" are more than willing to share with the "have nots" when it comes to coding.

When we originally asked for GMCP affliction data, it was denied on the grounds that it further increased the gap between those who do and don't have the ability to code. What happened? Nemutaur went away and built a system that tracks afflictions based on some logic and also what server-side curing attempts to cure (I think, I could be misremembering how he explained it), so the gap was already there. Then he gave it away for free to anyone who wanted it, which meant that the gap suddenly disappeared. All refusing to give us GMCP aff data did was make it harder for him - and anybody else who may have been inclined - to do so.

Honestly seems like it would be easier, and more beneficial, to make manualling easier than to make automating harder. Those who enjoy automation can continue to do so, and those who can't/don't/won't should be able to compete at much the same level as those who do.

Ayami

I assume we all agree that the ship has long sailed on automating defense, and the only issue becomes 'do we also want offense to become automated?' If both are automated what's the point of people performing combat at all, instead of doing rock-paper-scissors to determine the victory? Or, like other roleplay games, simply roll dice? The simpler you make the game, usually, the more advantage automation has until you reach the point where combat is simply dnd-style roll dice + add modifiers and decide who wins based on that. At that point, sure, automation is no longer an issue. For a lot of people, the fact that automating would be the ideal isn't as big a deal as it's simply boring to fight, essentially, an npc in a multiplayer game. If I want to fight an AI, why don't I just play Baldur's Gate?

There's of course the option of simply doing both: make manualing simpler and automation harder at once, which I suggested earlier. That's by far the most challenging approach, though, I think. It'd require sitting down and doing massive redesigns of large parts of Achaean combat.

Austere

edit: disregard

Nakari

Antonius said:

The "need" for automation may be a barrier to entry, and may discourage new players, but I'm not convinced that it discourages them any more than learning that they'd need to put in months, if not years, of practice to become a truly competent fighter. Most players may not be capable of creating anything more complex than a trigger, and a lot of them may not be willing to learn how to do more than that, but I think it's been continuously demonstrated that the "haves" are more than willing to share with the "have nots" when it comes to coding.

This has been my experience exactly. I'm simply not competent at keeping track of/predicting who has what in my head, and even starting to seriously practice that would take hours and hours of spars and learning the affliction list like the back of my hand and I'd probably have to code a bunch of echoes, too. And to be clear, I need to do this, and I plan to do this, but my meagre scripting knowledge means that I can actually fight and have fun in the process. 

I've never been into combat enough to seriously sit down and get competent at fighting, and I'm not going to actually be good until I do, really. But I'm also no good at coding, and yet have found that a very accessible way for me both to actually have fun with the beginnings of pvp, as well as learn more about how combat works.

Tael

Jules said:

Ayami said:

Naverre covered what I'm talking about better than I could in his fully/partially observable part. We essentially agree though, Achaea as  it is right now is trivial to write a full automated offense for. I was under no impression that Achaea's current uncertainty/etc is at a level where it'd be difficult to automate, I was simply saying increasing the complexity an AI would need would be a good idea.

That's what has been so weird about the past page and a half or so.  Everyone seems to agree with the basic idea you brought up, which was "there are a few things that (some) humans still do better than computers (with certain types of poker exemplifying those things), so if you incorporate more of those things into the game, you help (some) humans have an advantage over things that are automated, so it encourages people to back off automation somewhat", but even with everyone seeming to understand and agree on that basic point... from there it all seems to unravel a bit with people partly misunderstanding and partly actually disagreeing with each other - often in fairly long posts (probably somewhat unavoidable, but the forums have a pretty low tolerance for long posts, and there's also a very normal tendency to not read multiple lengthy posts as carefully). 

Anyway, the basic idea you brought up makes sense, asking the question "what things are (some) humans still better at than any computer program"?  That said, there are fewer and fewer things that humans are better at than a well-programmed application, and the subset of people who truly excel at those things is small, too. 

