Thinking about Pirates

[bvanevery]

Details he has an air force

he has an air force

MY 2216.  Or not.  Maybe someone gave him the plane.  If he built it himself, then it implies he's also got missile launchers.  And just hasn't built troops that use them yet.  Well, we'll see if I survive.  But if I just get whipped, by an air force I can't do anything about, I'll quit.  Watching AI factions get every advantage, and basically being unable to do anything about it except beg / sugar daddy, is not much of a game.

And it's not like this guy has ever been interested in making friends with me.  No matter the game, even though I've typically been Democratic.

Details Morgan won't be able to protest

Morgan won't be able to protest

Some of my sea bases have gotten large enough that I'm having trouble keeping people happy.  I wouldn't mind having my mindworms pack a bit more of a punch.

[bvanevery]

Details the weak point

the weak point

MY 2217.  I send a Scout to see if the river is clear for infiltration.  A Former is in the way, so I have to send up a Recon Rover to kill it.  This will probably get it killed next year, but speed is of the essence.  And then I realize, I must be tired.  The Former wasn't actually in the way.  Oh well!  Anyways I'm infiltrating now.

Yeah he's got Doctrine: Air Power.  It's not some gift.  He doesn't have any allies.  He's got 4 planes in the region that I can see in bases.

[bvanevery]

Details you gotta be kidding me

you gotta be kidding me

MY 2220.  I just did that mission.   You really expect me to fight these behemoths 'blind', most of the game?  Lal is Democratic Planned Knowledge for pete's sake, with -2 PROBE. This is a total ripoff.  At least in WTP there's this protocol for implanting listening devices, and they last a certain period of time depending on how diligent the faction is about PROBE.  Which sure as heck shouldn't be Lal.

Yeah I've had enough of this game.  That was the last straw.  The game is viable, but godawful tedious.  I've got better things to do with my free time than pull teeth.

Air force finally got around to killing one of my advancing probe teams.  I was hoping it might notice me on the flank, but no such luck.  All this adds up to, it's not basically possible to stand in a land battle line against a gargantuan Thinker AI enemy, when you don't have the techs for basic advancement safety.  I'm prototyping my Silksteel armor but it's a drag, and there's nothing stopping Lal from just making Missile Needlejets anyways.  The rate he can produce stuff is absurd.  2 turns in little cities.

In the future, if there is much of a future, I'll have to regard the perimeters of these AI empires as toxic waste zones.  I might have done ok in the water, far away from them, and just sneaking a probe in here and there.  Although, if I have to do that sort of thing too much, it's godawful tedious too.  I'm also still not seeing what an intervention plan is supposed to be.

Mainly, I'm being reminded of a lot of holes in the original rules, that caused me to mod for 3.5+ years.

[bvanevery]

Gotta turn off the things that weren't improvements.  Looking in thinker.ini, I notice:

Code: [Select]

; Set the amount of priority AI factions have for the preferred social model.
; Allowed values: 0 (no priority) - 1000. By default the priority is moderate.
social_ai_bias=10
Totally cranking that to 1000.  Enough of this Morganic Green and Deirdre Planned AI rubbish.

Here's the one that got me started looking:

Code: [Select]

; Probe team infiltration on other factions can expire randomly when this option is set.
; Chance for discovery is rolled each turn and depends on the comparative PROBE ratings of the
; attacking and defending faction. If discovered, a dialog box will be displayed for both.
counter_espionage=1
That's a zero!

This one actually looks like an improvement over previous versions of Thinker:

Code: [Select]

; In vanilla game mechanics, eco damage on AI bases is notably reduced from player levels.
; This patch increases AI eco damage to the same level than what the player bases have.
; In addition, constructing Tree Farms and similar facilities always increases the clean
; minerals limit even before the first fungal pop.
eco_damage_fix=1
Well this next one explains why exploration has not been quite as profitable in either WTP or Thinker.  I'm ok with it, but I definitely did notice, compared to my own mod.

