Author Topic: SMACX AI Growth mod  (Read 59518 times)

0 Members and 1 Guest are viewing this topic.

Offline Douglas

Re: SMACX AI Growth mod
« Reply #930 on: December 27, 2019, 08:26:44 PM »
It was only a Fission PB, and most of the line was supply crawlers stretching far to either side of the base. There might have still been an open path to it, I'm not sure of all the details, but if so it required taking a detour far to the side that would take multiple turns of movement. I only needed three actual needlejets to block the blast radius itself, and could have used crawlers for that too if I put them a little farther out.

Offline bvanevery

  • Emperor of the Tanks
  • Librarian
  • *
  • Posts: 4433
  • €701
  • View Inventory
  • Send /Gift
  • Allows access to AC2's quiz & chess sections for 144 hours from time of use.  You can't do without Leadship  Must. have. caffeine. -Ahhhhh; good.  
  • Planning for the next 20 years of SMACX.
  • AC Text modder Author of at least one AAR
    • View Profile
    • Awards
Re: SMACX AI Growth mod
« Reply #931 on: December 28, 2019, 12:54:56 AM »
I've seen the AI found a base at sea, just to get a platform to launch a nuke from.  I rage quit a game because of that once.  It wasn't just that the AI did this, it's that my policy of global flooding didn't work.  Anything I'd destroy, the AI would just rebuild.

Offline bvanevery

  • Emperor of the Tanks
  • Librarian
  • *
  • Posts: 4433
  • €701
  • View Inventory
  • Send /Gift
  • Allows access to AC2's quiz & chess sections for 144 hours from time of use.  You can't do without Leadship  Must. have. caffeine. -Ahhhhh; good.  
  • Planning for the next 20 years of SMACX.
  • AC Text modder Author of at least one AAR
    • View Profile
    • Awards
Re: SMACX AI Growth mod
« Reply #932 on: January 22, 2020, 09:16:30 AM »
CHANGELOG
Changed from version 1.38 to 1.39:

FACTIONS:

- Cult of Planet: changed 2 errors in Faction description.  Said it got +1 PLANET and +1 SUPPORT.  Both are wrong, and have been wrong for a long time.

WORLD SIZE:

- Giant planet: changed to 126x252.  Playtester ih8regin suggested that allowing a map coordinate to exceed the range of an 8-bit integer, might expose more pathfinding bugs.

CONQUER FOCUS:

- Command Nexus: moved to C3 Adaptive Doctrine.  When it was given by C3 Neural Grafting, the Morganites would typically be the ones to get it, because that tech is partly a Build tech.  This feels inappropriate.  I think a faction doing pure Conquer research should have the best chance of getting it, so I've put it in a pure Conquer part of the tech tree.
- Single-Sided Surfaces: now has C3 Field Modulation as a prereq, instead of C3 Neural Grafting.  The predefined units granted by this tech, only have 1 ability.  Field Modulation is a pure Conquer tech and Neural Grafting isn't, so this gives a pure Conquer faction the best chance to discover the tech first.
- Nonlinear Mathematics: now has C3 Adaptive Doctrine as a prereq, instead of C3 Field Modulation.  This keeps it from having the same prereqs as C3 Single-Sided Surfaces.  Thematically, Field Modulation feeds the defensive techs, and Marines feed the offense techs.
- Doctrine: Initiative: now has E3 Monopole Magnets as a prereq, instead of C3 Adaptive Doctrine.  I had accidentally made this tech and C4 Advanced Military Algorithms have the same prereqs.

EXPLORE FOCUS:

- Doctrine: Loyalty: set growth=0.  A Punishment Sphere is not a practical means to grow in the early game.  It only has value when base sizes are much larger.  This tech is now a pure Build research barrier, in the same way D1 Information Networks is a pure Discover research barrier.  Factions without a Build focus can trade for it, or eventually discover it by popping a supply pod.
- Neural Grafting: set growth=0.  It doesn't give the Neural Amplifier anymore.  It doesn't offer any mindworm, growth, or exploration benefit.
- Bioadaptive Resonance: now has E3 Centauri Genetics as a prereq, instead of C3 Neural Grafting.  This is for Explore research focus continuity.  Having 2 abilities is no longer guaranteed, nor is Non-Lethal Methods.
- Doctrine: Air Power: set growth=1.  Needlejets do not aid substantially in the exploration of Planet by the time they appear in the game.  Nor are Needlejets capable of popping supply pods.  Air Transports are not good at moving Artifacts or Colonists as a base or airbase is required to load them.  Although Needlejet Colonists could settle new territory quickly, the AI doesn't know how to use them.  The Explore research focus should be more directed towards mindworms and not distracted by techs that give little exploration benefit.
- Nanominiaturization: set growth=1.  Fuel air cells speed up Needlejets, at a noticeable additional cost.  They are a tweaky and rarely used feature of the game.  Needlejets do not aid substantially in the exploration of Planet by the time they appear in the game.
- Fusion Power: set growth=1.  The only benefit for exploration is ships move a little faster.  Psi combat does not care about reactor size.
- Orbital Spaceflight: set growth=2.  By the time a Sky Hydroponics Lab is launched, most of Planet's surface is already known anyways.  They do not provide substantial amounts of food until all bases have Aerospace Complexes, which are expensive and take a long time to complete.
- The Will To Power: now has C6 Organic Superlubricant as a prereq, instead of C6 Photon/Wave Mechanics.  The former isn't great lore, but it has Explore focus continuity.  The latter doesn't, and that made it a hard barrier to Explore research.
- Industrial Nanorobotics: set growth=1.  Completing the Space Elevator allows orbital insertions.  They can be useful for popping the last supply pods, but there are almost none left by this stage of the game.
- Graviton Theory: set growth=1.  Although orbital insertions are useful for popping the last supply pods, there are almost none left by this stage of the game.  Gravships cannot pop supply pods.
- Quantum Power: set growth=0.  Ships move a little faster, but exploration is over by now.  Psi combat does not care about reactor size.
- Singularity Mechanics: set growth=0.  Ships move a little faster, but exploration is over by now.  Psi combat does not care about reactor size.