Just to be clear since I don't think my earlier rambling really was, while I agree that there are things that make automation much harder, the point I was trying to disagree with is that uncertainty (the kind of uncertainty in Achaea, specifically the variety that started this whole conversation - the kind that Aetolia doesn't possess) is one of them. That's actually a very easy thing to deal with when automating things. If Aetolia adopted Achaea's herb-curing, it would be pretty trivial to optimise these undesirable, one-button systems - they would be less effective, but so would players, the relative effectiveness of players and this sort of automation would be unchanged (if anything, I'd bet it would make player reaction times even slower relative to the automation).

The other point I was trying to disagree with, which I think you're getting at a little here, is that you could necessarily add things that make automation more difficult. There aren't many things that humans are better at than computers. Most things that humans are better at than computers are going to be very, very hard to implement in Achaea. Humans are better at a lot of visual tasks, but we only have text. Humans are better at semantic tasks, but the inverse of that is that you can't automate creation of these problems or automate verification of answers, so a human would ultimately have to create each challenge-response by hand, and the manpower of combat designers is never going to equal the manpower of players, who could just solve all of the tasks and then code every response into the system. You could maybe create a scheme similar to what captcha uses to avoid that problem, where other people's answers serve as the verification, but either way, it's not going to be easy and I'm not sure Achaea has the playerbase to actually allow captcha-style answers to converge. And even if it turns out that we have the playerbase for them to converge, I'm not sure what kind of captcha system you could implement with only text - it's pretty telling that captcha systems overwhelmingly rely on the difficulties inherent to computer vision to prevent automation.

The other thing humans are better at than computers, in general, is dealing with very many conflicting constraint satisfaction problems at once. That's the thing that makes full, naive automation of Achaea very, very hard. These sorts of problems get very, very daunting, which is why many complex games that want to implement player-like AIs actually build a bunch of sub-AIs that attempt to satisfy individual constraints and then use a blackboard AI to actually make decisions on the basis of the output of the sub-AIs. And expert human players are just about always better than blackboard systems (when such AIs beat humans, it's usually because, even though they're dumber, their faster reactions give them too large an edge). But Achaean combat doesn't actually involve very complex decisions - there really is a very clear "right answer" the vast majority of the time. Writing a naive AI to find that answer would be a herculean task, but writing an AI to emulate a player's combat decisions would not be nearly so difficult.

Aaaand I guess that was sort of rambling again. Sorry.

Jukilian

Seriously, can we move the automation discussion elsewhere?

I'm not convinced that it's on topic, really.

Kuy

Let us all take a moment to breathe in just how crazy Magi has become after this round of classleads.

*inhale*

*die*

Trilliana

While this could probably be in rants, my random Achaea thought lately has been the fact that I'm disliking how it seems that the game is becoming factionized, fraternization is frowned upon everywhere, and extremely combat heavy (Mainly with this month's promo, but meh, more reason to not buy credits for me)

And what I mean by factionized is what's going on for a few years now. Occies being called back to - for the most part- Ashtan, Necromancy to Mhaldor, Forestals back to Eleusis, and now with this fight between Cyrene and Targossas. 

On one hand, I like the idea that Cyrene could be true neutral, but in the end, they'll still be considered neutral-good because that's just so firmly a part of Cyrene that I doubt it would ever change. 

And @Antidas it wasn't -you- per say, it was also how things were dealt with. But, that's my opinion. 

Exelethril

I like factionalization, makes things less convoluted. This gives each player a distinct identity to choose from/play in rather than a patchwork of ideals that don't quite match up but are allowed to co-exist because of personal opinions.

Trilliana

Exelethril said:

I like factionalization, makes things less convoluted. This gives each player a distinct identity to choose from/play in rather than a patchwork of ideals that don't quite match up but are allowed to co-exist because of personal opinions.

Erm... Trilliana/I have a distinct identity. It takes me everywhere but Mhaldor, but even so, I do have a lot of conversations with Mhaldorians. The fraternization makes me not want to leave Cyrene, which closes off a lot of potential roleplay out of discomfort. I DON'T like the idea that I'm closed off from people that could lead to growth of my character and honestly I felt that way while in Eleusis. I was an outsider even though I'd joined Eleusis. It's off-putting, frustrating, and in the end damned boring. 

