Optimal Land Use: Cities, and Crawlers, and Formers, oh my!

[Earthmichael]

Although it would be interesting to see purple trees or something, I personally don't use Prog factions enough to make it worth anyone's time to do it.

[Yitzi]

By the way, an important note: Crawlers are only overpowered as compared to "standard" (farm/solar on non-rocky with occasional* mirrors, condensers, and boreholes mixed in).  Forests are less former-time-consuming and therefore incomparable before Hybrid Forests and strictly inferior after Hybrid Forests (until advanced specialists and/or 6-nut spaces).  (Yes, early Hybrid Forests are extremely overpowered IMO.)  For this reason, there probably won't be many crawlers by me in the Market Forces game, as Hybrid Forests are even worse in terms of high-power.

*Say, no more than 1 mirror on average affecting a square, and no more than 1 borehole per base on average.

[Earthmichael]

I have got a GREAT IDEA.  You go ahead and play Market Forces from the start yourself, with the same non-conquest restriction proposed by Kirov.  Play with all of your improved ideas.  Then let's compare the result based on the turn where each of us win by cornering the market, and see who gets there first!  Then we can compare notes to see where each of our strategies stand at varous tech levels, and see what worked best!

I played a few more turns of Market Forces today.  Have you started yet, Yitzi (or anyone else)?  I would enjoy hearing your perspective on the scenario.

[Yitzi]

I played a few more turns of Market Forces today.  Have you started yet, Yitzi (or anyone else)?  I would enjoy hearing your perspective on the scenario.

Not yet, I'll probably start tomorrow.

[Yitzi]

Ok, I've played some, and don't think I'm ready for this.  Morgan is definitely a strong faction if played right...but an accelerated start (giving him one of the more useless early projects) is no help, starting on such a small landmass is no help, and playing on Transcend with the resulting hostile AI is DEFINITELY no help (even if you can often keep them happy up to a point, most won't make treaties, and Morgan needs his treaties and pacts).  After Domai got the Planetary Energy Grid ahead of me, I decided that I'd had enough.

[Earthmichael]

I found only Angels would make a long term treaty/pact; the others may make a short term treaty if you bribe them, but they won't hold to it.

I was not able to get the Planetary Energy Grid either.  But after I got rolling, I was able to get several other good secret projects.

Since you were critical of my Morgan development, I was hoping to see your alternative development strategy.

So given the small land mass, do you still agree with your assessment that you should expand out to sea?  Or do you think my strategy of raising land has more merit that it appeared at first glance?

[Yitzi]

So given the small land mass, do you still agree with your assessment that you should expand out to sea?  Or do you think my strategy of raising land has more merit that it appeared at first glance?

It has more merit than it appeared at first glance, for the simple reason that I didn't realize that tidal harnesses have twice the build time of solar collectors, and didn't take into account that sea formers cost twice as much as formers, so it actually is worth it, simply because forests are so cheap and are still strong with hybrid forests.  (However,

However, I still maintain that hybrid forests are overpowered before AEE at the earliest*; without them, it would be a lot easier to compare what is and is not too powerful.

*They allow a 4-former-turns square to produce resources comparable to a rainy/rolling square with a farm, solar, 1000+ feet, and a nearby mirror, which has a former-time cost of over 20.  And that's not counting the fact that they negate boreholes' extra ecodamage (giving you something to spend those extra former turns on; the only limit is the fact that the "soft limit" of the "clean minerals" cap is still quite hard, which has its own problems), and give a third-tier boost to psych and economy way before any other third-tier facilities.

[ete]

The way I see it, Forests+tree facilities, Rocky Road Mines, and Advanced Terraforming (Mirrors, Condensors, Enrichers, Boreholes) are all fairly reasonably balanced with each other in terms of former/facility/tech investment and resource return (and all very useful), and Sea improvements have their own set of advantages which makes them interesting (easy to mass Nutrients, plenty of places for useful Tidal harnesses right away unlike Solar which needs high altitude). The only clear imbalance I see is basic terraforming options like Farm, Solar, and non-rocky Mine are near-useless in all but a few niche scenarios. I think trying to bring all the other useful and viable forming options DOWN to the effectiveness level of farm/solar will result in a much less interesting game. I'd suggest either attempting to bring Farm/Solar UP to the level of the others, or accepting that it's not the baseline by which you should measure all other forming options and letting it be largely useless.