Code: [Select]

; Significantly reduce the frequency of supply pods on random maps and
; place nutrient/mineral/energy bonus resources on the same tiles instead.
rare_supply_pods=1
I totally don't understand the purpose of excluding these 3 factions from a slate of random opponents.  It might explain why my opponents have been awfully consistent from run to run.

Code: [Select]

; Select whether to exclude some factions from random selection in the game setup.
; Factions can be selected individually but at least 7 factions must always remain enabled.
;  1 = Gaians
;  2 = Hive
;  3 = Univ
;  4 = Morgan
;  5 = Spartans
;  6 = Believe
;  7 = Peace
;  8 = Cyborg
;  9 = Pirates
; 10 = Drone
; 11 = Angels
; 12 = Fungboy
; 13 = Caretaker
; 14 = Usurper
; 15 = First custom faction slot if available etc.
skip_faction=12
skip_faction=13
skip_faction=14
So, erasing those entries.

And last but certainly not least, in alphax.txt:

Code: [Select]

1, 2     ; Numerator/Denominator for frequency of global warming (1,2 would be "half" normal warming).
Gotta get back to the real game, and have the AI factions suffer the consequences for all their condensers and boreholes.  Oceans do seem to rise more than I remember in Thinker a long time ago, but it's not enough.  It shouldn't be easy to do piles of eco-damage willy nilly.  Setting to 1,1.  I wonder how this will combine with eco_damage_fix=1.

Ok, that looks like it could actually be an appropriate game for Pirates.  I will remember not to create any broad coastal front on land, i.e. I am not actually the Roman Empire.  It just doesn't work like that.  I wonder what kind of game system would be needed, to have it work like that.

I probably lucked into the right idea in the 1st game.  A flanked "monsoon jungle" peninsula.  The error was overextending the flanks.

[Induktio]

Details strong garrison

strong garrison

He's got a probe team on defense in this trivial forward city that I expected to be nearly empty.  And an Impact Speeder to kill my probe team if I back off.  I guess he was just in a war with Morgan, and of course they just made peace.  I keep underestimating Lal, despite that he's been a complete jerk in every game I've seen him in.  Does he especially have it in for the Pirates for some reason, or is he always this way to everyone?  Why isn't being Democratic making him happy?

What's really interesting here is that the AI doesn't have explicit priorities of keeping "one probe, one defensive unit, some offensive units" in one city, but these were emergent results with the way the front garrisoning algorithm works. Now it also seems to be quite robust in the way it handles multi front wars while not getting confused.

> I've definitely seen him be Erratic in the stock binary, but this Lal has never been in a good mood.
> Alternate possibility: diplomatic AI was always programmed to prey on perceived weakness.

The AI will definitely act more intransigent if it ranks higher than the player faction. Sometimes they will not even accept commlink attempts. There's only a couple of places where Thinker directly patches diplomacy dialogue, so the behaviour seen here should mostly be vanilla AI. That one example of AI attempting to blackmail and then asking for a pact is weird, but it should be only vanilla AI logic.

> Totally cranking that to 1000.  Enough of this Morganic Green and Deirdre Planned AI rubbish.

It is the way it is, because vanilla game mechanics place very high penalties on being in Free Market while airpower is available or not having the ability to pop boom. And restricting the AI choices this way would make either or both of those true for most factions, thus making the game much easier for humans. This is not necessarily true if the SE choices are modded, however.

> At least in WTP there's this protocol for implanting listening devices, and they last a certain period of time depending on how diligent the faction is about PROBE.

There's some point to say the infiltration should always be guaranteed to last a certain amount of time, but sometimes espionage missions fail outright. About WTP's version I'm not sure what purpose the introduction of specific infiltration devices actually serves.

It would possible to remake the mechanic such that upon infiltration, you're getting a guaranteed X turns of infiltration and then some random/probe rating specific factor modify the addition. Infiltrating again when you have, say 5 turns of 10 turns left, could renew the counter back to 10 turns. This value would have to be displayed somewhere in the GUI though. One option would be to add a new screen on ALT+T menu to display the infiltration status for all factions. This way it's also not required to introduce any listening device abstraction in the game.

