Author Topic: About The Recent Schedule, and TLF Updates  (Read 7945 times)

Offline ptarth

  • Arcen Volunteer
  • Hero Member Mark III
  • *****
  • Posts: 1,166
  • I'm probably joking.
Re: About The Recent Schedule, and TLF Updates
« Reply #15 on: June 26, 2014, 01:39:27 am »
  • re:Racial AI
    • And thus we see what happens when I try to be funny. I was suggesting this would be Arcen's next major announcement.
    • There is no comprehensive Racial AI thread.
    • The best is the simple paragraph I wrote up in the Issues Poll
      • Racial Strategic Level Mechanisms and AI
        • Further develop a "mental state" for each race. Allowing them to measure and evaluate the past, present, and future of the system (other races strengths and weaknesses). Then give them the ability to use that information to their own ends.
        • Currently the races are not active participants in the game. They don't have a goal they are seeking towards. They have sometimes end the game by military dominance of the system, but they aren't making choices designed to get them there. It just happens. This could be changed. Designs goals for the races and the mechanisms to obtain that goal. This would characterize their behaviors in game and then the player has to decide how they are going to deal with the race as it reaches toward that goal.
        • For example, the Thoraxian Queen (in a particular mood) might be dedicated to the idea of a conquest victory of the solar system. Currently she can be warlike, but she doesn't do anything except make her attack more often. Deeper goals she would evaluate her goal, determine her strengths and weaknesses, and then compare it against the strengths and weaknesses of the other races. She then might end up plotting a way to get the Skylaxians to suicide their great armada into the Acutians thus enabling her lesser fleet to take over both of their empires. Or she might decide to bide her time and continue to build up armadas. Or she might think to ally herself to the Skylaxians and to get them to share technologies increasing her own strengths, get the Skylaxians into wars and steal the planets taken from the enemies and gradually build up her own strength at the cost of the Skylaxians. Then once she is big enough, she betrays them and takes all their stuff.
    • This is obviously not very detailed or particularly useful.
    • Some brief organizational thoughts
      • Planetary and Racial AIs will have to be separated at some point.
        • Grand Strategy will be the Role of the Racial AI
          • Combat Strength: Current and Future predictions, relative to the strengths of the other Racial AIs
          • Research Strength: Current and Future predictions, relative to the strengths of the other Racial AIs
          • Manufacturing Strength: Current and Future predictions, relative to the strengths of the other Racial AIs
          • Racial Relationships: Who we should try to be friends/enemies with, who we should make our enemies. And who there friends/enemies are.
          • Alliance Relationships: What sort of assistance can we get if we make alliances with others, also pro & con federation.
          • Tactical War AI: Where our ships should go and what they should do there. When do we start building drop ships versus fleets.
          • Strategic War AI: How our allies and our combat strength fares against our enemies combat strength.
        • Planetary Strategy will be the Role of the Planetary AI
          • How the planet is faring
          • What can be done to make the planet better
          • Trades and other ways to make the planet better
          • Telling the Racial AI what we have and how that it fits into what it wants.
          • Budget Allocations
          • Fleet, RCI, building, population representations
      • Give the Planetary AI the ability to "understand" RCI.
        • We need more production = Increase Economy
        • Our hostile planet is killing us = Increase Environment/Planet Habitability
        • We need more fleets = Increase Public Order
        • Our people cannot maintain a stable population - Increase medical
      • Give the Planetary AI the ability to respond to changes in RCI need by building buildings, asking the Racial AI for tech priorities, asking the Hydra player for help, allocating budget to RCI improvement.
    • Once Racial and Planetary AIs are created, then you can add personality weights to shape the different Races to behave differently.
      • The Acutians will promote Industry and be less concerned with the environment.
      • The Burlusts will be eager to go after the weakest races even if their is a chance of intervention.
      • etc
    • I'm not sure how useful this is, but you asked.
  • re: RCI debates
    • I owe NichG one internet for the SimCity source.
    • Short Version - The players and the devs have a different version of what RCI means.
    • The devs have been slowly giving in.
    • The experienced players are currently upset that they have options that are useless Dispatch RCI Improvement.
    • The devs think the solution might be to remove Dispatch RCI Improvement, but are trying to make less drastic changes (re: Federation Points).
    • Temporary Quick & Ugly Solution
      • No guarantees
      • Balance RCI on min/max of -100/+100, 0 neutral
      • Add variables for min/max RCI and then set percentages for each label description and RCI trigger points. This way you can change them easily, instead of having to do it all by hand every time.
      • Plan that Technologically advanced races will have an RCI of 70 in 100 years.
      • Make all RCIs have trends all the time. Increase trend size to +/-0.25 per month, and have a 2% chance of switching per month.
      • Reduce maximum impact of Events to 20 for small events, 30 for medium events, 50 for major events
      • RCI buildings have a maximum of 1 (or divide all following values by the max number)
      • Increase building effects to +10 one time increase and +0.1 per month
      • Give Hydraal player the ability to reduce RCI by 1 of a race. Do not give them a reputation or relationship penalty for it.
      • Long Version - Go read the RCI change thread
« Last Edit: June 26, 2014, 01:51:39 am by ptarth »
Note: This post contains content that is meant to be whimsical. Any belittlement or trivialization of complex issues is only intended to lighten the mood and does not reflect upon the merit of those positions.