[Yitzi]

The way I see it, Forests+tree facilities, Rocky Road Mines, and Advanced Terraforming (Mirrors, Condensors, Enrichers, Boreholes) are all fairly reasonably balanced with each other in terms of former/facility/tech investment and resource return (and all very useful), and Sea improvements have their own set of advantages which makes them interesting (easy to mass Nutrients, plenty of places for useful Tidal harnesses right away unlike Solar which needs high altitude). The only clear imbalance I see is basic terraforming options like Farm, Solar, and non-rocky Mine are near-useless in all but a few niche scenarios. I think trying to bring all the other useful and viable forming options DOWN to the effectiveness level of farm/solar will result in a much less interesting game. I'd suggest either attempting to bring Farm/Solar UP to the level of the others, or accepting that it's not the baseline by which you should measure all other forming options and letting it be largely useless.

The problems with bringing farm/solar up to the level of forests+hybrid forest or advanced terraforming all over the place (which is what you need to balance forests+hybrid forest or many of the better crawler-based strategies) are:
1. It makes power imbalances grow too fast in the early game, putting a heavy focus on early-game strength, and speed over long-term effectiveness.  Some people like to play like that, I do not.  As such, it's certainly worth at least having some way of playing without that.
2. The fact that advanced terraforming creates extra ecodamage suggests that having it all over the place is not what was originally intended; it was meant to be used sparingly, not as much as you could (and to compete with forests, it has to be done a lot.)

So I would dispute your core contention: I think that bringing everything down to a lower level would result in a longer, and therefore more interesting, game, by slowing growth.

[ete]

The problems with bringing farm/solar up to the level of forests+hybrid forest or advanced terraforming all over the place (which is what you need to balance forests+hybrid forest or many of the better crawler-based strategies) are:
1. It makes power imbalances grow too fast in the early game, putting a heavy focus on early-game strength, and speed over long-term effectiveness.  Some people like to play like that, I do not.  As such, it's certainly worth at least having some way of playing without that.

How early game are we talking? Because changes to forest pre-Tree Farm are going to be hard to make, and post tree farm.. I think you should be getting pretty great growth, because you've got to manage a large empire so real-time speed of play drops dramatically (unless you automate everything, which leads to really dumb things) despite an increased growth curve.

If you want to make a "I like to play games with lower growth curves" mod, that's fair enough, but that seems like a very different mission statement from balancing the game, and the former being included almost certainly won't help the uptake of the latter by players, because games are extremely long as it is and there seems like no strong balance reason to extend them.

Also, no matter what you do (within reason) early game advantages and disadvantages are going to grow, because the game is built around exponential growth. This is one of the things that makes SMAX great in my opinion, it lets you actually win games and makes everything count.

2. The fact that advanced terraforming creates extra ecodamage suggests that having it all over the place is not what was originally intended; it was meant to be used sparingly, not as much as you could (and to compete with forests, it has to be done a lot.)

I'm pretty happy with the way Adv forming works. I don't agree that it needs to be used in huge amounts to "compete" with forests, the advanced things can and should simply be phased in replacing old forests/mines as and when former turns become plentiful and the clean mins limit is raised.

So I would dispute your core contention: I think that bringing everything down to a lower level would result in a longer, and therefore more interesting, game, by slowing growth.

hm, with the length of most games on reasonable sized Transcend or from what I've seen in multiplayer, intentionally slowing the game down seems to be a not very good idea. I'm pretty sure the extreme pace of development in late game is a very intended and very useful feature, since it means that even when each turn is taking most of an hour you still get to see some progress and can reasonably end the game.