[bvanevery]

Sometimes they will not even accept commlink attempts.

Lal did that to me a lot this game.

That one example of AI attempting to blackmail and then asking for a pact is weird, but it should be only vanilla AI logic.

2 attempts.  Lal and Zak both did it.

Totally cranking that to 1000.  Enough of this Morganic Green and Deirdre Planned AI rubbish.

It is the way it is, because vanilla game mechanics place very high penalties on being in Free Market while airpower is available or not having the ability to pop boom. And restricting the AI choices this way would make either or both of those true for most factions, thus making the game much easier for humans. This is not necessarily true if the SE choices are modded, however.

Of course, in my modding I changed the whole SE table.  My Capitalist (Free Market) doesn't have that problem, or that level of benefit.  It's quite onerous going back to old school Free Market.

Nevertheless, it's Morgan's thing.  He's the only one who has to do it.  I don't have any problem with Morgan being the runt weakling faction he always was.  You may not want to write a "works better in Free Market" AI for him, but someone might.  And if no one ever does, I won't miss it.

I was frankly quite surprised when Morgan started one of my games doing Free Market.  It looked decidedly suboptimal.  I can't remember which of these games that was.  He didn't like die or anything.  Your "protection radii" probably prevents summary execution by Recon Rover, same as my recommendation of Huge maps usually does.  And somehow the mindworms didn't get him.  Anyways if he can survive to midgame with Free Market, I don't see why he should be getting a gift of going Green.  He's not a human player.  He is not empowered with the agency of a human playing the game.  He is a capitalist pig ideologue.

Aside from being out of character, it ruins diplomatic mechanics.  Deirdre goes to war because you're not Green, when she's running Planned.

BTW you can pop boom in Free Market.  You have to spend enough money to make a Golden Age.  +2 for that, +2 for Children's Creche, +2 for Democratic.

There's some point to say the infiltration should always be guaranteed to last a certain amount of time, but sometimes espionage missions fail outright.

3 years isn't acceptable.  There's no play balance here.  I'm not in a multiplayer deathmatch where I want razor's edge difficulties in making basic maneuvers.  I'm not complaining about infiltration being "overpowered".  I'm dealing with AIs who cheat and look at my stuff anyways.  I'm not going to pay the mouseclicks to produce and then move these units over and over again, every few turns.

About WTP's version I'm not sure what purpose the introduction of specific infiltration devices actually serves.

Instead of shocking you with summary loss of capability, it's giving you a countdown warning of when you're going to lose the capability.  This way you don't have to spend the whole game stressing about whether your infiltration is going to suddenly vanish and leave you caught short.  You don't have to keep running infiltration missions "just in case".  It's clearly a quality of life improvement.

It would possible to remake the mechanic such that upon infiltration, you're getting a guaranteed X turns of infiltration and then some random/probe rating specific factor modify the addition. Infiltrating again when you have, say 5 turns of 10 turns left, could renew the counter back to 10 turns. This value would have to be displayed somewhere in the GUI though. One option would be to add a new screen on ALT+T menu to display the infiltration status for all factions. This way it's also not required to introduce any listening device abstraction in the game.

I'm sure you could do all kinds of things.  You could also just lift WTPs code, as it has already implemented a solution that works well.  You can see how many devices you have left, when you try to perform another probe mission.  That's adequate feedback on how long you have to go, "checking" on your status.  Rather than being able to ALT+T that info at any time.

If you're going to make infiltration "not forever", then you've already crossed the bridge of purity compared to the original game.  Might as well take a solution that works just fine.

The play mechanical impact of disappearing infiltration, is I tend not to infiltrate until it's actually needed.  Because pushing a unit all the way across a Huge map, is very tedious.  If the infiltration is then discovered and exposed 20 turns later, there's just no point.  I won't have even gotten any units over there.

