Alpha Centauri 2

Sid Meier's Alpha Centauri & Alien Crossfire => Modding => Topic started by: JoGr223 on January 20, 2022, 12:11:23 AM

Title: Sea factions game analysis and my new sea faction
Post by: JoGr223 on January 20, 2022, 12:11:23 AM
first let's consider some unique things about sea factions assuming they have no special attributes; compared to land factions:
I'm assuming small/standard map and 40-60% sea levels, Thinker 3.0 mod/patch (I do all games on small here, ofc transcend)

pros:
2x free recycling tanks at the start
sea colony pods have 4 moves -> better base placement (especially early ones)
no enemies in the water for the first 30-60 years
1-1-4 scout means they can get comm links earlier than others and use it for tech brokerage
terraforming:
   tidals give always +3E (after restrictions lift), while solars only depending on latitude
   farms likewise +3N which makes it especially strong on very desert/fungal like maps, as there are no deserts in the water
   squares can give even more when facility boosters are made (thats minor bonus as its late)
   u can kill enemy bases with terraformers by lowering levels: sometimes it can be as cheap as 80EC + only 8 former turns (so 8 formers to insta-kill); note: it's usually much more expensive, idk whats the formula here but usually it costs like 300-4k

cons:
sea colony pods cost more than 2x more than land: much slower faction growth - basically u have to always build recycling tanks for new base before u can establish it - top1 disadvantage
no 2-1 "rolling and rainy" squares in water, only 1-1
less frequent bonus resources - need to control bigger territory, therefore:
big distance between bases:
   harder to defend
   harder to speed build SPs with crawlers - another top1
   bigger inefficiency?
terraforming:
   no forests - top2
   no proper mines with 4 minerals (only +2 after 8 turns of terraforming) - top3
   no boreholes
crawlers:
   more expensive sea crawlers: slower pay-off start from them
   no forests or 4m mines - much less ROI
moving to land requires use of transports - so at least -1mineral just for that (likely at some point u would need land at very least for some mines/boreholes)
having hybrid of sea+land bases makes further inefficiency in bases cooperation (despite too big territory):
   need to use transports
   sea/land formers have restricted area of operation

Given such perspective Vanilla Pirates are just ridiculous - with baseline (no attributes) they would already be at huge disadvantage, but no, they additionally get crippling -1effic and -1growth by game designers.

So I'm trying to make a faction that can stay in the sea most game or for long time and still be balanced (or lets say have 80/20 ratio for land/sea bases count). My early attempt shows that it's easy to make a mistake of making faction that is balanced in the water but is OP when it just moves to land ASAP, hence the only solution is to have some kind of disincentive/restriction against going land.
The best would be something like +20% cost of land colony pods, but since I can't implement this (or it's not possible?) I settled with "house rule" restriction that would apply only to human players unfortunately :)

My sea faction is a mod for pirates, looks like this:

name: Nautilus Seasteading Organization
Quote
#DATALINKS1
^LEADER: {Professor Ulrik Svensgaard}
^BACKGROUND: {Unity oceanographer}
^AGENDA: {Libertarianism}
^TECH: {Doctrine: Mobility, Doctrine: Flexibility}

#DATALINKS2
^+1 ECONOMY {Unrestrained entrepreneurship}
^+1 EFFICIENCY {Survivalist attitude is necessary to live outside land}
^-1 GROWTH {Culture does not place high emphasis on family}
^-1 POLICE {Constitution limits central government power}
^House rule: May not build land colony pods until year 2121/2131/2141/2151 depending on map size: tiny/small/standard/large+ {Citizens unwilling to settle on land}
^{May not use Police State Politics.}
^Free {Naval Yard} with discovery of {Doctrine: Initiative}: skilled shipwrights
^Enhancements may be built in ocean and trench squares with the discovery of Advanced Ecological Engineering: {Trained for life at sea}
^Sea colony pod & sea former prototypes {free}: advanced oceanic technology
^Bonus mineral from ocean shelf squares: {Culture and technology adapted for the ocean}
^{Marine Detachment ability} free for all naval combat units upon discovery of {Adaptive Doctrine}

AI settings: "-1, 0, 0, 1, 1,"

I made a benchmark isolated 1 player scenario, where the goal is to build PTS ASAP on standardized map; I run recently same scenario but on land standardized map for a few factions; obviously it's not possible to completely port this kind of land map to sea, so the standardization was limited to number of special resources which was almost the same as on land map, but energy squares instead of rivers. I run this without +1 effic and build PTS in 2152; came to conclusion +1effic is needed as they were 4 years behind Gaia, 8 years behind morgan and 10 years behind univ. and since my land benchmark map is kinda realistic I came to conclusion on random map they would likely be 5+ more years behind others due to the fact that porting this land map to sea map as close to possible created unrealistic sea map "too good to be true", one that likely wouldn't occur with random maps generator due to "less frequent bonus resources" on sea.