Offline NichG

  • Full Member
  • ***
  • Posts: 125
Re: About The Recent Schedule, and TLF Updates
« Reply #16 on: June 26, 2014, 01:44:49 am »
I'll chime in and say that the RCI Redesign thread is currently in a somewhat unsettled state as of yet, in the sense that there isn't a particular solid recommendation we've all been able to come to agreement on. It may hopefully get more clear in the next couple of days since ptarth and I have been building a sort of toy model simulator so we can pass it around and see what actually ends up surviving the theory-crafting stage.

So that may be a good thing to put off reading for a few days if time is limited.

Offline jonasan

  • Jr. Member Mark III
  • **
  • Posts: 97
Re: About The Recent Schedule, and TLF Updates
« Reply #17 on: June 26, 2014, 03:24:15 am »
as a side note to the ongoing discussion.... but in relation to the original post......

thanks very much for the ongoing linux support chris! getting (pretty much) the whole Arcen library on linux will be great!

i hope your able to get things working with SDL 2 so that we can have controller support for the games through steams big picture mode controller configuration..... with that i will be able to play some local co-op shattered haven natively!

Offline MaskityMask

  • Full Member Mark II
  • ***
  • Posts: 168
Re: About The Recent Schedule, and TLF Updates
« Reply #18 on: June 26, 2014, 03:35:33 am »
Thanks for the answers :)

Though, ptarth, I have to say that I kinda figured out that anti federation alliances ARE the end game threat so they would still be important part of the game's arc. I mean, if federation forms before fear empire and federation is composed of races that are willing to share tech with each other, they are going to be much stronger than any non federation race so only chance for non federation races to be a threat is to form alliance as well, but if they dissolve upon creating alliance right away, its kinda anticlimactic. So basically, anti federation alliances are only thing that is going to be challenge after federation has formed, unless you get unlucky and one of federation members get wiped out right after federation is created or you go for sake of challenge to create federation with unusual races. So yes, I'd think that anti federation alliances are bit more important than what you suggest they would be

Anyway, yeah, advanced start game options where you can start with some of anti federation alliances already existing would go long way of creating alternate challenges and even more replay value :) Bonus points if they allow for super rare events to be happening more often and such stuff.

Speaking of super rare events, I get that they are supposed to be rare, but... I'm not actually sure if they work in the game. I mean, since they are so rare that nobody has seen them, I'm not sure if they are actually bugged or not <_< I mean, I've only seen ultimate weapon buildings, acutians using planetcracker on their own and evuc gas giant igniting in versions close to release version. I never even saw end result of acutian planetcracking since they got wiped out.(they don't seem to do it even when they are down to lost planet and have moons left so I don't know what actually makes them to do planetcracking) The ultimate weapon buildings appeared only two playthroughs in different versions, so I don't know if they currently even work or if they just have so obscure conditions they can't form, even if year is 100 years after start, race has all technology and races have -1000 dislike of each other they can't build them, so I don't know if they are just THAT rare or if they are bugged nowadays.

I guess alternate start conditions/scenarios would be one way to solve that problem too. I mean, I get it that game is supposed to be partially random so that each playthrough is different, but I think having some special scenarios available for sake of challenge would be interesting. Like, not just having scenario where one of antifederation alliances already exist, also have scenarios where evucs start with gasgiant, have high possibility of igniting it and their planet is located so that explosion radius kills everyone if they succeed, scenario where acutians are trying to planetcrack someone and you have to stop them, scenario where federation races are both robotics ones and skylaxians have built cyberwarfare ultimate building, thoraxian first race scenario with them having hunter fleet building and having pirate empire already existing so that hunter fleet has something else to do besides hunting the player. So you'd be on timer where you have to win game before thoraxians kill all pirates and start hunting you instead. Etc etc, I can figure a lot of scenario ideas. (Would be also good way of having peltian fear empire scenario along with asteroid perturber :P Since I really want to see those furballs being danger to all life in solar system, I'm still working on creating peltian fear empire which is really hard since they keep getting wiped out. Bonus points for good laughter if the races peltians wiped out to create fear empire are acutians, burlust and thoraxians)

Note: I don't really have context of how uber powerful the uber rare event buildings are since I haven't figured out yet how to artificially make conditions were they are created xP I know that acutian one is really gamebreaking though since -1 economy to all planets per month is a lot. Though I guess it would also make for interesting scenario... Game where everyone but acutians have their economy ruined. Huh. Yeah, I think having scenarios where those buildings already exists from the start might make for good challenge

Anyway, got sidetracked there when I started to ponder about possible options for it, but yeah, I agree that advanced starting conditions would be an awesome option :D

But yeah, I don't know much about RCIs, but I can say that if rci reached low or high end of spectrum, it never goes down or up again so it feels frustrating to player since options make it feel like you are supposed to be able to affect it yet your attempts to put a dent to numbers is meaningless since even if you build all rci buildings, give race all rci boosting tech, tell race to boost rci value and dispatch yourself to raise rci values, you can't make -1000 rci back to zero again since it goes down to -2000 like a snowball down to hill no matter what you do
« Last Edit: June 26, 2014, 03:51:07 am by MaskityMask »

Offline nas1m

  • Master Member
  • *****
  • Posts: 1,268
Re: About The Recent Schedule, and TLF Updates
« Reply #19 on: June 26, 2014, 04:55:30 am »
@Chris: Thanks for elaborating. The vast majority makes sense to me ;).