UNIT COSTS:

- Infantry chassis: now costs 2.  Unarmored infantry units with an offensive weapon are no longer a cheap deal.  They now have expense proportional to the cost of the weapon.  This is achieved both by increasing the Infantry chassis cost and raising weapon costs.
- Speeder chassis: now costs 3.  This is to make it incrementally more expensive than Infantry.
- Hovertank chassis: now costs 4.  This is to make it incrementally more expensive than a Speeder.
- Needlejet chassis: now costs 16.  This is calibrated to cost the same as a Hovertank chassis.
- Copter chassis: now costs 16.  This is to match the Needlejet chassis.
- Gravship chassis: now costs 20.  This is calibrated to cost twice as much as an Infantry chassis.

- Particle Impactor: now costs 4.  Same as the original game.  The new policy is that weapons up to the Missile Launcher are Fission Era weapons with low costs.
- Gatling Laser: now costs 5.  Same as the original game.
- Missile Launcher: now costs 6.  Same as the original game.
- Chaos Gun: now costs 13.  The new policy is that weapons up to the Tachyon Bolt require a Fusion Reactor to be affordable.
- Resonance Bolt: now costs 29.
- Plasma Shard: now costs 58.  The new policy is that weapons up to the Graviton Gun require a Quantum Reactor to be affordable.
- Chronoton Gun: now costs 69.
- Graviton Gun: now costs 106.
- Singularity Laser: now costs 213.  The new policy is that weapons up to the Black Hole Gun require a Singularity Reactor to be affordable.
- Black Hole Gun: now costs 255.  This is the maximum cost any weapon can have.  It's only an 8-bit unsigned quantity and rolls over if it's any higher.
- Terraforming Unit package: now costs 3.  Since the Infantry chassis is more expensive, this reduction was needed for land based Formers to stay at the same cost.  It has the effect of making Sea Formers cheaper.  This shall not be regarded as a problem.
- Supply Transport package: now costs 13.  This makes a Fusion Infantry Supply cost 50, same as an Artifact.
- Probe Team package: now costs 3.  Since the Infantry chassis is more expensive, this reduction was needed for land based Probe Teams to stay at the same cost.  It has the effect of making Foil Probe Teams cheaper.  This shall not be regarded as a problem.

- Plasma Steel Armor: now costs 2.  Since the Infantry chassis is more expensive, armors have to be less expensive to match the costs usually expected in the early game.  There isn't room for much gradation of cost.  Plasma is now always better than Synthmetal in every respect.  This was generally true in practice anyways.
- Pulse 3 Armor: now costs 3.
- Resonance 3 Armor: now costs 3.
- Silksteel Armor: now costs 3 and has armor rating 4.  I wanted an armored Garrison unit with only a Gun weapon to cost 40.  However, the integers only allow me to have costs of 30 or 50.  I chose 30 and lowered the protective value of the armor to compensate.
- Photon Wall: now costs 4.
- Psi Defense: now costs 4.  In the original game, it had the same cost as a Photon Wall.
- Probability Sheath: now costs 7.  Requires a Fusion Reactor to be affordable.
- Neutronium Armor: now costs 10.
- Pulse 12 Armor: now costs 11.
- Resonance 12 Armor: now costs 11.
- Antimatter Plate: now costs 24.  Requires a Quantum Reactor to be affordable.
- Inertial Damper: now costs 38.  Requires a Singularity Reactor to be affordable.

PREDEFINED UNITS:

- Scout: set cost=1.  I want to retain 2 kinds of Scouts that only cost 10.  The new cost for No Armor would make it cost 20.
- Trance Scout: set cost=1.
- Clean Scout: set cost=2.  I want it to cost 20.  The new Infantry chassis cost would make it 30.
- Clean Trance Scout: removed.  E4 Bioadaptive Resonance no longer guarantees that 2 abilities are available.  The AI wasn't building it anyways.  Various other units with 2 abilities were already removed in previous revisions because they were too expensive.
- Clean Police: removed.  The new Infantry chassis cost would make it cost 40.  I don't want special cases for most Scout designs.  It was a bit of an exploit that most abilities could be gained with a Scout for free.
- Clean Plasma Police: removed.  C4 Single-Sided Surfaces no longer guarantees that 2 abilites are available.
- 3-Res Police: removed.  E4 Bioadaptive Resonance no longer guarantees that Non-Lethal Methods is available.  The AI wasn't building it anyways.
- Clean Synth Speeder, Clean Synth Police Wagon, Synth Police Wagon: removed.  With the new Infantry and Speeder chassis costs, Synthmetal armor is no longer free on a Speeder.
- Clean Synth Garrision: reinstated and now available with C1 High Energy Chemistry.  This tries to distract the AI from building too many garrison units and running out of SUPPORT.  To give the AI the best possible chance, Synthmetal Armor does not need to be prototyped.
- Transport: removed.  The AI does not need a predefined Transport unit to design and build them.
- Clean Transport: reinstated in the same slot that the Transport used to be in.  The Pirate AI obsesses about building Transports regardless of whether any Transport unit is predefined.  At least it can be made to build Clean ones so as not to run out of SUPPORT.
- Clean Recon Rover: removed.  The AI was building a lot of them for combat.  It should build just plain Recon Rovers.  It's better for the AI to invest in clean armored units for standing defense.
- Clean 3-Pulse Laser Foil: removed.  The AI obsessed about building it in the Fusion era when it was no longer appropriate, and losses were high.
- Clean Colony Pod, Clean Scout: removed bogus 30th ability flag.  These units are base types with specially assigned costs and are available at the beginning of the game.  It would be irrational for the player to design similar units.
- Trance Scout: removed bogus 30th ability flag.  This unit has a specially assigned cost, is only available when Hypnotic Trance is discovered, and is then immediately available without prototyping.  It would be irrational for the player to design a similar unit.
- Formers: put back in the same slot in the predefined unit table as the original game.  When Clean Formers are available, the AI mostly prefers to build those.  It doesn't matter what order they're in.
- Clean Sea Formers: removed.  The Terraforming Unit package is cheaper now, so a plain Sea Formers only costs 20 minerals.  Most factions start on land and do not have a problem running out of SUPPORT from building too many Sea Formers.  The Pirates do not seem to obsess about building Sea Formers, and they have a +1 minerals bonus on ocean shelf squares to compensate for any SUPPORT problems.
- Sea Formers: reinstated in the same slot in the predefined unit table as the original game.  Available with B2 Industrial Base.  The Foil chassis is not guaranteed to be available yet, but this tech is a prereq for E3 Ecololgical Engineering which gives the Aquafarm.  It would be annoying not to have a Sea Former available for that.  This unit is only predefined as a safety valve anyways, to overcome a bug where Sea Formers are sometimes never designed by the AI.  The Pirates are supposed to have this unit available from the start anyways.  If they suffer the bug, at least they are a pure Build faction and B2 Industrial Base is directly on their research path.