It makes the Hunter-Seeker Algorithm a bit more powerful, because it's always going to shut off infiltration eventually.  That said, tech is currently so rapid in Thinker that the time window when HSA is dominant, may not be as problematic.  Then again, shortly after I remember achieving an Algorithmically Enhanced infiltration that 1st game, I got nuked.  So, whatever.

[bvanevery]

Details here we go again

here we go again

Alright, what is this, the 5th game now?  I'm determined this time not to make "landward" mistakes.  No broad land fronts.  No overly exposed sea fronts either, like the time I flanked prematurely.  Try to stick to the concept of the "perfect circle" in the water.  Try to avoid being the butt of overly aggressive AI diplomacy and troop movement.  This is also the 1st game I'm playing with "better" Thinker.ini settings.

Details hi ugly

hi ugly

One implication of that, is we get to see what Alien factions are capable of.  The good news about battling an Alien is you can gas them with impunity.  No need to wait for a boring permission from the Planetary Council.  I've done this plenty of times in The Will To Power mod.  However, doing so can create a power vacuum that other factions exploit, so I need to keep in mind a principle of geopolitical balance.  It will be 'interesting' to see what Thinker does with their directed research capability.  Do they just grab planes immediately and stomp everybody, including me?

I'm realizing I'm unlikely to ever go Planned, so I'm likely to be at war with the Aliens.  Since I'm forcing them to obey their Planned imperative, the political dynamics will be more like the original game.  That is to say, unless you're Yang, it's really hard to get along with these guys!

Oh and uh, I kinda forgot, that Thinker hasn't rebalanced their factions.  With their free armor and so forth, they could stomp all over the rest of us.  Glad I'm somewhat safer in the water, knock on (drift) wood.

Otherwise, I get to play with the obnoxious Peacekeepers again, who have never liked me in any game.   The Data Angels again, who have been unimpressive every game.  The Spartans again, whom I have trashed, with sugar daddy technology input.  The Believers, I think are new.  They are well known nuke lobbers, although at least they don't research well.  They're certainly on the short list of factions capable of becoming huge though.  Which Yang has proven, leads to nukes.

I get a merciful respite from Morgan's unwanted presence.  So tired of him.  That said, I was mildly interested to see what constraining him to Free Market would do.  Not enough to want to suffer him another time though.

Details the strategic radar

the strategic radar

Looks like a straightforward "settle right about here" kind of start, with no obviously better option.  Deep oceans aren't the place to build the cradle of your seafaring civilization.  Sitting on that mineral special looks like about as good a place for a capitol as any.

[Induktio]

> Your "protection radii" probably prevents summary execution by Recon Rover, same as my recommendation of Huge maps usually does.

What do you mean by this?

> That said, tech is currently so rapid in Thinker that the time window when HSA is dominant, may not be as problematic.

Tech cost is currently tuned so that it's nearly as fast as vanilla except in the late game phases. But if the tech stagnation option is enabled, that should really slow things down, because it's +100% cost instead of +50% like in vanilla.

Anyway, most design stuff involves tradeoffs. If you're restricting social engineering or any choices strictly on some narrative limitations, then it can't be prioritizing best AI because for the human players these alternative choices are always available when allowed by the game mechanics. So that's an issue where one has to strike some balance and it's done in this case with the social_ai config variables. It may not be the best for diplomacy dialog, but at least the social AI plays by the same rules than humans.

[bvanevery]

> Your "protection radii" probably prevents summary execution by Recon Rover, same as my recommendation of Huge maps usually does.

What do you mean by this?

Code: [Select]

; Balance faction starting locations more evenly on random maps and
; maintain a minimum distance to any nearest neighbors.
faction_placement=1
This is probably more beneficial to the AI factions than to the human player.  It does not allow me to immediately eliminate a nearby faction with a Recon Rover rush.