A couple of quick notes:

RCI
- Not being able to dispatch to influence RCI might well reduce frustration. I see a couple of issues, though:
-- Currently RCI for any given planet seems to either snowball to any of the extremes or stay roughly at equilibrium fast without changing much after that during the course if the game - leading to games feeling samey once RCI has chosen its path
-- Dealing with the Boarines might become frustrating if nothing can be done by the player to influence RCI
-- All of this might be mitigated by a new quest that allows to adjust a planet's RCI *significantly* e.g. by the hundreds
-- In general more quests would go a long way improving the game imho
-- Being able to hurt RCI referred to adding a new dispatch, not being disgruntled about a preexisting one that has been removed ;)
-- ptarth's quick and dirty approach to improve RCI doesn't look to shabby as well ;) - if keeping RCI as something the player can manipulate is okay with you

Planetcrackers, Gas Giant Ignition, Ultimate Buildings
- I definitely feel that the probability of major events (planet cracker, gas giant ignitions, ultimate buildings) should be increased or be configurable - I have not yet seen *any* of these in my games :(!
-- An Advanced start Option to Control this would be totally sufficiant

Variety and Warlike races
- Thoraxians and burlusts still have a huge amount of trouble to fullfill their designed role as the menace of the solar system - they get tanked by the skylaxians/arcutians in each and every of my games...

Community Interaction
- In general the polls set up by ptarth would be a great place for you to start regarding Community questions

- Finally, thanks for looking into my gripes ;D
« Last Edit: June 26, 2014, 08:36:12 am by nas1m »
Craving some more color and variety in your next Bionic run? Grab a boost and a couple of custom floors!

Offline x4000

  • Chris McElligott Park, Arcen Founder and Lead Dev
  • Arcen Staff
  • Zenith Council Member Mark III
  • *****
  • Posts: 31,651
Re: About The Recent Schedule, and TLF Updates
« Reply #20 on: June 26, 2014, 08:56:13 am »
Ye gods, that's alot of text.  Ugh.  Normally I try to read absolutely everything in a thread, but

Now you know how I feel! ;)  Minus the ugh part, but there is just overwhelming amounts of stuff.


My own biggest problem is very definitely the RCI.  So very many things are based on it.  But right now, there's no such thing as strategy in dealing with it.... it's up to the whims of the Random Number Gods.  If an RCI value starts to plummet?  There's little you can do.  The RCI values overall seem to be based on a really large "scale", so to speak, with values that are bad or good being in the thousands. 

But the things you can DO about it, and the effects that things like buildings have on it, in other words, the possible rate of any given change, seem to be based on a scale where 100/200, or negative 100/200, are the "really bad" or "really good".  In other words, the numbers are WAY too skewed in relation to each other, and seem to actually be from different sections of the development process over the beta.  It's as if an RCI balance change was needed at some point, and the current scale of thousands was created.... but the numbers in actual EFFECTS were utterly forgotten and simply not changed from the previous model.

Really what we did at one point was uncap it, as originally it was based on a -100 to 100 scale, as you guessed.  So was influence and attitude, eventually.  The severe effects from RCI were not introduced until later, though, so it's most likely actually just an artifact of us going a bit overboard with making RCI meaningful after people complained they were meaningless.

As for the buildings, earlier in the beta they were a flat addition to RCI amounts, which turned out to be a really bad model.  So we shifted them to the current model, but made them conservatively slow.

I'll chime in and say that the RCI Redesign thread is currently in a somewhat unsettled state as of yet, in the sense that there isn't a particular solid recommendation we've all been able to come to agreement on. It may hopefully get more clear in the next couple of days since ptarth and I have been building a sort of toy model simulator so we can pass it around and see what actually ends up surviving the theory-crafting stage.

So that may be a good thing to put off reading for a few days if time is limited.

Gotcha, makes sense.  I think I may a couple of stabs at it based on some comments in this thread, though -- mainly splitting out the controversial bits into some options in Advanced Start, so that people can experiment with it in the game.  If something is so widely loved that we think it needs to be there for everyone, then we can take away the option and merge it in.  Of course that still may not suffice in terms of my first guesses here, but since this is such a hot topic I figure that some low hanging fruit are something that I should aim for.

As an aside, I'm planning another TLF release today, so I'm going to try to get to a lot of the things you guys want. :)

as a side note to the ongoing discussion.... but in relation to the original post......

thanks very much for the ongoing linux support chris! getting (pretty much) the whole Arcen library on linux will be great!

It's my pleasure!  I'm not presently a linux user, but I have been at various points in the past dating back to RedHat 5.  And I used to run Mandrake DNS and file servers.  I use so much stuff that is windows-dependent that I can't reasonably make a switch, but like so many people I really wish there was an alternative, and I see linux as being that eventually.

i hope your able to get things working with SDL 2 so that we can have controller support for the games through steams big picture mode controller configuration..... with that i will be able to play some local co-op shattered haven natively!

I'm not sure on that bit, as I haven't checked out that part of the code in the steamworks framework at all, but at some point it would be nice for sure. :)  Probably that's something we might approach when developing Airship Eternal, since it would make good use of it.  And then we would backport that to our other titles that use controllers in a meaningful way.  First I have to get my hands on one of the steam controllers, though, heh.

Thanks for the answers :)

You bet. :)

Speaking of super rare events, I get that they are supposed to be rare, but... I'm not actually sure if they work in the game.