[Yitzi]

How early game are we talking?

Let's say anything before fusion, though I think that there are parts that shouldn't come into play until the vicinity of quantum power (and I'd favor delaying Transcendi until Threshold.)

Because changes to forest pre-Tree Farm are going to be hard to make, and post tree farm.. I think you should be getting pretty great growth, because you've got to manage a large empire so real-time speed of play drops dramatically (unless you automate everything, which leads to really dumb things) despite an increased growth curve.

The issue isn't real-time speed of play, it's the ability to turn an advantage into a larger advantage.  Keep in mind that Tree Farms aren't that late; beelining let's you get them only 10% of the way up the tech tree, and even if you go "tier by tier", you're only about halfway up the tech tree.  Even fusion is only slightly more than halfway up the tech tree, which is way too early for growth to expand to the point of having such a huge influence on the game.

If you want to make a "I like to play games with lower growth curves" mod, that's fair enough, but that seems like a very different mission statement from balancing the game

It is related to balancing earlier-game strength with later-game strength,

because games are extremely long as it is

Really?  How many MP games do you know of where someone actually completed the tech tree (i.e. reached Threshold of Transcendence)?

Also, no matter what you do (within reason) early game advantages and disadvantages are going to grow, because the game is built around exponential growth.

Not really.  Some things involve exponential growth, but all sorts of features (most notably the tech cost formula) suggest that exponential growth is not what the game is designed for.

This is one of the things that makes SMAX great in my opinion, it lets you actually win games and makes everything count.

Being able to actually win games is accomplished either by endgame offense being stronger than defense (which it is, with Blink and orbital insertions), or a way to win without conquering everyone else (i.e. transcendence.)  What fast exponential growth lets you do is let you actually win games when you're only halfway up the tech tree.

I'm pretty happy with the way Adv forming works. I don't agree that it needs to be used in huge amounts to "compete" with forests, the advanced things can and should simply be phased in replacing old forests/mines as and when former turns become plentiful and the clean mins limit is raised.

You misunderstand: Obviously it doesn't have to be used in huge amounts to compete with forests if the rest is forests.  But having farm/solar with occasional boreholes/mirrors/condensers is not enough to compete with forests (which also might have a few boreholes of their own).

hm, with the length of most games on reasonable sized Transcend or from what I've seen in multiplayer, intentionally slowing the game down seems to be a not very good idea. I'm pretty sure the extreme pace of development in late game is a very intended and very useful feature, since it means that even when each turn is taking most of an hour you still get to see some progress and can reasonably end the game.

Firstly, I'd also like to make the game harder naturally, so that "default" single-player is Librarian (where AI is fairly balanced against human players in terms of advantages, and the AI is cooperative enough to allow diplomacy-based strategies.)  Secondly, if each turn is taking an hour in MP, you've probably got other problems: Loose time controls give 5 seconds per base or active unit and 16 per event, suggesting that even with 30 bases (a very high number; at "cover all the territory inside radii" spacing, you can't fit 30 bases per faction on a standard map), and with 10 active units per base (also very high unless you're in a major war), you're expected to take roughly half an hour per turn on "loose" settings.  (And yes, it'll be a fairly slow game.  TBS games are designed to be somewhat slow)  If you can provide a savegame of where it took an hour per turn, that might help figure out what's going on.
Thirdly, I have nothing against growth picking up in the actual late game (say, around the point where future tech starts to become available); it just shouldn't be so high when the game (as measured by the tech tree) is only half over.

Now, of course a lot of this is my preferences and the rest is simply the way the game was probably designed to be played; if someone prefers the "a builder game ends when the tech tree is only half finished" style, they can of course play that (and then all that's needed is to strengthen farm/solar to only need occasional mirrors/condensers to compete with forests), but I don't think that's how the game was meant to be played, and it certainly doesn't give time to appreciate all the game has to offer (e.g. winning by transcendence).