The attachment limit per post is 5, and I eventually ran out of room at the top of the thread.  Consequently, this becomes the new home of version 1.39. It was downloaded 86 times before I moved it here.
« Last Edit: June 25, 2020, 07:16:20 AM by bvanevery »

Offline bvanevery

  • Emperor of the Tanks
  • Librarian
  • *
  • Posts: 4433
  • €701
  • View Inventory
  • Send /Gift
  • Allows access to AC2's quiz & chess sections for 144 hours from time of use.  You can't do without Leadship  Must. have. caffeine. -Ahhhhh; good.  
  • Planning for the next 20 years of SMACX.
  • AC Text modder Author of at least one AAR
    • View Profile
    • Awards
Re: SMACX AI Growth mod
« Reply #933 on: February 07, 2020, 08:20:45 AM »
There's a problem with the University starting the Hunter-Seeker Algorithm way too quickly.  It has consistently occurred in my last several games with them.  I started a test game with them in it, just to be sure.  I got the Believers, and I did fine for being on a small island with the Cyborgs.  We traded techs and kept up.  But the University started the HSA on turn 78.  Previous game was similarly before turn 100, I think also in the 70s.  At least, they got C4 Advanced Military Algorithms then.  They didn't start the HSA because they had 3 other SPs to build.  I changed the weights on MilAlg before this test game, but it didn't solve the problem.

What seems to be happening, is the University climbs rapidly up the Discover tree, as I intended.  It gets to D4 Pre-Sentient Algorithms.  Then it doesn't have the Build prereqs for D4 Applied Relativity.  Rather than learn those, it seems to pathologically beeline up the Conquer part of the tree, completely ignoring Build and Explore.  It's a strange behavior as they're a Passive, pure Discover faction.  I would have thought the distribution between the techs would be more random, but it doesn't seem to be.

They're also very willing to give away C3 Adaptive Doctrine, which gives the Command Nexus.  2 games in a row, they've traded it to me for nothing.  So of course I built it.  Don't know what's up with that.

I could just make the HSA more expensive, but I don't feel that's enough.  I don't want the U. being able to get this so fast.  At times I've not even had probe teams by then, due to bad luck.  Even with a partial Conquer focus.  I can't have the U. completing the HSA before other factions even get a chance to steal from them.

This is a PITA because I'm somewhat lacking flexibility in my tech tree nowadays.  I'm going to have to stare at it really hard to figure out how to reshuffle this.  If I have to make a sacrifice somewhere, I think it would be ok to make the U. less scary fast on tech.  I'm really not sure what's freed up with them in the early tech tree, but they just go gangbusters now.

Maybe I need to take the University's free Biology Labs away and clip their tech wings.  At one point in development, it was definitely needed.  But maybe with the new "Discover beeline" tech tree, they don't need it anymore.
« Last Edit: February 07, 2020, 08:36:34 AM by bvanevery »

Offline bvanevery

  • Emperor of the Tanks
  • Librarian
  • *
  • Posts: 4433
  • €701
  • View Inventory
  • Send /Gift
  • Allows access to AC2's quiz & chess sections for 144 hours from time of use.  You can't do without Leadship  Must. have. caffeine. -Ahhhhh; good.  
  • Planning for the next 20 years of SMACX.
  • AC Text modder Author of at least one AAR
    • View Profile
    • Awards
Re: SMACX AI Growth mod
« Reply #934 on: February 11, 2020, 07:53:46 AM »
I made Pre-Sentient Algorithms a Conquer 5 tech and put the Hunter-Seeker Algorithm in it.  This delayed the University until turn 115, which is still too early.  They got it before they even got D3 Intellectual Integrity and Knowledge!  I think whenever the University can't research a Discover tech, they pathologically choose a Conquer tech to research next.  You'd think it would be a random selection between Explore, Build, and Conquer, but apparently it is not.  Just goes straight up the Conquer tree.

Because they research so many Conquer techs, I'm tempted to change their personality from Passive to Erratic.  I'll see what further testing yields.

I could put Discover breadcrumbs in the tree, breaking the rigidity of my category separation.

Offline 杨主席

Re: SMACX AI Growth mod
« Reply #935 on: February 22, 2020, 08:38:45 PM »
Really impressive work, man. :D I've only just discovered this mod after years of not playing SMAX. Currently playing as the Cult of Planet (which I just realized was named "Fungboy" in the original files) and already made some terrible strategic decisions, but I really appreciate how smooth the changes make the game feel. What kind of new strategies have you discovered so far while playing on your own mod? I'm still a terrible player and my recollection is very hazy but I recall that back in the day, from what I gathered online at the time, you basically could do some combination of forests, crawlers, and needlejets and the game was pretty much over by then.

Offline bvanevery

  • Emperor of the Tanks
  • Librarian
  • *
  • Posts: 4433
  • €701
  • View Inventory
  • Send /Gift
  • Allows access to AC2's quiz & chess sections for 144 hours from time of use.  You can't do without Leadship  Must. have. caffeine. -Ahhhhh; good.  
  • Planning for the next 20 years of SMACX.
  • AC Text modder Author of at least one AAR
    • View Profile
    • Awards
Re: SMACX AI Growth mod
« Reply #936 on: February 22, 2020, 09:03:46 PM »
Hoo boy, Cult of Planet.  In my mod, you want to go Extremist.  It's designed for them and them only!  They are immune to the negative effect of using it.  It's a +2 SUPPORT bonus, no downside.