It's true that those are definitely a bit too rare, heh.  Some of those secret weapon buildings need to be something that come about in a different fashion.  I have it currently set in a certain way that they unlock and can be built, but the main building queue conflicts with them a lot.  Not in the sense that they are bugged, but in the sense that it is just... really unlikely for them to build them, even as a last-ditch effort, and they should be less rare than THAT, heh.

RCI
- Not being able to dispatch to influence RCI might well reduce frustration. I see a couple of issues, though:
-- Currently RCI for any given planet seems to either snowball to any of the extremes or stay roughly at equilibrium fast without changing much after that during the course if the game - leading to games feeling samey once RCI has chosen its path
-- Dealing with the Boarines might become frustrating if nothing can be done by the player to influence RCI
-- All of this might be mitigated by a new quest that allows to adjust a planet's RCI *significantly* e.g. by the hundreds
-- In general more quests would go a long way improving the game imho
-- Being able to hurt RCI referred to adding a new dispatch, not being disgruntled about a preexisting one that has been removed ;)
-- ptarth's quick and dirty approach to improve RCI doesn't look to shabby as well ;) - if keeping RCI as something the player can manipulate is okay with you

Adding new quests to help and hinder RCI by the hundreds might be some good opportunities to provide some good variety and whatnot, and new power to the player without it being an "always there" option.  I like that.

Having the Boarines manage their own RCI better might be something of interest, too, we'll see.

Variety and Warlike races
- Thoraxians and burlusts still have a huge amount of trouble to fullfill their designed role as the menace of the solar system - they get tanked by the skylaxians/arcutians in each and every of my games...

Huh, that's interesting, that's not what I was seeing.  But I will have a look.

Community Interaction
- In general the polls set up by ptarth would be a great place for you to start regarding Community questions

- Finally, thanks for looking into my gripes ;D

Good point!  And sure thing. :)
Have ideas or bug reports for one of our games?  Mantis for Suggestions and Bug Reports. Thanks for helping to make our games better!

Offline x4000

  • Chris McElligott Park, Arcen Founder and Lead Dev
  • Arcen Staff
  • Zenith Council Member Mark III
  • *****
  • Posts: 31,651
Re: About The Recent Schedule, and TLF Updates
« Reply #21 on: June 26, 2014, 09:05:33 am »
  • re:Racial AI
    • And thus we see what happens when I try to be funny. I was suggesting this would be Arcen's next major announcement.
Ah, I see. :)

  • Further develop a "mental state" for each race. Allowing them to measure and evaluate the past, present, and future of the system (other races strengths and weaknesses). Then give them the ability to use that information to their own ends.
  • Currently the races are not active participants in the game. They don't have a goal they are seeking towards. They have sometimes end the game by military dominance of the system, but they aren't making choices designed to get them there. It just happens. This could be changed. Designs goals for the races and the mechanisms to obtain that goal. This would characterize their behaviors in game and then the player has to decide how they are going to deal with the race as it reaches toward that goal.
  • For example, the Thoraxian Queen (in a particular mood) might be dedicated to the idea of a conquest victory of the solar system. Currently she can be warlike, but she doesn't do anything except make her attack more often. Deeper goals she would evaluate her goal, determine her strengths and weaknesses, and then compare it against the strengths and weaknesses of the other races. She then might end up plotting a way to get the Skylaxians to suicide their great armada into the Acutians thus enabling her lesser fleet to take over both of their empires. Or she might decide to bide her time and continue to build up armadas. Or she might think to ally herself to the Skylaxians and to get them to share technologies increasing her own strengths, get the Skylaxians into wars and steal the planets taken from the enemies and gradually build up her own strength at the cost of the Skylaxians. Then once she is big enough, she betrays them and takes all their stuff.

The tricky thing with traditional AI like that is that there are just so many edge cases, and there are so many ways for the player to figure out what the current mental state is, and then exploit it.  I don't like coding overarching central AI for those reasons, and I instead prefer to code distributed AI that is more reactive in the moment and thus dynamic, etc.  I've written (and lectured to other AI developers) at length about it in the past, actually: http://arcengames.com/designing-emergent-ai-part-1-an-introduction/

Myself and the traditional AI approaches just don't meld well, because I just don't believe in them.  Keith, on the other hand, has had a lot of success in AI War building traditional AI mechanisms on TOP of my distributed AI approaches, and that hybridization has been super powerful.  But the core thing is that it's something that doesn't try to control the entire AI, it's just a more complex subroutine in the distributed decision making, so to speak.  That's kind of what I would envision here.