My main goal was to create balanced and interesting to play sea faction for humans; how it fares in the hand of AI is not a decision factor, but i did run two quick games where AI controls them :
1 game  2154 they were the weakest; with 6 sea and 4 land bases; dem/planned/wealth
2 game 2135 they are the strongest (but small margin compared to few others top AIs)! with 4 sea and 2 land bases; none/FM/none
so results may vary for AI, with this limited testing it seems they are ok; note that I altered critical setting for Thinker: 1st game was on minimal distance of bases, 1 square; 2nd on default - 2 sqares (base_spacing=3);

I also played a few games vs AI as Seasteaders and here again found without +1effic they are too weak; my results varied a lot depending whether I had early enemies who didn't let me get on land - if I engage into any early war then I'm unlikely to get ahead for a long (50+) years if ever since I may not be able to get any SPs; and if u dont go land ASAP then it may be not possible to colonize it without stealing land already taken, which leads to harsh war :)
if u got straight to land-only game (make no sea colony pods), then +1eco +1effic is likely OP; but also from the coherence/lore/storyline perspective I found it pointless if u make a sea faction that doesn't really stay sea - just like if let's say believers had no research penalty; I mean any kind of penalty against going land should be core for any kind of sea faction I suppose
Title: Re: Sea factions game analysis and my new sea faction
Post by: bvanevery on January 20, 2022, 01:21:50 AM
Some assumptions to examine.

Thinker mod still has +1 minerals in the water for an AQUATIC faction, right?  Like in the stock game.  If so, this is widely acknowledged to be a huge, overpowered advantage.  Sure you don't have any single big source of minerals, but you don't need them.  Kelp will grow you big bases, and all you have to do is plant it.

You should try playing The Will To Power mod where the +1 minerals is taken away.  You definitely have to think harder about what you're doing as the Pirates.

Nobody's making you spread out your sea empire, that's your choice.  Granted, if you get map generation where there isn't a lot of shallow water, you might not be able to make a "nicely round" empire.  But you're still not required to fling yourself far afield.

Completion scumming at the beginning of the game, will get your sea colony pods done in a hurry.  Again try WTP if you want more pain in that regard.  You only get this paltry 20 mineral bonus instead of a completion.

In the original game, my mod, and Thinker, you can spend money to complete SPs.  In WTP it's onerous, because it uses a 6:1 cash ratio for SPs, instead of 4:1 like everyone else.  Anyways, you don't have to use supply crawlers.  Also, sea supply crawlers are a thing.  Except again in WTP, where supply crawlers are not a thing.
Title: Re: Sea factions game analysis and my new sea faction
Post by: JoGr223 on January 20, 2022, 10:11:29 AM
Thinker mod still has +1 minerals in the water for an AQUATIC faction, right?  Like in the stock game.  If so, this is widely acknowledged to be a huge, overpowered advantage.
why you think this is overpowered ? that would be OP only on really dry land maps, where most players get 0-0, or 0-1 squares, no 1-1 and 2-1; most maps land will have a lot of 1-1 and some 2-1; so I'd say +1m for aqua is just making it equal to land factions on most maps; my preference for maps is weak/rare/dense so here aqua has worse area, on average/average/average I'd say it's only equal in this regard of 1-1 squares (and still worse than land given all other things I mentioned);

but this does make interesting disincentive for land factions no to go sea as they don't get this benefit

Quote
Sure you don't have any single big source of minerals, but you don't need them.
you mean u dont need them to enjoy the landscape and fresh sea air ? I agree; otherwise I question your skill if u say u don't need minerals to win game


Quote
Kelp will grow you big bases, and all you have to do is plant it.
many big bases are only good on citizen level where u dont have drone problems; on transcend this is not true; so for transcend play it's much better to have 2-1 rolling and rainy on land and have forest spread instead of kelp spread
Title: Re: Sea factions game analysis and my new sea faction
Post by: bvanevery on January 20, 2022, 03:24:29 PM
Thinker mod still has +1 minerals in the water for an AQUATIC faction, right?  Like in the stock game.  If so, this is widely acknowledged to be a huge, overpowered advantage.
why you think this is overpowered ?
Because the ocean already gives you abundant food and abundant energy, enabling you to grow big bases, so those +1 minerals really add up.  Also bear in mind that in the later game, you have the Subsea Trunkline as well as an increase in the amount of minerals that sea minerals give you. 

Quote
Quote
Sure you don't have any single big source of minerals, but you don't need them.
you mean u dont need them to enjoy the landscape and fresh sea air ? I agree; otherwise I question your skill if u say u don't need minerals to win game
Note the bold.

Do you have an After Action Report demonstrating your success or failure with oceanic tactics?  I have several; go look in the AAR forum.
Title: Re: Sea factions game analysis and my new sea faction
Post by: Nexii on January 20, 2022, 05:36:10 PM
Probably with crawler spam default Pirates are not OP. Sea crawlers and formers are a lot more costly.

So yea if you nerf crawlers, aquatic factions become OP as they have much higher FOP - 3/1/3 and that's without even getting aquafarm, thermocline, or +2 ECON

Faction boosts I've found have a lot less impact than unit cost and FOP when trying to balance out a third type of faction that relies on fungus
Title: Re: Sea factions game analysis and my new sea faction
Post by: bvanevery on January 20, 2022, 08:49:48 PM
Fusion powered sea crawlers are easy to belt out quickly.  I've filled up oceans with them, harvesting the energy for my super research capitol.  I honestly can't tell you how huge a difference it made, as I got definitely feelings of "been there, done that" after a game or two of such.  But I have to disagree with the idea that there's anything difficult about belting out sea crawlers.  You can do that.
Title: Re: Sea factions game analysis and my new sea faction
Post by: Alpha Centauri Bear on January 24, 2022, 10:19:17 PM
Thinker mod still has +1 minerals in the water for an AQUATIC faction, right?  Like in the stock game.  If so, this is widely acknowledged to be a huge, overpowered advantage.
why you think this is overpowered ? that would be OP only on really dry land maps, where most players get 0-0, or 0-1 squares, no 1-1 and 2-1; most maps land will have a lot of 1-1 and some 2-1; so I'd say +1m for aqua is just making it equal to land factions on most maps; my preference for maps is weak/rare/dense so here aqua has worse area, on average/average/average I'd say it's only equal in this regard of 1-1 squares (and still worse than land given all other things I mentioned);

Hmm. Did you actually played a lot of vanilla games with Pirates or by Pirates? As AI they are 80% of time first in development and 20% of time they are second if they fall into some pretty disadvantageous ocean configuration. When I played them it took some time to build transports while bases still small but by turn 50 I usually have largest empire colonized land and sea alike.