Like, murder Yang in his crib, for instance.  In that "peninsula flank" game, I started out close enough to Yang for him to become a problem for me, but not close enough to immediately mount an assault that could summarily eliminate him.  I had to fight him, after Thinker had already given him an enormous resource buildup.  It's only because I sugar daddied my techs, that I stomped him.  Tactically, in a fair fight, it would have looked a lot more like that Lal game that I just quit, as too tedious to manage.  My sea bases were pretty threatened for an interim period while I got caught up, having overextended those flanks to try to surround Yang.  Actually I did have some close calls with him settling sea bases around my own flanks, sneaking very spammy Synth Laser Foils around me, and nearly taking 1 of my cities.  Then I sugar daddied and that was that.

In my own mod, I have anti-rush protocols, but they're different.  Huge map is recommended, so distances are likely to be larger between factions, such that a rush is not logistically practical.  However I still have the stock game's faction placement algorithm, so you can end up right next to someone.  Which when you're used to playing on a Huge map, is very uncomfortable.

It does force players, and by inclusion AIs, to be less complacent.  Sure you usually have a lot of land to yourself, but you can't count on it.  You have to scout and make sure you have the means to survive, before you settle down to just growing growing growing.  Thinker on the other hand pretty much guarantees an AI is safe to grow grow grow, and doesn't have to do anything special to defend itself for some time.  Barring a sugar daddy, a Thinker AI faction will have grown plenty large enough to defend itself, by the time a human player or even another AI has shown up.

My other anti-rush protocol is that Synth armor is given almost immediately on Tier 1, without a need to prototype it.  Laser tech comes on Tier 2, and you do have to prototype it.  A Conquer AI can still learn that stuff fast enough to whip your ass if it starts close to you though.  Like Santiago or Yang, they can stomp you.  You might not have to fear a lot of factions that way, but the militant ones, you'd better not "press your luck" about how much defense you need.

Also in my mod, everyone starts with Clean Reactors available.  It's a bletcherous hack to solve the AI running out of SUPPORT problem, since I don't have access to the AI code.  It benefits the AI more than it benefits you the human player, which is how I justify it.  Once AIs trivially gain the Synth tech, they immediately start making Clean Synth Garrisons.  They stack their early bases with the things.  If you show up with a few Recon Rovers and think you're gonna trash one of my AI factions, you're going to be in for a rude surprise.  You can do it with a lot of Recon Rovers, if you've coughed up a lot of production in a hurry, but you're not gonna do it with a few.  That and and somewhat stronger defensive emplacements, means you're gonna bounce off the cities.  You could still pillage and create some trouble, but you're not gonna succeed at overrunning anybody.

Tech cost is currently tuned so that it's nearly as fast as vanilla except in the late game phases. But if the tech stagnation option is enabled, that should really slow things down, because it's +100% cost instead of +50% like in vanilla.

I'll try it if current Thinker settings adjustments prove unsatisfactory.  WTP settings were most unsatisfactory at some point in testing.  Even with some improvements in research pace, tech leakage is still a big problem in that mod.

Anyway, most design stuff involves tradeoffs. If you're restricting social engineering or any choices strictly on some narrative limitations, then it can't be prioritizing best AI because for the human players these alternative choices are always available when allowed by the game mechanics.

As you know from past interactions, I consider prioritizing AI performance at any cost, to be a huge design mistake in this game.  This is not Civ.  You are not a generic nationality, with a few different special military units for 'flavor'.  You are a faction with an ideology, in a diplomatic system where the clash of ideologies is core to gameplay.  Deirdre usually goes after Morgan and vice versa.  That's the gameplay.

Your approach is having your cake and eating it too, breaking the rules that are set out in the faction.txt files.  The game isn't "everyone chooses what they want".  The game is that human players choose what they want, AIs don't.  They are asymmetric that way.  The original design of the game is superior and should be respected in that way.  There's a reason the original 7 factions all had slightly shifted choices from one another.  It was to create a set of natural allies and natural enemies.  That all goes out the window when factions can do whatever they want, regardless of their design.

but at least the social AI plays by the same rules than humans.