  • This is obviously not very detailed or particularly useful.
  • Some brief organizational thoughts
    • Planetary and Racial AIs will have to be separated at some point.
      • Grand Strategy will be the Role of the Racial AI
        • Combat Strength: Current and Future predictions, relative to the strengths of the other Racial AIs
        • Research Strength: Current and Future predictions, relative to the strengths of the other Racial AIs
        • Manufacturing Strength: Current and Future predictions, relative to the strengths of the other Racial AIs
        • Racial Relationships: Who we should try to be friends/enemies with, who we should make our enemies. And who there friends/enemies are.
        • Alliance Relationships: What sort of assistance can we get if we make alliances with others, also pro & con federation.
        • Tactical War AI: Where our ships should go and what they should do there. When do we start building drop ships versus fleets.
        • Strategic War AI: How our allies and our combat strength fares against our enemies combat strength.
      • Planetary Strategy will be the Role of the Planetary AI
        • How the planet is faring
        • What can be done to make the planet better
        • Trades and other ways to make the planet better
        • Telling the Racial AI what we have and how that it fits into what it wants.
        • Budget Allocations
        • Fleet, RCI, building, population representations
    • Give the Planetary AI the ability to "understand" RCI.
      • We need more production = Increase Economy
      • Our hostile planet is killing us = Increase Environment/Planet Habitability
      • We need more fleets = Increase Public Order
      • Our people cannot maintain a stable population - Increase medical
    • Give the Planetary AI the ability to respond to changes in RCI need by building buildings, asking the Racial AI for tech priorities, asking the Hydra player for help, allocating budget to RCI improvement.
  • Once Racial and Planetary AIs are created, then you can add personality weights to shape the different Races to behave differently.
    • The Acutians will promote Industry and be less concerned with the environment.
    • The Burlusts will be eager to go after the weakest races even if their is a chance of intervention.
    • etc
  • I'm not sure how useful this is, but you asked.

Sure, I appreciate it!  Generally speaking that sort of AI control is kind of the antithesis of how I do things, but in your original comment and to some extent above, there's a lot of interesting kernels that can be adapted to the sort of systems that I do use.  For instance, races don't need memory -- in fact, having memory is demonstrably bad, because it makes them predictable.  But how races weight various options and react to specific circumstances is something that would make a lot of sense to make more unique along the lines of what you are saying.  There is already a fair bit of that that is completely random (in terms of a race choosing a favorite technology group to pursue at the start of each game, for instance), but there's not enough, and certainly not enough that is specific to the flavor of a specific race.
[/list][/list]
Have ideas or bug reports for one of our games?  Mantis for Suggestions and Bug Reports. Thanks for helping to make our games better!

Offline Riabi

  • Former Arcen Staff (tigersfan)
  • Master Member Mark II
  • *****
  • Posts: 1,595
Re: About The Recent Schedule, and TLF Updates
« Reply #22 on: June 26, 2014, 10:00:22 am »
Ye gods, that's alot of text.  Ugh.  Normally I try to read absolutely everything in a thread, but

Now you know how I feel! ;)  Minus the ugh part, but there is just overwhelming amounts of stuff.

Yep, I used to feel this way too. Now that I don't work here, I get to pick and choose what I want to read. :)

Offline TheDarkMaster

  • Newbie Mark III
  • *
  • Posts: 28
Re: About The Recent Schedule, and TLF Updates
« Reply #23 on: June 26, 2014, 02:31:24 pm »
With the major events, I started a whole thread discussing what I thought would be a good idea for them: http://www.arcengames.com/forums/index.php?topic=15846.0

In short, the idea is that these are the things that should flavor each game and make it distinct.  They're things that make the player have to respond to them and change their playstyle so that once you figure out one strategy that works you can't just do only that strategy and always win.  Having how frequently these events can happen be tied to the difficulty level or a different option in the menu isn't a bad idea, but they really are needed and you need to make sure that at least one or two trigger near the start of the game to give you a good reason to play the game after you've beaten it once or twice without intentionally sabotaging yourself.

