Question: What does the research rate curve look like in your games?

[Yitzi]

I'm considering tinkering with the research rate formula, but to get an idea of what changes would make sense as alternative options I'd like to know what the turns per tech looks like at various stages of the game under the current formula.  So I'm asking the community: How long does each tech take at each stage of the game, and at what difficulty level and what modding (if any) and what factions?

[gwillybj]

Normal research rate without techstag isway too fast for me. I never got to use anything for more than a few turns before it was obsolete, so now I play with techstag on and the research rate in alphax.txt set at 50%. That makes for 20-60 turns per tech depending on the faction's settings in its txt, social engineering, and progression of the game. And plenty of time to play with all the neat toys Sid and Brian gave us.

[Nexii]

I've been modding/playing the following:
-Transcend, Normal sized map
-Mix of HvH and vs AI
-An SE set that is overall more penalizing, but encourages researching SE techs earlier
-Slower terraforming for forest & fungus remove (12T), faster on farm/solar/echelon
-Higher support cost (divisor 4)
-Drone fixes in (mode 31)
-Ecodamage fixes in (generally lenient but 0 CMs)
-Base 7 SMAC factions unmodified
-Cheaper colony pods/formers (20 for inf, 30 for rover), but higher facility costs from -2 IND start
-IndAuto for M-cap lift, Gene for N-cap/Condensor, OptComp for E-cap/Echelon

The amount of overall aggression/war has much more impact than all of these tweaks cumulative.  When you're pressured at war you might not get to put anything at all into labs, else you get wiped out.  In peace you can run peace SEs, plus pacts, plus with global trade pact the economy increase gets rather crazy.  It would seem that these things alone can influence your tech rate by a factor of 10x or more (when considering base maintenance/drone control...labs can be viewed like discretionary income).  I feel in the early game tech may be a bit too slow, since there isn't a good source of land energy before boreholes.  By mid-game the tech rate is somewhat too fast, and probably way too fast after satellites.  Again these depend on the state of war also.  I think you want a fairly fast tech rate though considering Transcend is only 400 turns by default.

Do you have the curve formula for tech research?  I assumed it was based on # of techs you had and the map size only, but I may be off.

[ete]

I'm in the minority here, but I speed up tech rate by 2x-5x in my transcend games unless I'm playing on a custom supermassive map. If I don't the game is over (in a "i am dominant and it's just a matter of clearing up at my own pace" way) before I get to much the fun tech.It's dull to have loads of cool options only when you entirely don't them need to win.

Also, even if the tech per turn rate is really fast late game, the tech per hour spent on AC is often quite a lot lower because lategame turns take MUCH longer if you're being at all aggressive, and even if not it still slows you a whole lot.

[Yitzi]

Normal research rate without techstag isway too fast for me. I never got to use anything for more than a few turns before it was obsolete, so now I play with techstag on and the research rate in alphax.txt set at 50%. That makes for 20-60 turns per tech depending on the faction's settings in its txt, social engineering, and progression of the game. And plenty of time to play with all the neat toys Sid and Brian gave us.

Normal research rate without techstag isway too fast for me. I never got to use anything for more than a few turns before it was obsolete, so now I play with techstag on and the research rate in alphax.txt set at 50%. That makes for 20-60 turns per tech depending on the faction's settings in its txt, social engineering, and progression of the game. And plenty of time to play with all the neat toys Sid and Brian gave us.