[ete]

The issue isn't real-time speed of play, it's the ability to turn an advantage into a larger advantage.  Keep in mind that Tree Farms aren't that late; beelining let's you get them only 10% of the way up the tech tree, and even if you go "tier by tier", you're only about halfway up the tech tree.  Even fusion is only slightly more than halfway up the tech tree, which is way too early for growth to expand to the point of having such a huge influence on the game.

If you want to make a "I like to play games with lower growth curves" mod, that's fair enough, but that seems like a very different mission statement from balancing the game

It is related to balancing earlier-game strength with later-game strength,

Yes, smaller advantages lead to bigger advantages.. But I'm far from convinced that AC has that balance wrong.

because games are extremely long as it is

Really?  How many MP games do you know of where someone actually completed the tech tree (i.e. reached Threshold of Transcendence)?

Long as in how long it takes to complete, which is well into the months. And unless one player gives up due to massive powergraph advantage, it would go on a lot longer.

Also, no matter what you do (within reason) early game advantages and disadvantages are going to grow, because the game is built around exponential growth.

Not really.  Some things involve exponential growth, but all sorts of features (most notably the tech cost formula) suggest that exponential growth is not what the game is designed for.

The tech formula cost has to go up a lot, but honestly the factors of increase it uses are so tiny compared to the exponential growth that I would point to it as strong evidence of the game being designed for exponential growth.

This is one of the things that makes SMAX great in my opinion, it lets you actually win games and makes everything count.

Being able to actually win games is accomplished either by endgame offense being stronger than defense (which it is, with Blink and orbital insertions), or a way to win without conquering everyone else (i.e. transcendence.)  What fast exponential growth lets you do is let you actually win games when you're only halfway up the tech tree.

Why should only endgame offense be able to kill people? Surely in a balanced game offense should provide a very real threat throughout the game, especially for many player games where attacking one guy is very likely to put you at a disadvantage so you need good risk/payoff ratios otherwise no one will fight.

I'm pretty happy with the way Adv forming works. I don't agree that it needs to be used in huge amounts to "compete" with forests, the advanced things can and should simply be phased in replacing old forests/mines as and when former turns become plentiful and the clean mins limit is raised.

You misunderstand: Obviously it doesn't have to be used in huge amounts to compete with forests if the rest is forests.  But having farm/solar with occasional boreholes/mirrors/condensers is not enough to compete with forests (which also might have a few boreholes of their own).

ah, yes, i did misunderstand. In that case: both can use adv teraforming, so all that's happening is farm/solar still just sucks.

hm, with the length of most games on reasonable sized Transcend or from what I've seen in multiplayer, intentionally slowing the game down seems to be a not very good idea. I'm pretty sure the extreme pace of development in late game is a very intended and very useful feature, since it means that even when each turn is taking most of an hour you still get to see some progress and can reasonably end the game.

Firstly, I'd also like to make the game harder naturally, so that "default" single-player is Librarian (where AI is fairly balanced against human players in terms of advantages, and the AI is cooperative enough to allow diplomacy-based strategies.)

Making the game any form of challenge for a competent player at librarian will be an insanely huge challenge unless you plan to do it by making the AI cheat a whole lot more at librarian. It is however a very useful project and all players would benefit from smarter AI, not just those who share your feelings on exponential growth.

Secondly, if each turn is taking an hour in MP, you've probably got other problems: Loose time controls give 5 seconds per base or active unit and 16 per event, suggesting that even with 30 bases (a very high number; at "cover all the territory inside radii" spacing, you can't fit 30 bases per faction on a standard map), and with 10 active units per base (also very high unless you're in a major war), you're expected to take roughly half an hour per turn on "loose" settings.  (And yes, it'll be a fairly slow game.  TBS games are designed to be somewhat slow)  If you can provide a savegame of where it took an hour per turn, that might help figure out what's going on.

I said most of an hour (meaning the majority of, >50%, actually closer to 35-45 mins), and was not talking about MP but the lategame turns of a large VS Ai game with broken AI factions. It's due to there being a hell of a lot of bases all churning out something most turns (mostly build order queue facilities), and having to manage a huge army attacking a large AI empire.