Offline ptarth

  • Arcen Volunteer
  • Hero Member Mark III
  • *****
  • Posts: 1,166
  • I'm probably joking.
Re: About The Recent Schedule, and TLF Updates
« Reply #24 on: June 26, 2014, 03:57:14 pm »
  • Quote
    As for the buildings, earlier in the beta they were a flat addition to RCI amounts, which turned out to be a really bad model.  So we shifted them to the current model, but made them conservatively slow.
    • In my search for the history of RCI I don't recall seeing why they turned out to be a bad model. Can you clarify?
  • Rare Event Rarity
    • I believe the major problem is that games still don't last long enough or turn out to have such dire situations to force the interesting Rare Events
    • Instead we get: Burlusts are again able to defend their rock with 1TB of troops. You have to spend 20 years of intense population destruction to make any progress
    • Part of that problem comes from the primarily multiplicative combat strength mechanism being combined with the high birthrates of the Burlust(and their super building).
    • By being able to either finish games more quickly or make the end game more interesting than Burlust eradication, more players will see more rare end game events.
  • That ptarth guy
    • He is a genius.
    • He makes such nice comprehensive polls.
    • He is handsome.
    • And he loves to use bulleted tables in his posts
    • Did I mention his great sense of humor?
  • Traditional Decision Tree AI versus Individual Agent AI versus hybrid AI.
    • I feel I'm being mischaracterized here.(I don't think that means we have to duel to the death, but I might be mistaken. Please consult my second if this is the case for further arrangements)
    • The initial description was system-neutral not favoring nor endorsing any system of AI organization.
    • The second more in depth characterization was definitely hybrid in nature, the same system you speak about in your AI war thesis.
    • I'm blaming it on my lack of clarity, I need to use more bullet points.
    • Racial Memory -I suppose it depends on what you mean by memory. This is getting far afield, so I won't pursue it.
    • Randomness - This one seems to be a hot issue given your posts and articles on AI. There is a range of AI behaviors, for which you want some randomness. However, making true random behavior is not desirable (and the point you bring up in your own article). However, on a simple reading you come out in saying that you don't want predictability.
    • Quote
      having memory is demonstrably bad, because it makes them predictable
    • I believe you are caricaturing your case here. You want AIs that provide a range of behaviors when there are many reasonable options, however when there are not, you do want a predictable AI. That's inherent in the probabilistic agent-based AI approach. However, as it read, you sound like one of those GOG extremists (per other thread where GOG extremist was inadvertently misused leading to massive walls of text over the misunderstanding).
    • To continue to beat a dead horse, I refer to your dumpsterKEEPER retreat example in your AI articles.
      • I do not accept your response that the AI behavior is good (splitting and sending some ships to die).
      • I tentatively accept that it may be good enough(for an AI with the level of resources AI Wars AI has) and for the time commitment to further AI development.
      • I will suggest that the the better answer is that the AI needs to have agent accessible information about all systems.
      • To limit it to the specific issue, it needs to have "retreat plans" for every single system.
      • It should know how dangerous it is to cross through any system with ships.
      • Traditional AIs either have omniscience or have to scout. I believe scouting is more in line with AI Wars.
      • Given an up to date system danger plan (a function of both mobile defense availability and fixed defenses in the way of a retreat), shouldn't it be straightforward to calculate a retreat plan that is very reasonable?
      • It chooses the path based upon a probabilistic evaluation of reasonable danger by length of route ratio. Possibly adding some more logic based upon other units fleeing with it (more of us = safer travelling).
      • This would also mean the AI would allow itself to be encircled, a characteristic of human war strategy. Obviously then, the "danger" of being in an encircled system would also be something that could be used at the Strategy level AI to make decisions.
      • It would also mean that if you secretly build a massive defense system in a "unprotected" wormhole junction, which the AI does not scout, and then you force an AI force to retreat from an adjacent system, it may choose the "unprotected" wormhole as the best retreat path. It then would have the ships moving there be annihilated (i.e., you just created an ambush). Once the first ship jumps in, the AI would then have to choose to attempt to push through the ambush or to select a different retreat path (risking interception by the fleet that caused the retreat in the first place). These are good AI behaviors.
  • Don't feel you have to respond to this in detail. I spent much too long writing it and both our time is perhaps better spent on being productive.
Note: This post contains content that is meant to be whimsical. Any belittlement or trivialization of complex issues is only intended to lighten the mood and does not reflect upon the merit of those positions.

Offline x4000

  • Chris McElligott Park, Arcen Founder and Lead Dev
  • Arcen Staff
  • Zenith Council Member Mark III
  • *****
  • Posts: 31,651
Re: About The Recent Schedule, and TLF Updates
« Reply #25 on: June 26, 2014, 04:40:59 pm »
Ye gods, that's alot of text.  Ugh.  Normally I try to read absolutely everything in a thread, but

Now you know how I feel! ;)  Minus the ugh part, but there is just overwhelming amounts of stuff.

Yep, I used to feel this way too. Now that I don't work here, I get to pick and choose what I want to read. :)

Happy belated birthday, by the way! :)

With the major events, I started a whole thread discussing what I thought would be a good idea for them: http://www.arcengames.com/forums/index.php?topic=15846.0

In short, the idea is that these are the things that should flavor each game and make it distinct.  They're things that make the player have to respond to them and change their playstyle so that once you figure out one strategy that works you can't just do only that strategy and always win.  Having how frequently these events can happen be tied to the difficulty level or a different option in the menu isn't a bad idea, but they really are needed and you need to make sure that at least one or two trigger near the start of the game to give you a good reason to play the game after you've beaten it once or twice without intentionally sabotaging yourself.

Nice!

  • Quote
    As for the buildings, earlier in the beta they were a flat addition to RCI amounts, which turned out to be a really bad model.  So we shifted them to the current model, but made them conservatively slow.
    • In my search for the history of RCI I don't recall seeing why they turned out to be a bad model. Can you clarify?
Hmm... I'm not sure how much was documented, and when it was.  May have been during alpha.  For a long while it was a "+20 to RCI" value on buildings, and it would just be a flat boost.  I shifted away from that because the very concept of that is just a little odd.

  • Rare Event Rarity
    • I believe the major problem is that games still don't last long enough or turn out to have such dire situations to force the interesting Rare Events
    • Instead we get: Burlusts are again able to defend their rock with 1TB of troops. You have to spend 20 years of intense population destruction to make any progress
    • Part of that problem comes from the primarily multiplicative combat strength mechanism being combined with the high birthrates of the Burlust(and their super building).
    • By being able to either finish games more quickly or make the end game more interesting than Burlust eradication, more players will see more rare end game events.
Right, this is true.  A lot of the rare events are actually time-gated, too, so that they don't stomp new players.