So yes. It is proved to be overpowered by practice.

Your analysis is flawed because you compare tile to tile. While you should at the very minimum compare base to base and even better empire to empire. The average ocean tile is about same as average land one. Maybe slightly less valuable but just slightly. However, guaranteed immediate nutrient surplus generates higher population and higher number of bases pretty quickly. So you easily may have four time bigger population by turn 50 until some high tech land terraforming kicks in. And even with slightly inferior tile yield you still have triple total value or something. That is all without that aquatic bonus yet. With that you just grant +1 mineral per working tile which you already have in excess comparing to other factions and you get yourself largest empire production powerhouse.

but this does make interesting disincentive for land factions no to go sea as they don't get this benefit

I wouldn't say it is interesting. The game just falls into two largely not intersecting worlds effectively limiting player expansion and flexibility to escape from tiny starting island.
Title: Re: Sea factions game analysis and my new sea faction
Post by: bvanevery on January 25, 2022, 12:07:40 AM
I wouldn't say it is interesting. The game just falls into two largely not intersecting worlds effectively limiting player expansion and flexibility to escape from tiny starting island.
Yeah and for this reason my world generation doesn't allow tiny starting islands, and ships are available immediately on Turn 1.  It proved necessary.  I eventually found it to be the only way to put this problem out to pasture, once and for all.

It has the side effect of making oceanic exploration immediately available to everyone.  All those Artifacts that one could go fish for.

You also get better intelligence about the map and disposition of other faction's empires, since oceanic scouting is expected from the start of the game.

I mean we're not cave people walking across the Bering Strait.  Why shouldn't we be in the oceans immediately?
Title: Re: Sea factions game analysis and my new sea faction
Post by: JoGr223 on January 26, 2022, 05:01:53 PM
about Pirates faction:
As AI they are 80% of time first in development and 20% of time they are second if they fall into some pretty disadvantageous ocean configuration.
ok, let's fact-check this statement with simple scientific experiment  ;zak;: I will run up to 5 games to see if they are really either top1 in development (I assume you mean powergraph) in 80% cases. In test I will not play, just switch to observe AI and press enter to skip to some agreed year. If by your opinion they are so strong that means probably they get high growth-momentum from early on and they will happen to be top1-2 since early years ? Either tell me by which year does your statement hold true: I would assume it is at most year 60, since you mentioned year 50? Tell me what kind of setting this statement requires to be true - I'd assume any with at least big sea since you didn't limit this statement anyhow. Either tell me what setting do you usually play when you observed this (especially: was it transcend?) For test I propose transcend, standard map weak/rare/dense 40**% sea + game rules in attachement - as this is kind of settings I like. As opponents: morgan, univ, peace, gaia, hive, drones. Does your statement hold true only in vanilla SMAX or also in Thinker mod? IIRC your mod is based on Thinker mod ? ? So could I use Thinker 3.0 with default** thinker.ini params ? I'd prefer Thinker mod but if you think your statement is only true in vanilla SMAX then I'll run it there.

Your statement will hold true if in 4 games they are top1 and 1 game >=top2.

BTW: did you mean precisely "first in development" OR "first in development after me" ? If the former then question: what was your ranking in 80% of cases ?

** In Thinker you can actually set specific sea percentage in .ini so that would be preferable as this way experiment is better standardized; if u only allow vanilla then I propose 30-50% sea but on large map and make sure they start in big sea; even if it would turn out only 30% sea then still they should be at advantage since other factions get 70%/6=11,6% of land on average; so if this "big sea start" is around 15% of all water on map it should be fair compared to land factions
Title: Re: Sea factions game analysis and my new sea faction
Post by: bvanevery on January 26, 2022, 06:35:28 PM
If you want more benchmarks on AQUATIC factions with +1 minerals bonus intact, you might consider my SMACX AI Growth mod.  It is designed to run on the stock binary and makes no change to the game binary at all.  Just *.txt file modding.  Completely different alphax.txt and all factions rebalanced.

Lately I think my Pirates are back to being a powerhouse.  Historically I've noticed 2 issues that can cripple them:
I've done things to mitigate these problems.  1. is solved by providing a "backup" predefined unit, in case the expected unit designing fails.  I have Clean Reactors available from the start of the game so it's a Clean version of the unit, that's provided.  2. is merely mitigated by providing various Clean Transport units.  The Pirates still obsess about them, but at least they don't run out of SUPPORT for having done so.  This allows their productivity to just keep humming along.

A major difference between my Pirates and stock game Pirates, is mine pursue a Passive Wealth strategy.  They've got this huge ocean all to themselves, with abundant minerals, so why should they be invading anybody?  Why pester people off the coasts in this annoying "Santiago of the sea" fashion?  No reason for it, not a good strategy.  So they sit back and build, with all the abundant resources they've got.  And the results at times in my modding, have been terrifying.  Like #1 on the graph by a wide margin, planetary Governor, no question they're ahead.