I've been modding/playing the following:
-Transcend, Normal sized map
-Mix of HvH and vs AI
-An SE set that is overall more penalizing, but encourages researching SE techs earlier
-Slower terraforming for forest & fungus remove (12T), faster on farm/solar/echelon
-Higher support cost (divisor 4)
-Drone fixes in (mode 31)
-Ecodamage fixes in (generally lenient but 0 CMs)
-Base 7 SMAC factions unmodified
-Cheaper colony pods/formers (20 for inf, 30 for rover), but higher facility costs from -2 IND start
-IndAuto for M-cap lift, Gene for N-cap/Condensor, OptComp for E-cap/Echelon

The amount of overall aggression/war has much more impact than all of these tweaks cumulative.  When you're pressured at war you might not get to put anything at all into labs, else you get wiped out.  In peace you can run peace SEs, plus pacts, plus with global trade pact the economy increase gets rather crazy.  It would seem that these things alone can influence your tech rate by a factor of 10x or more (when considering base maintenance/drone control...labs can be viewed like discretionary income).  I feel in the early game tech may be a bit too slow, since there isn't a good source of land energy before boreholes.  By mid-game the tech rate is somewhat too fast, and probably way too fast after satellites.  Again these depend on the state of war also.  I think you want a fairly fast tech rate though considering Transcend is only 400 turns by default.

Do you have the curve formula for tech research?  I assumed it was based on # of techs you had and the map size only, but I may be off.

I'm in the minority here, but I speed up tech rate by 2x-5x in my transcend games unless I'm playing on a custom supermassive map. If I don't the game is over (in a "i am dominant and it's just a matter of clearing up at my own pace" way) before I get to much the fun tech.It's dull to have loads of cool options only when you entirely don't them need to win.

Also, even if the tech per turn rate is really fast late game, the tech per hour spent on AC is often quite a lot lower because lategame turns take MUCH longer if you're being at all aggressive, and even if not it still slows you a whole lot.

It looks like my question was unclear; I'm interested not so much in how long it takes on average, but rather how it changes and by how much , i.e. how does the early-game rate compare to the mid-game rate and how do both compare to the late-game rate?  (How the hours per turn changes, and what the major contributors are to that change, would also be nice to know, once ete brought it up.)

[Nexii]

Still not too clear.  Are you asking what influences the quickening tech rate (in terms of turns per tech) as the game goes on?  I would say it's mostly the things that give more raw energy:
- lifting energy cap + boreholes (which typically require condensor/TF to support)
- tree farm/hybrid to control eco, and grow bases to hab dome limit of 14
- satellites
- late-game specialists (not that influential until hab dome)

Mostly the last two are what fuels the very late game tech speed.  Of course labs improvements are key...Network Node, Research Hosp, Fusion Lab being more important than the later ones (since each is only 50% labs on the base, and the later ones cost more).

Far as real life time per turn increasing, I mostly fault formers as they occupy 50%-75% of actions on most turns.  Since all land is unimproved, essentially every tile within your empire has to be manually improved to play well (aside from spreading forests).  Clean modifier is a big contributor here.  Early game you might have 10 formers, mid 50+, late 300+.  Even taking away Clean doesnt really solve it...then you just end up doing +3 SUP SEs.  Armies can also take up a lot of time per turn...larger bases support bigger armies.  The bases themselves of course taking most of the rest.  Facilities can be queued up so this isn't quite as bad late. (aside from diplomacy/tech trade/evaluation).  Quicker ways to rush/see mineral count might help some there.

[Geo]

At my end, wartime unit shuffling takes up most of the turn time, followed by terraforming/yield tasks. Then it's 'abandon base checking' if I play a relocation game, with workshop tinkering coming next (especially annoying is to have to retire/obselete the same proposed design for the nth time). Last is energy rushing of production and diplo contacts that are the time consumers. So that's from top to bottom the time I spent/turn.

[Yitzi]

Still not too clear.  Are you asking what influences the quickening tech rate (in terms of turns per tech) as the game goes on?  I would say it's mostly the things that give more raw energy:
- lifting energy cap + boreholes (which typically require condensor/TF to support)
- tree farm/hybrid to control eco, and grow bases to hab dome limit of 14
- satellites
- late-game specialists (not that influential until hab dome)

Mostly the last two are what fuels the very late game tech speed.  Of course labs improvements are key...Network Node, Research Hosp, Fusion Lab being more important than the later ones (since each is only 50% labs on the base, and the later ones cost more).

