Sea factions game analysis and my new sea faction

[JoGr223]

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

#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

[bvanevery]

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.

[JoGr223]

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

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

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

[bvanevery]

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. 

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.

[Nexii]

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

[bvanevery]

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.

[Alpha Centauri Bear]

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.

[bvanevery]

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?

[JoGr223]

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  : 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

[bvanevery]

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:

• a bug where they don't make sea formers
• an obsession with producing way too many transports

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:

• my mod is designed for Huge maps
• my continent generation makes big continents and big oceans, like Earth

[JoGr223]

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.

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

[Alpha Centauri Bear]

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  : 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.

[Alpha Centauri Bear]

Just one first random game I have tried to do my experiment as well.

[bvanevery]

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?

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.

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.

[Alpha Centauri Bear]

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.
😁