This is not, and should never have been, a design goal.  Making Deirdre able to go Free Market or Planned is the worst thing you have done in this game.  We're not playing SMAC 20+ years later, because of Deirdre being a Free Market or Planned powerhouse.  We're playing the game now, because Deirdre is Deirdre, and Morgan is Morgan.  Maybe not every faction leader is as emblematic of the game as these 2, and it's well known that the expansion leaders, are not up to the same snuff as the original 7.  Or original 6 really, as Santiago is the tack-on diversity hire, who could never win a real war.  The game is:

• human rights vs. atrocities - Lal vs. Yang
• environmentalism vs. capitalism - Deirdre vs. Morgan
• religion vs. atheism - Miriam vs. Zhakarov

That's the core gameplay.  It's not a skin.  It's not the "win at all costs no matter what" game.

I'm glad to see at least nowadays, there are options to turn these most objectionable social changes off.  I will be verifying whether the game works 'right' when so doing.

The other big tiff we've had, is that all these condenser boreholes are supposed to have consequences.  You're not supposed to be able to cough these things out left, right, and center, and get off scott free.  If you researched all the techs necessary to make your Centauri Preserves and Temples of Planet, if you built all of that, if you actually had to stop expanding because you needed more Trance garrisons, if you actually had to restrain yourself making boreholes and factories because it was getting to be too much, well then the AIs and the humans would actually be playing the same game.  In the past, they weren't, at all.  I don't know how it is now.  I'm about to find out. 

I've got AARs of real global flooding scenarios I've lived through.  I've even beaten it, here and there.  With a ton of premediation.  Couldn't have done it, if I hadn't beelined for certain techs, and prepared with certain approaches.  This stuff is a pain, but it's there to provide a constraint upon what you're allowed to do in the game.  You have to earn all that production, not just have it right there for the easy taking.

After all these changes are made though, there's still the question of whether Infinite City Sprawl is, strictly speaking, the only valid strategy for the game rules as written.  My jury's out on that.  I will never, ever, be an ICS player.  It's deeply disgusting to me.   The game is supposed to be partly a builder game, and I want the builder approach to the game, to actually work.  I don't want the tedious "colonization over and over again" game, like the AI does.  It's boring as hell.

I might actually opine, that the people who really get into ICS, are usually avid multiplayers.  I don't have a bunch of survey evidence for that, but it's a hunch.  They could be getting their rocks off beating up other humans, at any cost, whatever mouseclicks are required to vanquish a human.  This is a very different gaming experience and motive than playing single player against AIs.  When you are beating other AIs, you either pat yourself on the back that your tedious exploit is somehow brilliant, or you look at the map and see it as an ugly waste of your time.  I'm in the latter camp.  There's no quality of life in settling piles and piles of little bitty cities.

Single vs. multiplayer design is a core tension in game development.  I think you are leaning towards the multiplayer end of the spectrum, but you can paint yourself as you see fit.  I don't see you as developing the single player builder narrative game mechanical experience, which is what made the game famous.  I see you as developing "opponents to beat, ways to beat opponents".  And that gets rather dry.

[bvanevery]

Details the 2nd colony

the 2nd colony

Settling my capitol is a trivial decision, but it begs the question where the 2nd base should go?  I mull it briefly, and decide that I must not get in the habit of depending on the land for anything.  It draws me into needing to defend the land.  So the sailing of the 2nd colonist, was a bit of a dead end and will require some backtracking.  Hopefully my Gun Foil will find somewhere better for it to be, shortly.  I think I will proceed southeast rather than west, attempting to have more centrality of impact on the map.  It won't be going far though.

[bvanevery]

Details what is the choice

what is the choice

I think getting Rec Commons has proven to be a pretty important early decision, so I will continue in this vein, as with previous games.  I'm doubting that mounting a war without bigger bases is a good idea.  Who knows if I even have anyone nearby to trash?  Logistically, I'm not supposed to expect a nearby enemy in Thinker.