That much I understood, though it's offset somewhat by increasing tech cost.

My question is how much it increases at various stages (e.g. is it a factor of 5 from early to midgame and 2 from midgame to late game, 2 and 3, 1 and 10, or what?)

Far as real life time per turn increasing, I mostly fault formers as they occupy 50%-75% of actions on most turns.  Since all land is unimproved, essentially every tile within your empire has to be manually improved to play well (aside from spreading forests).  Clean modifier is a big contributor here.  Early game you might have 10 formers, mid 50+, late 300+.  Even taking away Clean doesnt really solve it...then you just end up doing +3 SUP SEs.  Armies can also take up a lot of time per turn...larger bases support bigger armies.  The bases themselves of course taking most of the rest.  Facilities can be queued up so this isn't quite as bad late. (aside from diplomacy/tech trade/evaluation).  Quicker ways to rush/see mineral count might help some there.

I think that changes to unit cost and supply divisor are the way to cut down on army size.

Regarding formers...what are all those formers doing?  Even raising a square by 3 points and building farm/enricher/condenser-or-mirror is only 60 turns, meaning 30 with super formers, so 300+ super formers could fully improve one base every 2 turns, and by the late game there's simply no way you're getting a new base every 2 turns.

For rushing, I don't really think having rushing be that frequent is good for the game anyway; perhaps an across-the-board 50% increase to rush cost would make it less common and therefore take less time (though again, there are only so many facilities to be built in any particular base, and unless you're ICSing you shouldn't have such a ridiculously high number of bases.)

At my end, wartime unit shuffling takes up most of the turn time, followed by terraforming/yield tasks. Then it's 'abandon base checking' if I play a relocation game

How does a relocation game work?

with workshop tinkering coming next (especially annoying is to have to retire/obselete the same proposed design for the nth time).

You do know that automatic design proposals can be disabled, right?

In any case, I'm more interested at the moment in how the turns per tech at various stages of the game compare to each other.

[Nexii]

Gotcha.  I'd say its like ~12 turns/tech on average early, ~6 mid, ~3 late (maybe less even).  So something like a factor of 2 in each era?  It does depend a lot on how much tech is traded also because you could have mid-game tech but only early-game infrastructure (i.e. population, facilities, improved terrain).  This is what is happening in my pacifism game.  Everyone has a lot of E but not much M.

Personally I think rushing is fine as it is one of the more strategic portions of the game (should I rush a big base's multiplier, or save up to rush an SP, or grow horizontally).  Yes SP rushing with energy might feel cheesy but there has to be significant incentives to get a tech first (and some disincentive to trade immediately, this adds some strategy also), or everyone will just mindlessly steal tech and run full ECON build.  Anyways, 3:1 might be a better ratio (for all 3 of units, facilities, SPs).  2:1 isn't bad however IMO.  SPs could use a bit of a cost increase if it were lower.  These would be good variables to be able to set (if possible within alphax.txt.).  I never liked the 10 mineral rule or the unit rush formula how it is because having to manually calculate rush costs adds a lot of real life time and it's not that fun really.  I've been playing with the 10 mineral rule eliminated and it helps some.  I am finding that having the production mineral rollover option on saves a lot of unnecessary micromanagement.

The goal is usually to fully improve a base square every turn for pop boom, so whatever former time that takes.  Late game a large base can make a new base every turn.  Though it isn't always preferable to base spam due to B-drones, and your existing bases are usually better to rush due to higher base E, and less E loss due to EFFIC.  This is why I think a reduced EFFIC SE set is somewhat beneficial.  Anyways, there are incentives to expand late also mainly satellites. 

Fast combat resolution option also helps a lot with the time it takes to fight later wars.  The default makes combat with higher reactors much slower.

[Geo]

At my end, wartime unit shuffling takes up most of the turn time, followed by terraforming/yield tasks. Then it's 'abandon base checking' if I play a relocation game

How does a relocation game work?