  • Traditional Decision Tree AI versus Individual Agent AI versus hybrid AI.
    • I feel I'm being mischaracterized here.(I don't think that means we have to duel to the death, but I might be mistaken. Please consult my second if this is the case for further arrangements)
    • The initial description was system-neutral not favoring nor endorsing any system of AI organization.
    • The second more in depth characterization was definitely hybrid in nature, the same system you speak about in your AI war thesis.
    • I'm blaming it on my lack of clarity, I need to use more bullet points.
    • Racial Memory -I suppose it depends on what you mean by memory. This is getting far afield, so I won't pursue it.
    • Randomness - This one seems to be a hot issue given your posts and articles on AI. There is a range of AI behaviors, for which you want some randomness. However, making true random behavior is not desirable (and the point you bring up in your own article). However, on a simple reading you come out in saying that you don't want predictability.
    • Quote
      having memory is demonstrably bad, because it makes them predictable
    • I believe you are caricaturing your case here. You want AIs that provide a range of behaviors when there are many reasonable options, however when there are not, you do want a predictable AI. That's inherent in the probabilistic agent-based AI approach. However, as it read, you sound like one of those GOG extremists (per other thread where GOG extremist was inadvertently misused leading to massive walls of text over the misunderstanding).
Apparently we are both being mischaracterized. ;)  In terms of memory, I really am not a fan of it for AI in the main because it makes them slow to react.  I prefer AI that looks at the current state moment by moment and makes decisions based on that.  Obviously some memory is needed in terms of "what was I just doing?" so that the AI doesn't just flail about.  But that's more about following through on one action than something else.

The way that I understood your note was to try to use data from past actions of the player and the other AIs to predict their future actions.  That... gets very dangerous.  It's really easily abusable because players can do one thing for a long time, then immediately switch what they are doing and the AI will react slowly because the vast bulk of its data tells it to.  Either that, or it is having some special overrides to make it react more specifically to specific cases.  Either way, you wind up with mounds more code, mounds more bugs, and typically more exploits.  If the AI is making choices that are in the range of optimal rather than always optimal, and then remembering and following through on those once chosen (to a certain extent), then you get the ideal behavior, in my opinion.

The example from dumpsterKEEPER, if I recall it correctly, was the AI sending ships back and forth through wormholes indecisively.  That was a great example of the AI itself not remembering what IT was trying to do and thus causing problems by not committing to a decision.  Versus trying to anticipate traps based on your past decisions, which instead tends to make it actually easier for you to set counter-traps if you are a good player.  Even with the whole "commit to a decision" thing, there has to be some threshold after which the AI abandons a decision if it no longer thinks it is a good one.  Otherwise, again, you get an AI that is sluggish to react.

True randomness has never been my goal, and nor have I meant to sound like it is.  But programming decision trees is perhaps the antithesis of my general approach.

  • I do not accept your response that the AI behavior is good (splitting and sending some ships to die).
  • I tentatively accept that it may be good enough(for an AI with the level of resources AI Wars AI has) and for the time commitment to further AI development.
  • I will suggest that the the better answer is that the AI needs to have agent accessible information about all systems.
  • To limit it to the specific issue, it needs to have "retreat plans" for every single system.
  • It should know how dangerous it is to cross through any system with ships.

Interesting that you don't find the AI good in AI War.  But, ah well.  I seem to recall you are one of the uber players, so I'm not surprised you would find that opinion.  AI War is designed to be good up to a point for most players, and then be situationally difficult for people like yourself.  Basically getting into the "good enough given the circumstances" area there.

None of what you're saying there about agent-accessible information like escape routes and danger estimates and so forth is at all at odds with what I am saying.  The AI already does that sort of thing, and that's not a decision tree.  Rather, that's weighting of information.  The thing to do then is take something like the top 10% of good options, and choose one of those at random and then stick with it.  If the quality of options has a massive drop-off past some point, then ignore all those options, which may of course leave you with one.

That's actually more or less how the AI in AI War operates in a number of circumstances, but it hasn't been built up more because it hasn't really seemed needed by the playerbase.

Anyway, I guess the argument that we were pseudo-having (where we both thinking the other was saying the other was not, so actually we were more or less arguing on the same side against imaginary copies of ourselves):

False you: "As an example of the racial AI, the AI should calculate retreat paths, and then choose one, and then stick to it.  This should be something that is done very rigidly, and precalculated too soon, so that there is a high chance of the path becoming outdated and nonoptimal by the time it is needed to be used.  Also, always find the exact ideal path that is obvious, and take that."

False me: "No, it's stupid to make plans.  The AI should just run in random directions and see what happens.  Any sort of planning or target evaluation is holding me back, man!"

Actual both of us: "Given time and resources (which is always an iffy given), doing more complex and accurate evaluations of a larger variety of things is a good thing."

Actual me: "And that should be something that is done on the fly, and then kept as the instructions for a certain amount of time, unless the situation turns X amount bad, in which case things should override the normal 'stick with this' logic."[/list][/list][/list]
Have ideas or bug reports for one of our games?  Mantis for Suggestions and Bug Reports. Thanks for helping to make our games better!