Details chatty game

chatty game

I pick up Lal's comm frequency immediately.  Let's see if he's in the mood to give stuff away, so early.  Nope.  Doesn't want a Pact of convenience so I can see his starting location either.  I sign off, not bothering to find out if he already knows of other factions.

Details fungal wasteland

fungal wasteland

A sonar pod reveals that I'm probably next to the Sargasso Sea.  That could be good eventually, depending on what happens to pods when I pop them.  Right now it'll suck though.  Definitely an incentive to get some Transports into the middle of it though.  Meanwhile, should I plow through the fungus to get to a viable southeasterly settlement faster?  I'm inclined to say "yes" because I can retreat from any Isle that shows up, and an Isle hit won't kill me yet in any event.  At least, I don't think it can in open water.  Wouldn't stay put on fungus.

I sail slowly through the fungus without incident.

[Induktio]

Your approach is having your cake and eating it too, breaking the rules that are set out in the faction.txt files.  The game isn't "everyone chooses what they want".  The game is that human players choose what they want, AIs don't.  They are asymmetric that way.  The original design of the game is superior and should be respected in that way.  There's a reason the original 7 factions all had slightly shifted choices from one another.  It was to create a set of natural allies and natural enemies.  That all goes out the window when factions can do whatever they want, regardless of their design.

Well, I think you know what you wrote there isn't clearly true since the anti-ideologies (social models) specified in the faction.txt files are always parsed and even Thinker's social AI never chooses them and has never done so. That is clearly specified by the game rules. Have not seen anyone report bugs on that. The other part is more ambigious and related to how much narrative-driven you want a game to be. That's why there are many settings for the mod, because in this game there won't be some one-size-fits-all solution how the game should be played. Since this mod has been about AI improvement, there definitely should be an option to play against stronger AIs. Not that it's definitely required in any way, but if you constrain AI actions in significant ways, it's going to have its own consequences.

***

Related to my earlier comment "That one example of AI attempting to blackmail and then asking for a pact is weird, but it should be only vanilla AI logic." Now I can't seem to find the original message in which you noticed this AI diplomacy behaviour. Do you have a savegame of it just before you open the diplomacy or from the turn before it? That would be necessary for reproducing the issue.

[bvanevery]

Well, I think you know what you wrote there isn't clearly true since the anti-ideologies (social models) specified in the faction.txt files are always parsed and even Thinker's social AI never chooses them and has never done so.

I made a mistake about Deirdre going Free Market.  I don't think I've ever seen that, in any version of Thinker that I played in the past, although I wouldn't swear to it.  Nor do I think I've seen it in WTP.  I went back and edited my post, crossing out that part.  You may not have seen the edit.  She most certainly goes Planned though, and on that basis, my point stands.  Planned Deirdre lecturing you about going Green and then breaking Treaties over it, is downright obnoxious, and game mechanically broken.

What I have seen, is Domai going both Cybernetic and Thought Control, in the same game.  When of course he's supposed to be gunning for Eudaimonic.  It is a strange development sensibility, that it's ok to restrict what a faction is against, what you call an anti-ideology.  But not to similarly restrict what a faction is for, which is their ideology.  That mostly defines their anti-ideology, if such they have.  It gets a little fuzzy because of the 3-way SE system that sometimes provides false choices, like what the difference between Police State and Fundamentalist ever really would be.  That bugged me for a long time.

Your rationale is "players and AIs should be able to make equivalent choices" and I do not accept that rationale.  In addition to all the things I've previously noted, we give the player more choices because if we don't, they will complain.  We have to mollycoddle their sense of agency.  A computer opponent does not complain, so we make it fulfill a role.  The game has enduring value because the original devs created these ideologically competing roles.  It's not narrative skin.  It's a fully thought out diplomatic system that most other 4X games do not have.