Forests still work.  In fact, that's all that works.  You don't get to have factories, crawlers, or thermal boreholes until much later in the game now.

I suppose you could do ordinary Mines if you have enough food, but I only start building those once I've built Hybrid Forests anyways.  I'm scared of eco-damage, and you need to have a lot of food to work mines, when you don't have crawlers.

You get Soil Enrichers at the same time, so that's another option if you weren't into forests.  My Soil Enrichers are faster to build, to compensate for Condensers being slower to build.  They come at the same time, you don't get to have Condensers early.  This was to stop Thinker Mod from exploiting Condensers and Boreholes earlier in the game.

One benefit of my mod, is there's no penalty for building a Mine on farmland.  You can do both.  You're just assumed to be eating toxic metals in your garden produce.   :D  This was to make the AI's life easier.  Benefits the AI more than it benefits you.

I don't think Needlejets are going to benefit you as this great strategy, as they come a bit later, and they're definitely more expensive.  But, try it and see.  You tell me.

I made mag tubes available rather early in my mod.  I'm biased by the idea that I'm going to build a rail straight into the enemy's heart, than zip down it with tons of ground troops.  That's not a Cult of Planet way to play though.  You would want the Xenoempathy Dome, so that you can attack with piles of mindworms from the fungus.

Hovertanks also come early enough to be an alternative to Needlejet based invasion.  And frankly, I always found a Cruiser full of Marines to be the fastest way to conquer the whole world quickly.

Offline bvanevery

  • Emperor of the Tanks
  • Librarian
  • *
  • Posts: 4433
  • €701
  • View Inventory
  • Send /Gift
  • Allows access to AC2's quiz & chess sections for 144 hours from time of use.  You can't do without Leadship  Must. have. caffeine. -Ahhhhh; good.  
  • Planning for the next 20 years of SMACX.
  • AC Text modder Author of at least one AAR
    • View Profile
    • Awards
Re: SMACX AI Growth mod
« Reply #937 on: February 24, 2020, 04:49:00 AM »
CHANGELOG
Changed from version 1.39 to 1.40:

FACTIONS:

- University: removed Information Networks starting tech.  Removed 1 free tech bonus.  Removed strange #xxx comment line.  The University's research was too fast.  They don't need any research boost.

DISCOVER TECHS:

Previously I believed that the shorter Discover tech tree was benefiting the University, but it isn't.  They actually research many non-Discover techs.  They are probably doing well for other reasons, such as having Clean Reactors available from the beginning of the game.  Ergo, the Discover tech tree does not need to be especially short.

- Information Networks: now a D2 tech.  Set wealth=1 and growth=1.  These are breadcrumbs so that non-Discover factions can learn it.  Otherwise it becomes a hard research barrier for most factions.  I'm not worried if a Conquer faction doesn't learn it, as they want immediate military results.
- Knowledge: moved to D3 Secrets of the Human Brain. 
- Secrets of the Human Brain: now a D3 tech.  Set wealth=1 and growth=3.  Knowledge gives a JUSTICE bonus which can be worth a small amount of money, and a PLANET bonus.  It now gives the Virtual World.
- Optical Computers: set power=2, wealth=2, and growth=2.  It now gives the Planetary Datalinks, which can help all kinds of research.
- Intellectual Integrity: now a D4 tech.  Set power=0 and growth=0.  It now gives the Universal Translator, which is only good for more research.
- Thinker citizen type: moved to D5 Digital Sentience.  The advantage isn't needed earlier, and I need to repurpose Pre-Sentient Algorithms.
- Applied Relativity: now a D5 tech.  Industrial Labs should come after Research Hospitals.
- Digital Sentience: set tech=4, wealth=1, and growth=1.  Having tech as 5, probably just makes it easier for non-Discover factions to research.  Cybernetic only gives a small JUSTICE bonus. 
- Nanohospital: moved to D6 Nanominiaturization. 
- Nanominiaturization: now a D6 tech.  Set power=2, tech=4, and growth=3.  It now gives Nanohospital.  Repair Bays and Fuel Nanocells are marginal military capabilities.
- Corporate Lab: moved to D7 Unified Field Theory.  Industrial Lab is a Tier 4 facility and this doesn't need to be coming at Tier 6.
- Unified Field Theory: set wealth=3 and growth=0.  It gives the Corporate Lab and no longer gives the Nanohospital.

CONQUER TECHS:

- Probe Team package: moved to C2 Planetary Networks.  Police State gives a PROBE bonus and it's more rational to actually have the ability to make use of the bonus.
- Field Modulation: now a C2 tech.  D1 Biogenetics was underutilized as a prereq, and this can be a lore fit for that.  The position of Field Modulation is unimportant as a Cloaking Device is very nearly useless.
- Adaptive Doctrine: now a C2 tech.  I don't like having to wait until Tier 3 to get Marines.
- The Command Nexus: moved to C3 Polymorphic Software.  It has to stay at Tier 3.
- Polymorphic Software: now a C3 tech.  I don't mind if the AI is delayed in using Probe Teams.
- Neural Grafting: set power=3 and wealth=4, making it a B3 tech.  This has been functioning more like a Build tech, and Tier 3 needs another Build tech.

UNIVERSITY AI:

- Covert Ops Center: moved to C5 Pre-Sentient Algorithms. 
- Pre-Sentinet Algorithms: now a C5 tech.  Set power=4, tech=0, and growth=0.  It now gives Covert Ops Center and the Hunter-Seeker Algorithm.  Now has B4 Planetary Economics as a prereq.

The University AI is cheating, using Directed Research when it should be using Blind Research, to beeline for the Hunter-Seeker Algorithm.  If I disable the HSA from the game, the University doesn't bother to research C5 Pre-Sentient Algorithms.  If I relocate it to E5 Centauri Psi, it makes an extraordinary effort to learn that tech.  If the HSA is available with C4 Advanced Military Algorithms, the University will start it by turn 75.  If the HSA available from C5 Pre-Sentient Algorithms, but still having only Conquer prereqs, the University will start it by turn 100..115.  With B5 Planetary Economics as a prereq, the University will start it by turn 130..140, and other factions can beat them to finishing it.  This isn't great lore, and it's bad research continuity, but it blunts the AI's cheating.