I had the Morganites in a similarly terrifying modality on land for awhile.  But I think that's back when I had way too many ECONOMY bonuses in my Social Engineering table.  They haven't been scary like that for a long time, although they're still quite capable.

Also to note:
Title: Re: Sea factions game analysis and my new sea faction
Post by: JoGr223 on January 26, 2022, 07:11:40 PM
If you want more benchmarks on AQUATIC factions with +1 minerals bonus intact, you might consider my SMACX AI Growth mod.  It is designed to run on the stock binary and makes no change to the game binary at all.  Just *.txt file modding.
im not into total-conversions mod; besides your paradigm of only txt modding will produce inferior results in terms of AI skill than source code modding like Thinker mod does.

Quote
my mod is designed for Huge maps
I prefer small maps as there you already control 30-50 bases which is a ton; standard should be called "huge" as there you may end up with 200 bases; this is just not fun to operate on such scale - it's too much grinding; on huge it would be likely 300 bases; and if someone plays with less bases than terrain allows then its just weak strategy
Title: Re: Sea factions game analysis and my new sea faction
Post by: Alpha Centauri Bear on January 26, 2022, 07:25:26 PM
about Pirates faction:
As AI they are 80% of time first in development and 20% of time they are second if they fall into some pretty disadvantageous ocean configuration.
ok, let's fact-check this statement with simple scientific experiment  ;zak;: I will run up to 5 games to see if they are really either top1 in development (I assume you mean powergraph) in 80% cases. In test I will not play, just switch to observe AI and press enter to skip to some agreed year. If by your opinion they are so strong that means probably they get high growth-momentum from early on and they will happen to be top1-2 since early years ? Either tell me by which year does your statement hold true: I would assume it is at most year 60, since you mentioned year 50? Tell me what kind of setting this statement requires to be true - I'd assume any with at least big sea since you didn't limit this statement anyhow. Either tell me what setting do you usually play when you observed this (especially: was it transcend?) For test I propose transcend, standard map weak/rare/dense 40**% sea + game rules in attachement - as this is kind of settings I like. As opponents: morgan, univ, peace, gaia, hive, drones. Does your statement hold true only in vanilla SMAX or also in Thinker mod? IIRC your mod is based on Thinker mod ? ? So could I use Thinker 3.0 with default** thinker.ini params ? I'd prefer Thinker mod but if you think your statement is only true in vanilla SMAX then I'll run it there.

Your statement will hold true if in 4 games they are top1 and 1 game >=top2.

BTW: did you mean precisely "first in development" OR "first in development after me" ? If the former then question: what was your ranking in 80% of cases ?

** In Thinker you can actually set specific sea percentage in .ini so that would be preferable as this way experiment is better standardized; if u only allow vanilla then I propose 30-50% sea but on large map and make sure they start in big sea; even if it would turn out only 30% sea then still they should be at advantage since other factions get 70%/6=11,6% of land on average; so if this "big sea start" is around 15% of all water on map it should be fair compared to land factions

That would be good experiment. I would like to see results. If you can standardize it - it would be even better test.
I usually play on standard map. Don't remember ocean coverage. Probably one of the average too. For the composition of the other factions I usually play random of SMAC+SMAX together. So you can pick some subset or make it random too - should not matter much for that.
For the year, I cannot tell exactly. Probably year 50 should be good enough to let everybody develop. First few years could be quite erratic.

One important condition - it should be vanilla where they get +1 aquatic mineral. I recall they are not that powerful in WTP anymore. So use vanilla.

Keep in mind that my "statement" stems from perception and observation. I cannot guarantee these 80% and 20% to hold exactly true but this is what I can remember from past games that every time I check power bars Pirates are first across all (including me) and I am occasionally surprised if they are not. For my statement above that means precisely "first in development".
I don't by all means insist on these numbers above and would love to be corrected by statistics.



Another important condition. It should be Transcendent difficulty. Otherwise, human may easily go first and disrupt the statistics. However, if you are not going to play that may be not of a concern.



To reiterate, I am using "Faction Dominance Chart" as it is called in game.



I also like to adjust my time marker. I think 50 turns is too not enough for them to overcome their initial slowness. I recall they become a powerhouse toward turn 100 something. That doesn't meant they cannot do it by turn 50 in some games but turn 100 should be more consistent test.
Title: Re: Sea factions game analysis and my new sea faction
Post by: Alpha Centauri Bear on January 26, 2022, 08:05:12 PM
Just one first random game I have tried to do my experiment as well.

Title: Re: Sea factions game analysis and my new sea faction
Post by: bvanevery on January 26, 2022, 08:54:39 PM
im not into total-conversions mod;

The way most people use the term "total conversion modding", is when you take 1 game like Half-Life, and turn it into a completely different looking game like Counterstrike.  That's not the nature of my work at all.  Frankly, my modding work of the "big 3" modders is the closest to being like SMAC, while still being a mod that is trying to improve everything.  It is, quite simply, "much improved SMAC".

WTP in contrast, uses a rather different combat system among other innovations, does not colonize anything like the original game, etc.  Thinker mostly keeps original game rules but with some notable exceptions.  Like ignoring faction SE compulsions, so you get Deirdre who's Planned exhorting you to go Green or she'll make war on you.  Or nerfing the global warming disasters.  Something that, both fortunately and regrettably, WTP has reinstated something like the original gameplay.  I'm actually scared, terrified even, of global warming in WTP.  You haven't mastered SMAC until you've survived a mindworm apocalypse!