Kandra

I feel like factionalisation would be better as all or nothing. As it is, those cities that have a distinct faction with classes tied to it (eleusis-forest, targossas-devotion, mhaldor-necromancy, ashtan-occult) have a draw in that alone, while cyrene and hashan have... nothing to offer in that regard but restrictions based on not being the primary city of one of those factions. Could argue bards and runewardens for cyrene, and shamans for hashan, but those classes aren't actually factionalised and can exist in any city.

Aepas

I would like factionalisation more if every city had one unique class, and the rest were neutral.
Since Eleusis pretty much took the forests, that is -3 classes to every city. add necromancy and devotion and that leads to -7 classes for some cities. Occie makes it -8 classes in cities with no real alliance. Granted there are exceptions and the few people who exist outside of the norm, but this seems to be the way things work right now.

Exelethril

Aepas said:

I would like factionalisation more if every city had one unique class, and the rest were neutral.
Since Eleusis pretty much took the forests, that is -3 classes to every city. add necromancy and devotion and that leads to -7 classes for some cities. Occie makes it -8 classes in cities with no real alliance. Granted there are exceptions and the few people who exist outside of the norm, but this seems to be the way things work right now.

Could make Cyrene the city of arts/culture(Scarlatti) and give bard to Cyrene. Much potential rage   

As for Hashan.. not really sure what fits into their culture.

Tael

I think factionalisation in the game is great. It drives conflict, makes organisations more unique, and leads to more interesting and unique driving philosophies that make Achaean organisations distinct from generic fantasy organisations.

I think all of the "anti-fraternisation" nonsense the factions keep insisting on is downright cancerous in a game like Achaea. All it ends up doing is driving people to interact with their friends OOCly instead of ICly. It's a lazy, "zero tolerance" method of enforcing factional allegiance that lets people avoid the responsibility of having to actually pay attention to whether characters are staying true to their factional obligations, instead simply checking WHO to see if they spend too much time standing near people who aren't in the faction. It's yet more pearl-cluthing over the possibility that anyone might be betraying their faction, pretending as if removing those people "before they do any real damage" is actually a problem, when in fact any "damage" the average player can do to an org they're leaving is extremely limited and, if anything, makes the game's narrative more interesting, not less.

Tahquil

runewardens recalled to Cyrene. All cities left defenceless.

Nim

Not defenseless.

More like...

Neutrality insurance.

Jukilian

Tahquil said:

runewardens recalled to Cyrene. All cities left defenceless.

Shaman.

Edit: so basically just Targossas left defenseless.

Aerek

Sure, sure, use Shaman totems. The cheap Chinese knockoffs of real totems.

Or come pay Papa for the good stuff.

Aktillum

Exelethril said:

Aepas said:

I would like factionalisation more if every city had one unique class, and the rest were neutral.
Since Eleusis pretty much took the forests, that is -3 classes to every city. add necromancy and devotion and that leads to -7 classes for some cities. Occie makes it -8 classes in cities with no real alliance. Granted there are exceptions and the few people who exist outside of the norm, but this seems to be the way things work right now.

Could make Cyrene the city of arts/culture(Scarlatti) and give bard to Cyrene. Much potential rage   

Yes lets give one of the most complex PvP classes to the faction least interested in PvP.
But the minute they start ex-harmonicsing CIJ bards, Quackers and me are rollin' up on Blu like a mobsquad

Morthif

Would @Xinna go to Cyrene?

Exelethril

Aktillum said:

Exelethril said:

Aepas said:

I would like factionalisation more if every city had one unique class, and the rest were neutral.
Since Eleusis pretty much took the forests, that is -3 classes to every city. add necromancy and devotion and that leads to -7 classes for some cities. Occie makes it -8 classes in cities with no real alliance. Granted there are exceptions and the few people who exist outside of the norm, but this seems to be the way things work right now.

Could make Cyrene the city of arts/culture(Scarlatti) and give bard to Cyrene. Much potential rage   

Yes lets give one of the most complex PvP classes to the faction least interested in PvP.
But the minute they start ex-harmonicsing CIJ bards, Quackers and me are rollin' up on Blu like a mobsquad

Well, a genius programmer might really like Cyrene and script an auto-offense for bards, so they will press F1 and kill you while playing on a tablet while talking to you in full sentences.