EXTREMISTS:

- Doctrine: Loyalty: now an E2 tech.  Set wealth=3 and growth=4.  Extremist is only designed to benefit the Cult of Planet, who can use it without penalty.  It should be on the Explore research path for them.

BIG BASES:

- Hab Complex: moved to E2 Adaptive Economics.  It used to be at E3 Monpole Magnets, which can take a long time to research.  I went 120 turns as the Pirates without learning it, nor did anyone else in the game.  Meanwhile my kelp fed bases hit their size limit.

BUILD HAPPINESS:

- Cyberethics: set wealth=2.  I wish to de-emphasize the contribution of happiness to a Build focus.
- Monopole Magnets: set wealth=2.   I wish to de-emphasize the contribution of happiness to a Build focus.
- Sentient Econometrics: now a B7 tech.  Set tech=0, wealth=4, and growth=2.  It no longer gives the Corporate Lab.  The Self-Aware Colony can actually lower a city's effective police, because it doesn't use Non-Lethal Methods.

EXCESSIVE WEIGHTS:

- Advanced Military Algorithms: set power=4 and wealth=1.  Having power as 5, made it too easy to research.  Power gives a SUPPORT bonus but also JUSTICE and GROWTH penalties.  I want to deemphasize this as a Build tech.
- Planetary Economics: set wealth=4.  Having wealth as 5, made it too easy to research.
- Orbital Spaceflight: set power=4.  Having power as 5, made it too easy to research.
- Biomachinery: set power=3 and growth=4.  Having growth as 5, made it too easy to research.

MINIMAL BENEFIT:

- Biogenetics: set power=1.  A Biology Lab is not useful for conquest at the beginning of the game, as nobody knows how to make mindworms yet.
- Centauri Ecology: set power=1.  Roads are useful for conquest, but they aren't critical.  At the beginning of the game, any conquest is going to involve a lot of units walking or speeding over undeveloped land.
- Centauri Genetics: set power=1.  I want Mindworms and Spore Launchers to be mostly Explore techs, to differentiate them from a Conquer focus.
- Progenitor Psych: now an E3 tech.  Set power=1.  I want Hypnotic Trance to be mostly an Explore tech, to differentiate it from a Conquer focus.  I'm intrigued by the idea of it coming later, instead of being an easy early defense against mindworms.  This also makes human-Progenitor communication completely one sided.  Progenitors will initiate, not the humans.  In the stock game that would be a problem because the Progenitors are overpowered, and it's not fun to get stomped by an enemy you can't negotiate with.  But in my mod it's not a problem, because the Progenitors have the same strength as human factions.  When they won't talk, the typical military result is a stalemate.
- Advanced Ecological Engineering: set power=1.  Although faster roads, faster rails, and raising land are useful for conquest, they're simply not as important as better weapons and armor.
- Centauri Empathy: set power=1 and wealth=1.  Although Green allows one to capture mindworms, and they are powerful at the beginning of the game, having only a few of them is limiting.  Building a pile of Lasers is a much more certain way to kill an enemy.  The JUSTICE bonus is not going to make anyone any money at the beginning of the game.
- Planetary Networks: set wealth=1.  Although Police State and Non-Lethal Methods make people happy, it's hardly a monetary competititor with Democratic.  The JUSTICE penalty can also hurt money.  I want this more strongly differentiated from Build techs at the beginning of the game.
- Organic Superlubricant: set wealth=1.  Hovertanks are useful for terraforming fungus and rocky terrain, but the value of this should not be overstated.  They cost more and cheaper units can also get the job done.

SECRET PROJECTS:

- The Human Genome Project: moved to E3 Cyberethics.  Story-wise it bugged me having it at E3 Centauri Genetics.
- The Virtual World: moved to D3 Secrets of the Human Brain.  Set ai-tech=1 and ai-colonize=1.  Hologram Theaters are cheap in my mod.
- The Planetary Datalinks: moved to D3 Optical Computers.  It's a better lore fit.
- The Universal Translator: moved to D4 Intellectual Integrity.  Set ai-tech=1.  It isn't a terribly useful secret project.
- The Planetary Transit System: set ai-colonize=1.  I don't think the AI benefits much from it.
- The Hunter-Seeker Algorithm: moved to C5 Pre-Sentient Algorithms.  The University AI cheats, using Directed Research rather than Blind Research to go after this project, no matter where it is in the tech tree.  Moving it to Tier 5 slows the University down a bit.
- The Network Backbone: set ai-colonize=1.  It negates the Cybernetic POLICE penalty.
- The Empath Guild, Clinical Immortality: set ai-fight=0 and ai-mil=0.  2X votes is not explicitly oriented towards conquest.
- The Living Refinery: set ai-fight=0 and ai-mil=0.  A SUPPORT bonus is not explicitly oriented towards conquest.

PREDEFINED UNITS:

- Probe Team, Foil Probe team: moved to C2 Planetary Networks.  Police State gives a PROBE bonus, so let's provide the units actually needed to probe.
- Clean Sea Colony Pod: removed bogus 30th ability flag.  This unit has a specially assigned cost, so it would be irrational for a player to design a similar unit.
- Sea Formers: moved to E3 Ecological Engineering.  This unit is merely a safety valve, in case the AI doesn't automatically design the unit like it's supposed to.  The AI needn't build it prematurely.  It becomes more useful when nutrient restrictions are lifted and Aquafarm is available.
- Rover Formers: new predefined unit available with E2 Adaptive Economics.
- Clean Rover Formers: removed.  It's much more expensive than a Rover Formers.
- Clean 3-Pulse Garrison, Clean 3-Res Garrison: removed.  The AI doesn't build them, and they're much more expensive than a Clean Plasma Garrison.
- Cloaked R-Laser Speeder: removed.  It's much more expensive than a Cloaked Recon Rover, and it gives away a R-Laser prototype.
- 3-Pulse Probe Team: removed.  Although effective, they're twice as expensive as a regular Probe Team.
- 3-Pulse Police: removed.  The AI doesn't build them, and they're much more expensive than Plasma Police.
- Trance Scout: removed.  In the new unarmored cost regime, I did not like that Hypnotic Trance was getting a special dispensation for cheapness.