Thinker doesn't try to improve SMAC.  It tries to improve SMAC's AI, which is an important distinction.  And last I checked, only by improving colonization and terraforming.  It's just as dumb as the stock binary about combat, last I checked.  So it might out-produce me, but how is it going to out-fight me?

Quote
besides your paradigm of only txt modding will produce inferior results in terms of AI skill than source code modding like Thinker mod does.

That's a tall claim.  I haven't played straight Thinker in quite some time, but last I checked, it made no improvement in combat ability at all.  Its only claim to fame was colonization and terraforming ability.  So if you find a strategy that leverages combat, where the AI is weak, you can clean up.  Perhaps I should try it now, just to debunk your claim.

Your point of view is only valid in theory or principle.  In practice, mostly 1 person has to do all this work, of designing the game and tuning the AI systems.  Productivity wise, I can run rings around Thinker and WTP, because .txt modding is so much more tractable than binary modding.  And yet I've got 3.5+ years of calendar time into my modding.  Even with low hanging fruit, there's that much scope of complexity to the game.

Sure, 1 person with infinite development time could produce better results than I have so far.  But nobody in the real world has infinite development time.  Firaxis didn't.  I improved upon their work, because there was plenty of scope for improvement.  A binary coder killing themselves to do that development, can only last so long before they burn out.  I'm impressed that Thinker and WTP are both still at it, but I know from plenty of years in open source, that they're unlikely to keep it up forever.

At least I've probably crossed my personal finish line.  After 3.5+ calendar years, there's very little left that could be done to improve the game with only *.txt modding.  Maybe 1 weak area I still have, is I may have broken Economic Victory and perhaps should revisit it.  But nobody's clamored for it yet, and I don't care much for it myself, so it's really really low priority.

Quote
Quote
my mod is designed for Huge maps
I prefer small maps as there you already control 30-50 bases which is a ton; standard should be called "huge" as there you may end up with 200 bases; this is just not fun to operate on such scale - it's too much grinding; on huge it would be likely 300 bases; and if someone plays with less bases than terrain allows then its just weak strategy

AFAIAC my modding works on Huge maps just fine.  Without Huge, the game would be no challenge.  You can always stomp a close quarters enemy in the original game.  I've beaten up jerky close quarters opponents with mere Scouts.

Are you sure your expectations of AI quality, aren't heavily conditioned by playing on really, really easy maps?

WTP AFAIAC does not currently work on Huge maps.  It colonizes way too much.  Produces, defends, and attacks way too little.  Tim is aware of this and agrees that fixing colonization is the low hanging fruit.  The question is how / when.  Which again is the problem of, who is the 1 person with the energy to tune the AI up properly.  I did all that for my mod.  Put tons of work into it.  Already burned myself out.  Not really looking to be the burn-out guy for WTP too.  I limit my contribution to mostly playtesting, and if I happen to see something specifically actionable and easy that could gain a lot of results.  Mucking with its AI gore ain't it.  I'm supposed to be writing my own 4X TBS from scratch.

Haven't tried Thinker lately, so don't know.  You're the 1st person who's made me consider caring about that, in quite some time.  I really don't like this sales pitch that "binary modding is always better AI".  Because I haven't seen any evidence that that's actually true.

You really have a great deal of faith in your own opinions of how the game works, and what results one can get with modding.  Before we just end up in never ending pointy arguments about stuff, would you consider, that 3 of us have been at this rather a substantial number of production years now?  2 of whom are disagreeing with you about Pirates.  Could it be possible, that we know something from our testing that you don't.

But of course, do your own benchmarks.

And I'd also be interested to know, if you ever play my mod, and you comment on what is or isn't challenging about it.  Because it's basically stock AI given an environment where it can excel more, at whatever it's good at.  Mostly.  I did nerf probe teams, I can't stand that stuff.  And of course it's still going to have various original AI weaknesses.  But you're not going to be out-Boreholing the AI the way Thinker does, because you're simply not allowed to make them until late game.  If you've been relying on such tactics for your playing strength, in my mod work, you'll be in for a rude awakening.  You can do all that sort of thing eventually.  I didn't remove those play mechanics, but you probably should have won the game by other means before then.

Like you can claim "my AI is inferior" all you want, without actually playing it.  Doesn't mean anything until you do.
Title: Re: Sea factions game analysis and my new sea faction
Post by: Alpha Centauri Bear on January 26, 2022, 10:34:09 PM
Something that, both fortunately and regrettably, WTP has reinstated something like the original gameplay.  I'm actually scared, terrified even, of global warming in WTP.  You haven't mastered SMAC until you've survived a mindworm apocalypse!

I'm all for including other mods good finding. Let me know if some Thinker features need to be imported.