Rushing (sea)colony pods in conquered bases and abandonning them in the end.
IOW, very movement intensive.

[Yitzi]

Gotcha.  I'd say its like ~12 turns/tech on average early, ~6 mid, ~3 late (maybe less even).  So something like a factor of 2 in each era?  It does depend a lot on how much tech is traded also because you could have mid-game tech but only early-game infrastructure (i.e. population, facilities, improved terrain).  This is what is happening in my pacifism game.  Everyone has a lot of E but not much M.

So about a factor of 4 from beginning to end (by late, do you mean all the way up to transcendence?), and about evenly decreasing throughout?

Personally I think rushing is fine as it is one of the more strategic portions of the game (should I rush a big base's multiplier, or save up to rush an SP, or grow horizontally).

I agree that rushing is fine; having rushing be so common that it's a significant contributor to turn length, though, isn't.  The strategic element would probably be enhanced if you could only afford to rush once every few turns, as then you really have to pick it carefully.

Yes SP rushing with energy might feel cheesy but there has to be significant incentives to get a tech first (and some disincentive to trade immediately, this adds some strategy also), or everyone will just mindlessly steal tech and run full ECON build.

I figure that it's so expensive that it's ok.  Although I think the answer to discouraging techsteal+full ECON is more one of making it so expensive to steal tech from outlying bases that unless you've got a significant research disadvantage or want a tech right now, you're probably better off just teching yourself.

Anyways, 3:1 might be a better ratio (for all 3 of units, facilities, SPs).

The idea of making SPs and units more expensive than facilities are a good one, though.

These would be good variables to be able to set (if possible within alphax.txt.).  I never liked the 10 mineral rule or the unit rush formula how it is because having to manually calculate rush costs adds a lot of real life time and it's not that fun really.  I've been playing with the 10 mineral rule eliminated and it helps some.  I am finding that having the production mineral rollover option on saves a lot of unnecessary micromanagement.

Well, you can just apply an across-the-board percentage change by giving everyone a hurry bonus or penalty.

And with high production rollover there's no real reason to partially rush a lot of units; you might as well just do it fully and then your production will give you a leg up on the next one.

The goal is usually to fully improve a base square every turn for pop boom, so whatever former time that takes.

Even so, your older bases should be fully improved fairly quickly at that rate, so you should only be improving a few at a time...

Late game a large base can make a new base every turn.  Though it isn't always preferable to base spam due to B-drones, and your existing bases are usually better to rush due to higher base E, and less E loss due to EFFIC.  This is why I think a reduced EFFIC SE set is somewhat beneficial.  Anyways, there are incentives to expand late also mainly satellites.

Ah, just crowding bases on the theory that you can get population, and thus satellite production, faster with more pop booming bases.

Personally, I think it would be a more interesting game if satellites were harder to put up, easier to knock down, and booming were substantially harder (maybe even impossible pre-Eudaimonia).

Fast combat resolution option also helps a lot with the time it takes to fight later wars.  The default makes combat with higher reactors much slower.

Well, I suppose that's what it's for.

Rushing (sea)colony pods in conquered bases and abandonning them in the end.
IOW, very movement intensive.

And you do it in order to avoid the conquered base penalty?

[Geo]

And you do it in order to avoid the conquered base penalty?

Partly, yes. Other parts are to lift certain home bases over the pop cap, and faction 'philopsophy'.
IIRC, a way to avoid the penalty you mentioned is to let a conquered base be taken by yet another faction and then retaking it again.

[Nexii]

I would guess around a factor of 4 - not evenly decreasing per se.  I think the biggest jumps are at borehole and satellites.  Keep in mind a lot of what we saw before with mid and late being very fast is because the game is effectively won against the AI. 