Offline 杨主席

Re: SMACX AI Growth mod
« Reply #938 on: February 25, 2020, 02:18:19 PM »
Thanks for the tips! I didn't even realize that the Cult had extremist without any penalty. Perhaps I should have spent more time reading and less time staring at worm boils. It didn't make a difference in the end, however, as I got steamrolled by the Free Drones. I had entered in a war against them to keep the pirates happy, who were pressuring me to the west and were the most powerful faction at the beginning of hostilities. They sent me a single rover as I watched my bases fall one after the other to neo-communist conscripts. As Palpatine would say, "Ironic."

I have a few questions if you don't mind. Is there a changelog with all the changes, not just the changes from one version to another? Is there a tech tree to consult, as opposed to fumbling through the datalinks? I don't want to sound like a choosing beggar but I feel it's hard to plan without having a clear idea of where the new techs lead.

After being humbled with the Cult, I've started a new game with Morgan since you mentioned you had specifically beefed him up. It's a ton of fun. My empire is weakly defended and not that powerful (the University has developed EXTREMELY fast in the early game but the Hive ended up catching up somehow) but I am able to keep a tech edge with probes and simply buying stuff at a great price, even from people who hate my guts (which seems to be everyone except Lal). With +4 econ the take I haul in is just indecent and should inspire an "Occupy Borehole Cluster" protest against inequality or something to that effect. The Spartans are way too close for comfort so I try to keep them happy by declaring war on more distant enemies (sorry, Lal. She had too many drop infantry and needlejets for me to survive otherwise).

Offline bvanevery

  • Emperor of the Tanks
  • Librarian
  • *
  • Posts: 4433
  • €701
  • View Inventory
  • Send /Gift
  • Allows access to AC2's quiz & chess sections for 144 hours from time of use.  You can't do without Leadship  Must. have. caffeine. -Ahhhhh; good.  
  • Planning for the next 20 years of SMACX.
  • AC Text modder Author of at least one AAR
    • View Profile
    • Awards
Re: SMACX AI Growth mod
« Reply #939 on: February 25, 2020, 03:24:11 PM »
I got steamrolled by the Free Drones.

Looks like I did my job.   :D  The game has some shelf life for you yet.

Quote
Is there a changelog with all the changes, not just the changes from one version to another?

Yep.  readme_mod.txt.

Quote
Is there a tech tree to consult, as opposed to fumbling through the datalinks?

By selecting "HQ.. Laboratories Status.." or just hitting F2, you get a display of 1 tech, all the techs that connect with it, and an index of all the techs you have researched.  If you have infiltrated other factions, you can click on their icons and get the same info about them.  This UI is different than the Datalinks.  There is in fact a button at the bottom of the tech display which says Datalinks, that will take you to those.

Unfortunately, displaying tech relationships 1 tech at a time, is all that SMAC has got.  An overview is impossible inside the game.

I don't think anyone ever developed a 3rd party "stable" overview printout tool.  I know such have existed at various times, but I don't know of something that is being maintained and is known to be working.  Maybe it exists and I just haven't looked.

Some people have been interested in creating such a tech tree sheet manually for my mod.  But then I go and change the dependencies every release, so they are dissuaded.  I can't yet promise any feature freeze.  I think 1.40 is my last version, but I also said that a year ago, and look where I am now.  I'm sure that changes are on the wind-down, as I'm getting much more serious with moving on with the technology of a commercial project, and there's simply very little to fiddle with anymore.  But I can't promise that I'll never change something in the tech tree.  I haven't gotten enough playtester feedback and approval for the work, to make that kind of promise yet.  Someone could tell me that something sucks, and they could be right!

Quote
I don't want to sound like a choosing beggar but I feel it's hard to plan without having a clear idea of where the new techs lead.

SMAC was limited in what it offered.  However, as a matter of game design, one could go even farther with what they did offer.  They offered "Blind Research", where you don't just get to pick the next thing you're working on.  Well, what if like in real science, you don't even know what discoveries are possible?  It's a valid way to try to design a game.  It would have quite a burden for player expectations though, and those would have to be met somehow.  You think about the game in terms of planning your tech strategy, because you have that available to you.  Well what if you didn't?  You'd need something else to strategize about.

This kind of question, has me scratching my head and rubbing my chin, about "what I would do" in a similar 4X TBS title.  Choosing "what you're going to discover" is teleological.  It's non-sensical.  But as a play mechanic, what would I replace it with?

An even more vexing question, is what it means to "win".  In real life, is the United States "winning" ?  It won WW II, but it hasn't actually won anything since then.  If your goal isn't to take over the planet, then what's to win?  The idea of winning, is either archaic, or despotic.

Quote
With +4 econ the take I haul in is just indecent and should inspire an "Occupy Borehole Cluster" protest against inequality or something to that effect.

;rotflmao  I've had to do some rebalancing of ECONOMY over time.  At one point, AI Morgan was pressing Economic Victory very early in the game, for the low low price of only 1000 credits!  My answer to that was to make it impossible before B9 Global Energy Theory.

Offline 杨主席

Re: SMACX AI Growth mod
« Reply #940 on: February 25, 2020, 08:16:26 PM »
Quote
Looks like I did my job.   :D  The game has some shelf life for you yet.

Shelf life? Probably an entire supermarket left since I'm not even past Librarian  :D

Quote
SMAC was limited in what it offered.  However, as a matter of game design, one could go even farther with what they did offer.  They offered "Blind Research", where you don't just get to pick the next thing you're working on.  Well, what if like in real science, you don't even know what discoveries are possible?  It's a valid way to try to design a game.  It would have quite a burden for player expectations though, and those would have to be met somehow.  You think about the game in terms of planning your tech strategy, because you have that available to you.  Well what if you didn't?  You'd need something else to strategize about.

This kind of question, has me scratching my head and rubbing my chin, about "what I would do" in a similar 4X TBS title.  Choosing "what you're going to discover" is teleological.  It's non-sensical.  But as a play mechanic, what would I replace it with?