Thirdly, I have nothing against growth picking up in the actual late game (say, around the point where future tech starts to become available); it just shouldn't be so high when the game (as measured by the tech tree) is only half over.

By that point games tend to be already decided in SP (possibly MP), and will have been going for many months in MP.

Now, of course a lot of this is my preferences and the rest is simply the way the game was probably designed to be played; if someone prefers the "a builder game ends when the tech tree is only half finished" style, they can of course play that (and then all that's needed is to strengthen farm/solar to only need occasional mirrors/condensers to compete with forests), but I don't think that's how the game was meant to be played, and it certainly doesn't give time to appreciate all the game has to offer (e.g. winning by transcendence).

When a game ends depends on the ability of your foes, evenly matched players will still be able to have long games despite strong exponential growth, and I think that forcing the speed of growth down would simply delay already won games for longer than is enjoyable.

[Earthmichael]

The idea of measuring the progress of a game based on how far we are into the tech tree is interesting, but has nothing to do with reality.

The vast majority of games of a map the size of Planet or the balanced Vets map end long before the  tech tree has been completed.  I can't recall a MP game that I was involved in that even reached Quantum Power.

Technologies can be deadly at any stage.  Typically attack is about double defense as technology grows.  (I know, some of this depends upon the technology paths chosen, but it is in the ballpark.)

The point is, I have seen plenty of games conclude when the strongest attack technologies were missiles, and I have seen games on small maps fight most of the game with pre-missile technology.  So no, you don't need endgame technologies to complete a game.  If your game happens to linger long enough to get Tachyon fields, then I could see this putting the defense at a distinct advantage until Blink.  BUT, this just does not happen; I have yet to play in a MP game where a Tachyon Field was available to anyone.

[Yitzi]

Yes, smaller advantages lead to bigger advantages.. But I'm far from convinced that AC has that balance wrong.

Easy way to tell: How far up the tech tree do games tend to go?  How far should they go? (I feel the answer to the second question is "it should vary some, but often all the way top the top")

Long as in how long it takes to complete, which is well into the months.

With people playing how many hours a week?  IIRC, I found I can complete a game, all the way to the top of the tech tree, on a large map, in maybe 15-30 hours.  More than most multiplayer games, but a really good game takes time to be enjoyed properly.

The tech formula cost has to go up a lot, but honestly the factors of increase it uses are so tiny compared to the exponential growth that I would point to it as strong evidence of the game being designed for exponential growth.

Wait, because it increases slower than growth, that shows that the level of growth was intended?  How do you get that?

Why should only endgame offense be able to kill people? Surely in a balanced game offense should provide a very real threat throughout the game, especially for many player games where attacking one guy is very likely to put you at a disadvantage so you need good risk/payoff ratios otherwise no one will fight.

In a balanced game that isn't ready to end yet, offense will pose a threat if one person puts effort into offense and the other does not put comparable effort into defense.  That way it's a threat forcing action (if he focuses on combat and you don't, you will lose), but will not end the game prematurely (as it would if there's not a substantial gap between "can avoid being killed" and "can kill the other guy".)

As for many-player, there are numerous reasons to fight (for land, for bases, as an incentive to motivate or discourage certain sorts of behavior), provided that you have a combat advantage (probably because you're playing a combat-focused faction).  However, becoming too powerful is limited there by the danger of an alliance to take down someone who's a serious threat.

ah, yes, i did misunderstand. In that case: both can use adv teraforming, so all that's happening is farm/solar still just sucks.

And even farm/solar-with-moderate-advanced-terraforming (say, each solar collector is next to on average 1 mirror, and raising land isn't used much) still sucks as compared to forests.

Making the game any form of challenge for a competent player at librarian will be an insanely huge challenge unless you plan to do it by making the AI cheat a whole lot more at librarian.