I'm impressed that Thinker and WTP are both still at it

I can say, honestly, I am not at it 100% anymore. Mostly mopping some lingering issues and responding to user comments. I'd say most of the features I ever really wanted are implemented already and what's left is combat AI as you correctly pointed out. I tried to tackle it few times and realized that efforts are about 100 time more than I have already invested so I am just sitting there hoping I'll get 100 more free time at some point.
😁
Title: Re: Sea factions game analysis and my new sea faction
Post by: JoGr223 on January 26, 2022, 11:02:54 PM
2 of whom are disagreeing with you about Pirates.  Could it be possible, that we know something from our testing that you don't.

yes, thats why I proposed way to verify your views; I supposed it could be true Pirates AI is stronger than other AI in vanilla, since I now only play Thinker mod where they were v.weak; but they are still mostly average now as AI despite I gave them substantial boost in +1eco and +1 effic (effective +2 effic compared to original -1effic)

BUT YOU ARE WRONG !  ;danc

Did you actually played a lot of vanilla games with Pirates or by Pirates? As AI they are 80% of time first in development and 20% of time they are second if they fall into some pretty disadvantageous ocean configuration.
this is completely wrong as now evidenced by scientific experimentations ! very easy to reproduce my results

TEST SETTINGS:
transcend, large map weak/rare/dense 30+% sea + pirates have at least 50% sea access at game start + game rules in screenshot in previous post; opponents: morgan, univ, peace, gaia, hive, drones.

test consist of just opening game switching to "view only mod" (various factions) and pressing enter till 100years pass

smax vanilla 2.0

RESULTS:
coverage setting/estimated actual ocean coverage/pirates sea access (as % of all water); all games pirates sea start give them access to at least 15% of whole map, so 50% of sea (30%/50% at least)/rank:

game1: 30-50%/40%/90%/5th
game2: 30-50%/30%/65%/5th
game3: 30-50%/40%/90%/5th
game4: 50-70%/60%/95%/7th
game5: 70-90%/80%/100%/2nd

Notice how they suck in game2-3 despite starting in the jungle; mostly this is a factor which predicts strongest faction (even here: gaia in game1, lal and drones in game4)

I planned to only run this experiment on 30-50% settings, but since they were so weak I raised it in their favour; turns out it only helped with 70-90% which are obviously extremely stacked conditions in favor of pirates and against land factions; in this game they actually won elections around year 60, but lost another elections ~30years later and ultimately placed 2nd

anyway main point is that statement like "AI is good in games therefore faction X is strong" is wrong methodology - AI is stupid so you can't make conclusions based on how AI plays - unless you want to design balance for the comfort of AI players; even if they were strongest as AI they may still be weakest for human players; I run this test just to check whether Alpha Centauri Bear actually knows anything about this game  :P 8) since he made such bold and confident statement about pirates

also idk where this myth about pirates come from ? maybe cause they are just cool and on paper no enemies for a long time seems OP ? I created reddit poll on strongest factions for human players (not AI) and they won tying with hive (univ and aliens were banned from poll), each got 4 votes:
https://www.reddit.com/r/alphacentauri/comments/rkahk0/poll_rate_your_overall_best_faction_for/
maybe ppl just find it easier to play with them since no enemies

comparing to Thinker mod vanilla AI is such a joke, there they would have covered all map with 3-6x bases and 2-3x more tech and projects


proof:
Title: Re: Sea factions game analysis and my new sea faction
Post by: Alpha Centauri Bear on January 26, 2022, 11:54:54 PM
Hmm. Interesting. Maybe map size matters. Let me do the same testing.

AI playing certain faction good does not 100% prove that the faction bonuses themselves are good. I agree. However, it is a good approximation and the only test we can make. Everything else is just words.
Title: Re: Sea factions game analysis and my new sea faction
Post by: JoGr223 on January 27, 2022, 12:06:15 AM
Hmm. Interesting. Maybe map size matters. Let me do the same testing.

I also did 2 tests on standard/50-70% and results were conclusive (top3,top5)
Title: Re: Sea factions game analysis and my new sea faction
Post by: bvanevery on January 27, 2022, 12:29:38 AM
I'm impressed that Thinker and WTP are both still at it

I can say, honestly, I am not at it 100% anymore. Mostly mopping some lingering issues and responding to user comments. I'd say most of the features I ever really wanted are implemented already and what's left is combat AI as you correctly pointed out. I tried to tackle it few times and realized that efforts are about 100 time more than I have already invested so I am just sitting there hoping I'll get 100 more free time at some point.
😁

Well, yeah, I'm obviously not at it 100% anymore either.  I'm quite pleased to say the gap between my last 2 releases, was 6 months.  That's new for me.  Indicates I've moved into a totally new stage of maintenance.

Competent combat AI is the stuff that commercial 4X games are made of.  I'm not going to kill myself on SMAC AI in that regard, unless for some reason I suddenly saw how it could be easily improved.  Like if I was poking at OpenSMACX one day and saw, hmm, what if I changed these few lines.

Rather much like making Clean Reactors available from the beginning of the game, was a huge improvement in the AI's performance.  Because it had the serious deficiency of running itself out of SUPPORT.  It's a hack, that's totally justified if you have no intention of changing the game binary at all.

Wouldn't it be funny if there's actually more AI combat competence in the game binary, and Firaxis simply turned it off, before shipping?  Like it played too well.  I mean, how do you explain the hesitancy of some of these "infantry base surround" rushes?  Seems like they could go for the jugular a lot faster than they actually do.

What language I'm going to prototype a new combat system and combat AI in, is an issue that pops in and out of my head every few days now.  Then of course I get sidetracked again by real life things, like needing to replace my car battery.  I know way more about that now than I should have to.
Title: Re: Sea factions game analysis and my new sea faction
Post by: bvanevery on January 27, 2022, 12:49:00 AM
test consist of just opening game switching to "view only mod" (various factions) and pressing enter till 100years pass

100 years?  That's not a test of anything.  On a Huge map, I've barely got my own empire infrastructure underway by then.  I've probably popped a lot of supply pods, because I'm aggressive that way.  The only way you're going to get in a serious war with anyone in that timeframe, is if they happened to start right next to you.  And you probably have to choose such a war.  It's probably not going to be forced on you by actual map constriction.