Perhaps introducing rogue-like features in a tech tree would work? Conceptually, it would avoid the nonsensical teleological aspect you've highlighted. It would however be a nightmare to properly balance, or end up being mile wide and inch-deep like many rogue-like features.

4X games feel satisfying in part because you can see the consequences of your strategic choices unfold over the long term. Some random aspects of the game (the terrain you land on, the fortunes and positions of other civilizations, etc.) do not generally create frustration for the most part because if they act as creativity-boosting constraints for your strategic planning and add depth and character to your game's story-line à la Dwarf Fortress. However, if the randomness restricts the player too much and dwarfs (lol) the consequences of their own actions, that's when the frustration can set in e.g. a truly terrible start that even an experienced player can't do much about. In a way, this draws parallels with the hypothesis of autonomy, mastery and purpose being the trifecta behind intrinsic motivation. Overly punitive randomness attacks all three, which is why real life can be a ;miriam; sometimes.

Quote
An even more vexing question, is what it means to "win".  In real life, is the United States "winning" ?  It won WW II, but it hasn't actually won anything since then.  If your goal isn't to take over the planet, then what's to win?  The idea of winning, is either archaic, or despotic.

At this point, I guess the medium itself becomes the limitation. I have no idea what a game allowing paradigm shifts like this would look like. Perhaps to approach the complexity of real life it would necessarily have to be a simulation indistinguishable from it.

I'd definitely be down for a sandbox title based on the SMAX universe, though. There's the GURPS book, but that's about it.

Offline bvanevery

  • Emperor of the Tanks
  • Librarian
  • *
  • Posts: 4433
  • €701
  • View Inventory
  • Send /Gift
  • Allows access to AC2's quiz & chess sections for 144 hours from time of use.  You can't do without Leadship  Must. have. caffeine. -Ahhhhh; good.  
  • Planning for the next 20 years of SMACX.
  • AC Text modder Author of at least one AAR
    • View Profile
    • Awards
Re: SMACX AI Growth mod
« Reply #941 on: February 25, 2020, 09:30:21 PM »
What should the player have control over?  It's a deep question.  It needs solutions for the 4X genre to advance.  The player controlling everything, is not a solution.  The endgame is all that tedious unit pushing.

A losing condition, is definitely easier than a winning condition.  For instance the human race can lose by destroying itself in nuclear armageddon, allowing sufficient global warming to flood the whole planet and fry us all to death, failing to get into space before a big planet surface destroying asteroid hits us, or before the Sun goes red giant.  But if the human race were to survive all these things, it's still not clear what *winning* means, let alone for any subsection of the human race.  You just get to keep playing the game of Life.

Chairman Yang said it thus: "...life's only purpose, is Life, itself."

Offline bvanevery

  • Emperor of the Tanks
  • Librarian
  • *
  • Posts: 4433
  • €701
  • View Inventory
  • Send /Gift
  • Allows access to AC2's quiz & chess sections for 144 hours from time of use.  You can't do without Leadship  Must. have. caffeine. -Ahhhhh; good.  
  • Planning for the next 20 years of SMACX.
  • AC Text modder Author of at least one AAR
    • View Profile
    • Awards
Re: SMACX AI Growth mod
« Reply #942 on: February 29, 2020, 03:41:57 AM »
I hate Foil Probe Teams.  Yes, I know I put them in my mod.  For years I've used them as an exploit against the AI factions, and I only thought it fair that they get to do it back.  Problem is, it's way better for them than it is for me.

Sea bases are incredibly vulnerable to them.  And what do you have when playing the Pirates?  Sea bases.  It's generally not possible to make a "good" cluster of bases around your capitol in the ocean, to increase the cost of enemy mind control.  There aren't enough oceanic resources in one spot to make centralized empires like that.  One of your bases will inevitably be far away, sitting on some nice resource.  And here comes the enemy Foil Probe Team!

In my last game, I was able to shell a fair number of them into oblivion, because the AI was stupid enough to stop its movement just short of my sea base.  That's just the accident of where they left port from, and the distance covered.  Eventually though, a different faction came from a different direction and nailed me.

I quit the game.  I was cramped, stuck between lots of players, and couldn't even learn C2 Planetary Networks which gives probe teams.  My ally Zhakarov had the tech for some time but wouldn't sell it to me, because he's a dick.  He'd finally loosened up and probably would have sold it to me in a few turns, once I'd scraped the money together.  That's when this happened.

Happened the previous 2 games as well.  I'm in a "Pirate mastery" kick.  It seems like getting lots of completions from supply pods at sea, should be an exploit that would put them seriously ahead of everyone else.  But somehow, it's not working out.

Granted, one game my empire was fine, and I was big and scary, but I had such a long distance to cross to get to the Gaians.  Was sending all kinds of Recon Rovers and Impact Rovers across fungus, using the Xenoempathy Dome.  I stole the tech from the Gaians and had huge money to get things completed.  But pushing the units over land got really old.  I really didn't like this land-and-sea empire I was building, so I quit.

I wonder what it cost the Believers to subvert that base?  I suppose I should look it up.  I've tried to adjust the costs to be expensive, but I wouldn't be shocked if it isn't working.

Looking at the Believer's credits on the previous turn, I'm guessing they spent roughly 400 credits to take my sea base.  It is size 5 with 1 Medic.  It has a Children's Creche, a Recreation Commons, and a Biology Lab.  It is 16 squares away from my capitol as the crow flies, 18 by water.  That's not that far away.  I have 11 bases and +2 JUSTICE from Democratic Green.  The base is only suffering -2 energy due to corruption and waste.  The Believers have +1 PROBE from Police State.  My PROBE is 0, so my bases cost the standard amount to take over.  Normal factions in my mod have PROBECOST=150, a 50% increase in cost over the stock game.

I'm bumping PROBECOST to 200 for starters.  See if that ends up being more reasonable.