No, the AI cheating is exactly the sort of thing I don't want.  My idea is more to nerf some of the strategies that human players use to get ahead of the AI (easy pop booming, raising the clean mineral cap, build-a-shell-and-upgrade being the main ones).

I said most of an hour (meaning the majority of, >50%, actually closer to 35-45 mins), and was not talking about MP but the lategame turns of a large VS Ai game with broken AI factions. It's due to there being a hell of a lot of bases all churning out something most turns (mostly build order queue facilities), and having to manage a huge army attacking a large AI empire.

Well, if they're building queued facilities then they won't need attention every turn.  As for a huge army attacking a large AI empire...why not use a smaller army of stronger troops, or better yet (since you said late-game) blink drop troops (with gravship probe support) going after their less-defended bases?  Not to mention that a lot of the earlier stages (until they actually reach the front lines) can probably be done with go-tos.

By that point games tend to be already decided in SP (possibly MP), and will have been going for many months in MP.

They tend to be already decided by that point because early growth is so fast, exactly what I dislike.  As for having been going for many months...how many hours per player is that?

When a game ends depends on the ability of your foes, evenly matched players will still be able to have long games despite strong exponential growth

I suppose that does make sense.  As long as it's not "ICS if multiplying facilities were cheap" levels of growth; that results in a front-loaded focus that makes the game no fun at all.

Although that still raises the issue...how do you keep the economic boosts from being front-loaded?  If all the economic boosts show up with earlier techs, that makes early teching too important, overly boosting the tech factions and hurting the non-techer factions, and forcing people away from the tech paths that don't carry economic boosts.

The idea of measuring the progress of a game based on how far we are into the tech tree is interesting, but has nothing to do with reality.

That's the reality I'd like to change, at least in an offshoot system.

Technologies can be deadly at any stage.  Typically attack is about double defense as technology grows.  (I know, some of this depends upon the technology paths chosen, but it is in the ballpark.)

Well, except in early stages; before Synthetic Fossil Fuels, it's more like 1-1.5 times the defense (unless one side is playing momentum and goes for impact, and their target doesn't go for HEC to compensate.)

However, once you reach Probability Mechanics, tachyon fields mean a 3X modifier on base defense, so (unless they're just skimping on military techs) you can't really do more than chase their crawlers and formers indoors before they use their attackers (which also get that 2:1 bonus) to kill your units wandering about their territory.  (This goes especially if they've got a road or magtube network that you can't use because it goes through their bases.)

Other than the "after missile but before tachyon fields" range and the special case of early momentum factions, tech really isn't that deadly until the late game (defined by Matter Transmission and Graviton Theory).

BUT, this just does not happen; I have yet to play in a MP game where a Tachyon Field was available to anyone.

That's easy to fix: +1 to "max possible defense" from Advanced Subatomic Theory until tachyon fields become available should work, as you can get Advanced Subatomic Theory before the enemy can build a proper Missile attack force.



To both of you (and anyone else):
I'm starting another topic, to look at various ideas for how much tiles should be worth under various conditions.

[Earthmichael]

As for terraforming, I think all the various options grow with technology.

At low tech, I can:
1. Forest for 1/2/1
2. Farm/solar for 2-3/1/1-4
3. Mine rocky for 0/4/0

Early game, I do find some opportunities to use farm/solar for 3/1/3+, which is 7 FOPS, compared to 4 FOPS for Forest, or 5 FOPS for Forest with tree farm.  Only when I have built a Hybrid forest structure, which is quite expensive, do I get to 7 FOPS for Forest.

But at a similar tech level, I also learn soil enrichers and condensers.  I also learn to raise land and to make Echelon mirrors.  So I can convert many rolling squares into 4/1/5, which is 10 FOPS, much more than a Hybrid forest, and I don't have to build an expensive structure first.

Also, Boreholes can be built at a fairly early technology (only E4), and give 0/6/6, which is 12 FOPS.

So I don't see any reason to get bent out of shape over the 7 FOPS for a Hybrid Forest.