Caveat: the Conquer oriented factions may get Recon Rovers by then and do a rush at you.  Which if you haven't prepared for, is rather lethal in my mod.  Not a problem if you've prepared, but there's a "press your luck" game mechanic to how you develop your empire.  The AIs don't succumb to each other's Recon Rover rushes because they stack their defenses with Clean Synth Garrisons early and often.  You try that on the AIs, you'll just bounce.  Not unless you've got a lot of infrastructure to really pump out the rovers.

Maybe I didn't pay good enough attention to what the original terms of the squabbling were about.  But "the Pirates are overpowered as of 100 years" doesn't mean anything.  Of concern, is whether the Pirates are overpowered over the course of the game.  Can they whip your butt, or other factions' butts.  With their +1 minerals bonus.

I haven't tried stock game in a long time and frankly don't care about it.  The stock game can reduce their power for other reasons.  Like the bugs I mentioned, and the unsuccessful aggression strategy.  It annoys the human player and doesn't actually achieve anything for the Pirates.  Now you change a couple of bit flags in pirates.txt and, uh, things change!

Here are my pirates.  Of course they are meant to exist with an alphax.txt that fixes various other problems.
Title: Re: Sea factions game analysis and my new sea faction
Post by: Alpha Centauri Bear on January 27, 2022, 12:54:58 AM
Strangely Pirates don't do well in vanilla. Don't know why and don't know where I got this impression they are always ahead. Maybe in Thinker/WTP version before I removed +1 mineral. Maybe it gets them more optimized? Well, whatever.
Title: Re: Sea factions game analysis and my new sea faction
Post by: bvanevery on January 27, 2022, 01:09:23 AM
I can test my pirates.  I just don't want to use up my 1 laptop right now, that I'm typing on, for the task.  Spare laptop is in a room with a lot of construction dust that I don't want to go into.  Ditto the desktop computers.  This house is a mess.  The traditional time to do this sort of thing would be while watching TV.  And those rooms are gross right now.  I'm washing clothes, sheets, towels, and taking showers multiple times a day.  I vacuum a lot.  Gawd!

At some point, I could finally read my Mom's old college papers on the USSR.  And have my laptop running at the same time.  But right now, I've still got a bit of real work I need to do on my laptop.  Like figuring out car batteries, grr!

Testing resources... bane of my existence.
Title: Re: Sea factions game analysis and my new sea faction
Post by: JoGr223 on January 27, 2022, 12:16:35 PM
100 years?  That's not a test of anything. 

But "the Pirates are overpowered as of 100 years" doesn't mean anything.
I feel offended about this. no test or evidence of anything is precisely what your posts are about

Quote
my mod
Quote
my mod
Quote
my mod
x10 or so

please stop derailing this topic with these ads/spamming of "your mod"; i dont care about your mod; this is topic about MY MOD; and my mod is Thinker mod + new faction

Quote
I haven't tried stock game in a long time and frankly don't care about it.
I feel the same about your mod
Title: Re: Sea factions game analysis and my new sea faction
Post by: bvanevery on January 27, 2022, 03:49:37 PM
100 years?  That's not a test of anything. 

But "the Pirates are overpowered as of 100 years" doesn't mean anything.
I feel offended about this. no test or evidence of anything is precisely what your posts are about

Feel free to dial back your level of hostility in this forum at any time.  You're new around here.  It's ok to have a pointy argument.  It's not ok to keep taking potshots at my integrity.  I've most certainly "earned my place" as far as how this game mostly works, even if I don't know every phenomenon possible in the game.  No one could know... something you need to remember, when saying other people's claims are baseless.   "Tests", what do you think I've been doing in my modding work for 3.5+ years?  I've run more AI vs. AI tests than you can shake a stick at.  It's how I ensure my factions are balanced.

If I wasn't clear enough the 1st time: "100 years" is like this game's equivalent of saying "Hello World" in programming.  It doesn't prove the strength of a faction over the course of the game.  There are bonuses that factions get beyond the first 100 years, like the Pirate minerals increase, Subsea Trunkline becoming available, and the ability to work the Ocean not just Ocean Shelf.  200 years is bare minimum for determining how a faction can do.  300 years, aka Mission Year 2400, is the point at which you should start being able to answer, "Can this AI faction win the game, or has it stagnated?"  AI factions in my mod do actually win the game by about 2450, usually by Transcendence.  It happens regularly enough, at least when only playing other AIs.

please stop derailing this topic with these ads/spamming of "your mod"; i dont care about your mod; this is topic about MY MOD; and my mod is Thinker mod + new faction

Your thread is titled "Sea factions game analysis and..."

You opened a discussion of how to do a sea faction.  I've been there, done that, in spades.  I have relevant experience.  You may not like hearing the experience, but that doesn't change how the game works.
Title: Re: Sea factions game analysis and my new sea faction
Post by: Nexii on January 29, 2022, 04:32:33 PM
Strangely Pirates don't do well in vanilla. Don't know why and don't know where I got this impression they are always ahead. Maybe in Thinker/WTP version before I removed +1 mineral. Maybe it gets them more optimized? Well, whatever.

Pirates really suffer from the negative EFFIC. Sea empires tend to be the most spread since you have to put cities offshore. So pretty quickly half their bases make very little energy, which is the main resource from water (tidals are much stronger than mines).