« Last Edit: February 29, 2020, 04:52:17 AM by bvanevery »

Offline bvanevery

  • Emperor of the Tanks
  • Librarian
  • *
  • Posts: 4433
  • €701
  • View Inventory
  • Send /Gift
  • Allows access to AC2's quiz & chess sections for 144 hours from time of use.  You can't do without Leadship  Must. have. caffeine. -Ahhhhh; good.  
  • Planning for the next 20 years of SMACX.
  • AC Text modder Author of at least one AAR
    • View Profile
    • Awards
Re: SMACX AI Growth mod
« Reply #943 on: March 08, 2020, 06:13:30 PM »
As the University I thought I was oh-so-superior.  But I had to tangle with the Usurpers at close quarters.  They got Recon Rovers done faster than I expected, before I could produce a substantial defense.  Maybe I could have, but I didn't quite take them seriously.  I got the Command Nexus finished, with 2 Artifacts, a lot of money from exploring, and a fair number of turns just building it.  I didn't have the productivity to bang out units all that quickly though.  I hadn't done my usual forest plantings, because I had spread out a fair amount, in a partial fungal maze, and needed roads.

The Usurpers whacked one of my border cities with an Elite Recon Rover.  I thought, ok, I'll probe it and then retake it.  Took forever to get a simple probe team done.  By the time I'd finished my infiltration and theft, they had learned Plasma armor and deployed it.  So I've got these chinsy Lasers and not enough of them to do any real damage.  I didn't want to spend money to retake my base, since I'd just doubled the cost of that.  Didn't bother to find out what it would cost me, I just didn't wanna pay it.

They also attacked my capitol hard enough to completely empty it, but they failed to take it over and I retaliated the next year.  The Cyborgs showed up and almost did similarly to a border city, but I held them off.

The game was so bumbly fumbly that I quit.

I learned Monopole Magnets awfully early.  That's a case of beelining to the Hunter-Seeker Algorithm.  Seems to be true of me too, not just the AI.  Scient disassembled the game's code and found out how the cheating works:
Code: [Select]
if (PlayersData[factionID].SE_ProbeBase <= 0) {
if(tech_is_preq(techID, Facility[FAC_HUNTER_SEEKER_ALGO].preqTech,
(PlayersData[factionID].AI_Tech != 0) + 2)) {
if (!PlayersData[factionID].AI_Power) {
valueRet *= 2;
}
if (PlayersData[factionID].AI_Tech) {
valueRet *= 2;
}
}
}

This applies to factions with PROBE <=0.  If you don't focus on Conquer, you get a 2X multiplier.  If you focus on Discover, you also get a 2X multiplier.  The University is getting a 4X strong "pull" towards the HSA.  Basically you can blind research anything you want, long as it contributes to the HSA!  That means things that don't contribute, aren't going to be discovered.  At least until the HSA has been completed, I suppose.

All of the techs I discovered in this test game, were consistent with beelining towards the HSA.  It's actually pretty annoying to have designed a Discover tech tree, that Discover factions can't even use.

I suppose if I put the HSA far enough up the tree, all of the early game techs would be ancestors, and thereby get equal weight.  But that seems to be a pretty drastic price to pay.

Looks like I need to throw in the towel and put the Hunter-Seeker Algorithm in the Discover tech tree.  Otherwise any faction focusing on Discover - including the human player! - won't actually learn the early Discover techs.  Instead they learn a bunch of Conquer techs.
« Last Edit: March 08, 2020, 09:12:10 PM by bvanevery »

Offline bvanevery

  • Emperor of the Tanks
  • Librarian
  • *
  • Posts: 4433
  • €701
  • View Inventory
  • Send /Gift
  • Allows access to AC2's quiz & chess sections for 144 hours from time of use.  You can't do without Leadship  Must. have. caffeine. -Ahhhhh; good.  
  • Planning for the next 20 years of SMACX.
  • AC Text modder Author of at least one AAR
    • View Profile
    • Awards
Re: SMACX AI Growth mod
« Reply #944 on: March 11, 2020, 10:06:09 PM »
semi tech stagnation
semi tech stagnation

Playtesting 1.41, I've run into an oddly slow tech game.  It's MY 2377 and only now, is anyone learning D3 Secrets of the Human Brain.  I made D5 Applied Relativity a prereq for C6 Fusion Power, so nobody's discovered fusion reactors either.  The highest weapons and armors in the game are Chaos and Probability, but I don't think anyone's built any Chaos guns yet.  I just built 1 Probability boat to have it available if I want, but nobody's making it to my shores in any substantial force.

My Missile Needlejets are plenty for sinking the few incoming enemy ships.  I'm building a land bridge to the Spartans.  I've delayed the Xenoempathy Dome in my mod, but I just completed it, so now I'm sending over Locusts and will do a ground invasion with mindworms if necessary.  I don't really care about conquering the Spartans as they're a relatively weak power, I just want to force them to sign a Treaty.

The Data Angels are my opposition.  On the graph, they're supposed to be more powerful than I.  They certainly have double my population and are the Governor I think.  At least I'm not.  Because I'm the Believers, I've been immune to my bases being subverted, but I've still had to garrison them pretty tight with Firewall Probe Teams.  I even  built Covert Ops Centers in most cities, first time I've done that.  I don't have them in the same tech as the Hunter-Seeker Algorithm anymore, and that's yet another Discover tech that hasn't been discovered yet.

I'm wondering how exactly I'm going to stomp the Data Angels before MY 2500.  Total conquest?  The Aliens are in the game.  They're weak but not exterminated.  The Usurpers are my ally.

 

* User

Welcome, Guest. Please login or register.
Did you miss your activation email?


Login with username, password and session length

Select language:

* Community poll

SMAC v.4 SMAX v.2 (or previous versions)
-=-
18 (6%)
XP Compatibility patch
-=-
9 (3%)
Gog version for Windows
-=-
77 (29%)
Scient (unofficial) patch
-=-
28 (10%)
Kyrub's latest patch
-=-
14 (5%)
Yitzi's latest patch
-=-
85 (32%)
AC for Mac
-=-
2 (0%)
AC for Linux
-=-
5 (1%)
Gog version for Mac
-=-
10 (3%)
No patch
-=-
12 (4%)
Total Members Voted: 260
AC2 Wiki Logo

* Random quote

Go through, my children! The time of miracles is upon us. Let us cast off sin and walk together to the Garden of the Lord. With God's mercy we shall meet again on the other side.
~ Sister Miriam Godwinson ’Last Testament’

* Select your theme

*