Offline ptarth

  • Arcen Volunteer
  • Hero Member Mark III
  • *****
  • Posts: 1,166
  • I'm probably joking.
Re: About The Recent Schedule, and TLF Updates
« Reply #26 on: June 26, 2014, 05:16:03 pm »
  • Quick response because I was not clear enough (and created another misinterpretation).
    • When I said "for an AI with the level of resources AI Wars AI has", I meant: given its economic situation of generating ships. It doesn't generate ships like the player. For the AI, losing a fleet here or there doesn't really matter. The loss of a single fleet is equivalent to a rounding error.
    • I did not mean that the AI in AI Wars is poor.
  • AI Memory - I didn't mean to transmit the idea of Racial AI Memory being fundamental. I see it as a possible improvement once major low level AI work is accomplished, but nothing in the short or medium term.
  • I agree with you in AI development principle. I am also firmly in the Agent-based probabilistic AI (being controlled by upper level strategic AI) being the most productive AI development. We may differ in how much we think the overarching Strategy AI should influence lower level units, but that is a minor disagreement.
  • AI Wars
    • My hidden dirty secret is that I've never gotten around to really getting into AI Wars. I bought it when it came out and keep current on expansions, but I still haven't really played it.
    • The interface is not intuitive enough to me and I haven't committed the time to exploring it.
    • I am characterized by hard-core analytical playstyles.
    • I play Dwarf Fortress with spreadsheets and computational models.
    • I broke tLF in my first game.
    • I use more bulleted points than anyone else.
    • That should give me sufficient street cred even without much AI War Experience.
  • dumpsterKEEPER was referring to how the AI when retreating, broke its fleets up into different retreat routes. Some of them used intelligent routes and returned to the AI territories safely. Otherwise choose suicidal routes being annihilated by player static defenses.
  • Randomization and other Positions
    • I also firmly believe we are arguing from the same position but are having semantic difficulties.
    • Quote
      Actual me: "And that should be something that is done on the fly, and then kept as the instructions for a certain amount of time, unless the situation turns X amount bad, in which case things should override the normal 'stick with this' logic.
    • Actual Ptarth: "I would further suggest that there should be two parts to this. One part is updated on the fly, the other should be updated only periodically (when additional scouting is performed) to save on computation."
    • Actual x4000: "Well yeah, isn't that obvious"
    • Actual Ptarth: Touche. But given fallible human mental prowess, I try to be clear and explicit. Hence my awesome bulletpoints.
« Last Edit: June 26, 2014, 05:49:17 pm by ptarth »
Note: This post contains content that is meant to be whimsical. Any belittlement or trivialization of complex issues is only intended to lighten the mood and does not reflect upon the merit of those positions.

Offline NichG

  • Full Member
  • ***
  • Posts: 125
Re: About The Recent Schedule, and TLF Updates
« Reply #27 on: June 26, 2014, 07:59:11 pm »
On the AI thing, I sort of like the idea of some particular predictability in TLF because the various racial AIs aren't actually the player's enemy, they're part of the system that the player is playing. E.g. since the player is embodied as this super-manipulator who can get planetary governments to dance to his tune, it makes some sense to expose behavioral information to the player beyond what you'd want to do in something like AI War because the player's character having special abilities with regards to that kind of manipulation is a premise of the game.

That said, to avoid the 'false me' interpretation, I would say that those particular vulnerabilities should be very specifically chosen rather than just being a happenstance because the AI was susceptible to manipulation. Furthermore, exploiting those vulnerabilities should be something that costs time and resources to the player, so its not just a 'okay, I know how the AI will be dumb so I just have to minorly tweak my activities'.

For example:

- The first time the Acutian Economy drops below -50, they will almost always start a campaign of all sorts of ecologically destructive actions to repair it - trashing their own planet and those of their allies.

That's a sort of decision-tree like case, but its only one layer of the full decision-making process (that involves choosing which actions they'll do, what planets they'll target, and also side things like how much emphasis they're going to give on their armada versus infrastructure or whether or not they'll start a war or whatever). It also has a trigger which isn't trivial for the player to hit (e.g. the player can't just play Rock 20 times and then play Scissors 40 times like in the AI with memory example).

If the AI is too impenetrable, then that kind of meta-strategy becomes hard to formulate (and since there are a lot of easier and more effective strategies out there, players will tend to gravitate towards those instead).

Offline ptarth

  • Arcen Volunteer
  • Hero Member Mark III
  • *****
  • Posts: 1,166
  • I'm probably joking.
Re: About The Recent Schedule, and TLF Updates
« Reply #28 on: June 26, 2014, 09:03:27 pm »
This sort of behavior can be naturally produced by agent-level AI (as per x4000's approach). It is also the type of thing I've been continuously promoting as Planetary AI and lower level mechanisms.

To illustrate your example:

  • Each Race have economic and environmental RCI value weight parameters.
  • Each Race has a demand for various RCI parameters.
  • For the Acutians they place a value on production and economy, and a low value on environment.
  • The "demand" for economic RCI is a function of their parameters and their current RCI.
  • As economic RCI gets lower, the Acutians demand for it increases, eventually causing them to act to gain it.
  • The way they get it depends on another set of calculations. They consider their options and the ramifications of those options.
    • Sacrifice Public Order - We'd Rather not.
    • Sacrifice Environment - We're okay with it.
    • Steal from the Burlusts who have a massive army - That doesn't sound like a good idea
    • Steal from the Peltians, who are defenselesss - Yes.... Yes...
  • None of these are Decision Trees. Instead they are independently calculated and then probablistically selected from.
  • Most of the time the Peltians are going to get the short end of the stick. Sometimes you'll see Environment RCI go up in smoke. Rarely Public Order. And once in a Blue Moon, The Burlusts get a "present"
Note: This post contains content that is meant to be whimsical. Any belittlement or trivialization of complex issues is only intended to lighten the mood and does not reflect upon the merit of those positions.

Offline NichG

  • Full Member
  • ***
  • Posts: 125
Re: About The Recent Schedule, and TLF Updates
« Reply #29 on: June 26, 2014, 11:21:34 pm »
I actually meant to literally have a message that says to the player 'if you get Economy to -50, they will do X' and then follow through with that. Like the way the Federation joining conditions are presented now.

 

SMF spam blocked by CleanTalk