Anyways, I'm aware that Domai doesn't have an anti-ideology among the future societies.  That just opens him to a double opportunity to be jarring.  Although even more jarring, is that the game made him Eudaimonic instead of Socialist.  I have corrected that in my mod.  He is so obviously meant to be the socialist archetype, the worker doing workerism, and uses plenty of diplomatic language to that effect vis a vis capitalists.  I figure, the original devs shoved the Aliens in as Planned and didn't have room for Domai as yet another proponent of Planned.  So they made him Eudaimonic.  The expansion was not as well done as the original game, and this is an area where they dropped the ball.

there definitely should be an option to play against stronger AIs

When stronger is simply a form of cheating, by breaking the game's established rules and norms, I'm opposed to it.  You of course are taking the position that "AIs should play like cutthroat humans who don't care" and I'm never going to accept that point.  Nor would I accept a game about WW II where Stalin can do what he jolly well pleases vis a vis the Allies.  That might be a game about world war, but it is not about WW II.

You're breaking the simulation of the game's ideological politics, deliberately, according to what you think will be the best game mechanical strategy for advantage.  Why not just let Zhakarov go Fundamentalist because it would be the most effective for him, since he can pay for the loss to his tech?  The answer is, because he's Zhakarov and his ideological role is actually important to the game mechanics.

Not that it's definitely required in any way, but if you constrain AI actions in significant ways, it's going to have its own consequences.

Game design is consequences.  Of course, when one mods a game design, it gets fuzzy.  I've extensively modded a lot of stuff.  I have my doubts that you've ever played my work.  I'm wondering if you object to anything I did, as strongly as I've objected to 2 things you did.

There was a 3rd thing, at the beginning, but I think that was more about your understanding of alphax.txt and tech research weights at the time.  I'm going to hazard a guess that if you've been modding the game this long, your understanding of those weights has evolved a great deal.  For instance, you're probably aware of the stock binary's "beelines to Hunter-Seeker Algorithm" encodings, according to various research choices.

Separating topics, BRB after research.

[bvanevery]

Related to my earlier comment "That one example of AI attempting to blackmail and then asking for a pact is weird, but it should be only vanilla AI logic." Now I can't seem to find the original message in which you noticed this AI diplomacy behaviour. Do you have a savegame of it just before you open the diplomacy or from the turn before it? That would be necessary for reproducing the issue.

Details magnanimous insults

magnanimous insults

This is like what Lal pulled a lot earlier.

Details this is so passive aggressive

this is so passive aggressive

There's something wrong with the diplomatic AI.  It's incoherent.  You don't bully someone, then offer to be their friend.  Which will realistically lead to giving techs, not getting them.  Presently, I need all the friends I can get though.  So I take up this ally of convenience.  I really don't need a pain in the rear.

Unfortunately, didn't happen to save a game with that post, in MY 2207.  I do have an earlier save from MY 2204.  You might be able to get the same diplomatic result talking to Zak then.  I also found a save from MY 2209.

Details magnanimous cheat

magnanimous cheat

MY 2147.  This guy's got a lot of nerve.  Especially while standing next to my mindworm!

Details very strange diplomacy

very strange diplomacy

Aside from that really tempting offer you made me before, I'm seriously trying to cut down on the sugar daddy thing.  I'm really trying to remember to kick your ass, m'kay?

That one, unfortunately, was from the game where I was trying to be more abbreviated about my reportage.  Don't have a save from that year.  I have one from MY 2136 and MY 2149.

If I see this "Magnanimous threats" phenomenon again, I'll try, in whatever sleep deprived state I'm in, to remember to save a game right after it happens.  And also the auto-save from the year before it happens.  But, I may fail.

[bvanevery]

Details two headed hydra

two headed hydra

MY 2103.  I got cloned.  I'm also running into the problem of taking too long to settle my 2nd base.  I don't like the land based resources I've been offered.  Maybe I'm a bit of a dummy moving along the coast, if I want to see resources at sea.  Maybe my Gun Foil should have been scouting for me, but the next supply pod was in the opposite direction.  By grabbing it, I happened to double my navy.  I was hoping for a completion event for my Transport.  I made a 2nd Scout so I can do some cheap exploration of the land in front of me, then head into the Sargasso Sea with the Transport alone.