If vanilla AI prioritized EFFIC more, I think Pirates would do okay in the vanilla game. So it's probably that they go for EFFIC more in Thinker if I had to guess.
Title: Re: Sea factions game analysis and my new sea faction
Post by: Nexii on January 29, 2022, 04:47:14 PM
As for Pirates balance in Thinker I'm not sure it makes sense to boost up an aquatic's faction bonuses. If they just settle onto land then they win easily. The imbalance is more of a land vs sea thing to solve. You could either make sea units cheaper, or sea FOP a bit higher, or adjust land vs sea terraforming times.

It's funny cause I once considered sea FOP to be overpowered in crawlerless play. In peaceful games the high energy focus leads to incredible levels of commerce. I recall one game where Believers were on the sea the most and out researching everyone (many treaties/alliances and global trade pact). It kind of opened my eyes to the potential. I feel like Pirates are a faction that would do better with Pacifist mindset. Is the Thinker AI more war-like or about the same? That could be part of it too. I feel like a sea faction should run away with it if everyone is peaceful, sea tiles make 5E each with +2 econ and thermocline. Much faster to plonk down than boreholes.
Title: Re: Sea factions game analysis and my new sea faction
Post by: bvanevery on January 29, 2022, 04:52:53 PM
Pirates really suffer from the negative EFFIC.
My pirates don't have EFFIC (JUSTICE) or GROWTH penalties.  They also focus on Explore Build, not Explore Conquer.  A pure Build focus definitely centralizes an empire's development more.  An Explore focus tends to scatter it to the winds, especially on water.  I haven't thought much about their Explore Build hybrid strategy lately.  I'd make them pure Build except that that's Morgan's schtick.  In my mod, all 14 factions have a unique combo of research priorities.  Which do, for some reason, also affect AI behavior.  I've got "coverage" of all the possible AI strategies, except for "study nothing" and "study everything".

As for Pirates balance in Thinker I'm not sure it makes sense to boost up an aquatic's faction bonuses. If they just settle onto land then they win easily.
Fortunately not a problem in the stock binary.  Pirates will mostly stay in the water.  It's not like Thinker's "these are pirates??" landlubbering stuff.
Title: Re: Sea factions game analysis and my new sea faction
Post by: JoGr223 on January 29, 2022, 05:40:35 PM
I feel like Pirates are a faction that would do better with Pacifist mindset.

yea, I designed them this way, kind like sea-morgan; but I overlooked some stats since I was mostly focusing on human play; I adjusted today "-1, 0, 0, 1, 1," and after 2 games of AI tests I see they are much stronger

Quote
Is the Thinker AI more war-like or about the same?
I noticed in late game (~120+years on standard map) Thinker AIs tend to go total war with each other, but not sure, might depend on many factors; early game if u play "Aggressive AI' settings - then definitely, or when I recently played "hive the noble peaceful man" after 50-70 years or so everyone gradually break pacts with me and then i get sudden attack from morgan; another game as morgan in ~2271 year: I get sudden "pact break and attack" from lal, despite morgan I run dem; note this game i was observcing AI all game and IIRC morgan and lal were allies most/all games but then I took over and play morgan in 2268.
Title: Re: Sea factions game analysis and my new sea faction
Post by: bvanevery on January 29, 2022, 08:50:34 PM
I adjusted today "-1, 0, 0, 1, 1," and after 2 games of AI tests I see they are much stronger

Yeah I did those exact settings in my mod version 1.8.  June 10, 2018.  3.5+ years ago.  Also changed their Agenda from Power to Wealth.  I'm on mod version 1.52.  Been doing them that way for a long time.  Have, at times, had to adjust tech tree stuff to enhance their play.  Like what part of the tech tree is Doctrine: Initiative in.  And fix aforementioned pirate specific bugs.
Title: Re: Sea factions game analysis and my new sea faction
Post by: EmpathCrawler on January 29, 2022, 10:15:39 PM
I feel like Pirates are a faction that would do better with Pacifist mindset.

yea, I designed them this way, kind like sea-morgan; but I overlooked some stats since I was mostly focusing on human play; I adjusted today "-1, 0, 0, 1, 1," and after 2 games of AI tests I see they are much stronger

Quote
Is the Thinker AI more war-like or about the same?
I noticed in late game (~120+years on standard map) Thinker AIs tend to go total war with each other, but not sure, might depend on many factors; early game if u play "Aggressive AI' settings - then definitely, or when I recently played "hive the noble peaceful man" after 50-70 years or so everyone gradually break pacts with me and then i get sudden attack from morgan; another game as morgan in ~2271 year: I get sudden "pact break and attack" from lal, despite morgan I run dem; note this game i was observcing AI all game and IIRC morgan and lal were allies most/all games but then I took over and play morgan in 2168.


Were you and Lal the top two in the power rankings? I'm pretty sure that the AI is designed so that the top two factions won't get along as a sort of mandatory Social Darwinism, even if your politics and behavior align. I think that's the standard behavior and that Thinker doesn't do much to diplomatic AI, at least not according to the notes on GitHub. You may have also triggered some AI anti-human logic by taking Morgan over.
Title: Re: Sea factions game analysis and my new sea faction
Post by: JoGr223 on February 01, 2022, 07:47:45 PM
IIRC morgan was top2 and lal top3; hive was top1; ill post more on this game soon
Templates: 1: Printpage (default).
Sub templates: 4: init, print_above, main, print_below.
Language files: 4: index+Modifications.english (default), TopicRating/.english (default), PortaMx/PortaMx.english (default), OharaYTEmbed.english (default).
Style sheets: 0: .
Files included: 31 - 840KB. (show)
Queries used: 15.

[Show Queries]