It's mostly units that I have more an issue with how the rush cost works - because it's non-linear that adds a lot of micromanagement.  Although due to the higher cost I feel only facilities are usually worth rushing.  A better player may be able to comment - if this is best or if going full labs is superior play. Anywhere from 2:1 to 4:1 on E to M is fairly balanced on other things.  Keeping it 2:1 on facilities and 4:1 on SPs, and making it 4:1 on units would be decent.  I'd probably play around with that.  Probably doesn't have a huge impact since most games the majority of E goes into labs not econ.  The key is keeping the incentive to tech, usually this was for SPs more than facilities due to the denial aspect.  Before you would pre-build crawlers, but if that's out then it's just energy rushing.  Which I don't have such an issue with because you can know the cost of an SP and how much energy you are making a turn.  That's a fairly basic calculation - at least to me.

I'm still playing through a trial game.  Right now the factions are all 5-15 turns for discoveries in the early-mid game.  Amplified a lot by pacts, but infrastructure is relatively poor.  It will take some time and granted I'm not an expert player.  One curious thing about SMAC is just how much strategy there is in the game.  I'm starting to feel that perhaps very high tech steal costs won't be ideal.  The steal cost should scale with the lab cost that it would take to research.  Early game a probe team costs a lot, later game it doesn't.  That is, early on even the mineral cost makes it questionable to steal.  Trading of course is always ideal but if you fall behind, there needs to be a catchup mechanism (which is pacts and stealing).  Governor only helps the #2 player catch the #1 player.

Satellite caps which get raised by late-game techs might work better, though that does encourage more smaller bases. 

I tried some games without booming, thinking it would be more interesting also, but it kind of wasn't.  It just ended up slowing down the game too much, everyone's bases stagnated around size 6-7.  Takes a lot of former time to make condensers to grow fast past that - and then you're better off just making more and more bases with PS.  Booming is what makes building vertically more worth it than horizontally.  Now you could eliminate booming but it would require a more gradual N cost curve.  Each worker costs more N, which means more and more former time.  I do agree this could reduce micromanagement some if done right since I know really good players will go through every base just to make sure it's +2N.

Older bases tend to be booming a lot in the early and mid game, so they require a lot of former time to stay improved.  If you haven't improved the terrain then the boom actually isn't going to help you much.  With Forests at 4 former-turns that means you need 4 formers per base to improve a tile per turn (2 if Super).  Much more than this however to do advanced terraforming.  Granted once bases hit their pop cap, then yes that means the formers can go elsewhere.  50/300 may have been a bit of an exaggeration but I find it's not too far off.  Reducing former time on everything may help with the massive number of formers required.  But at the same time, it may speed the game up too much - because advanced terraforming is so powerful.  May be worth trying the game with Condensor, Borehole, and TF/HF slightly weakened.  They have to be good enough to pursue though, so it's tricky. 

[Yitzi]

And you do it in order to avoid the conquered base penalty?

Partly, yes. Other parts are to lift certain home bases over the pop cap

Wait, you can use colony pods to exceed the hab facility caps?  That seems to me like a sort of exploit...

and faction 'philopsophy'.

Meaning?

I would guess around a factor of 4 - not evenly decreasing per se.  I think the biggest jumps are at borehole and satellites.  Keep in mind a lot of what we saw before with mid and late being very fast is because the game is effectively won against the AI. 

But there isn't a big jump at transcendi?  Or is that after you've already won?

It's mostly units that I have more an issue with how the rush cost works - because it's non-linear that adds a lot of micromanagement.

If you just rush fully, it doesn't add micromanagement.  You do lose any production over the carry-over cap that way, but fortunately the carry-over cap can be increased.

Although due to the higher cost I feel only facilities are usually worth rushing.  A better player may be able to comment - if this is best or if going full labs is superior play.

I'd guess it depends on the situation.

The key is keeping the incentive to tech, usually this was for SPs more than facilities due to the denial aspect.  Before you would pre-build crawlers, but if that's out then it's just energy rushing.

But that's expensive, as it should be.  I think the answer is to make techsteal difficult enough that getting the tech first equates to a potentially substantial head start on the project, which can be overcome by a similarly large production advantage but will give you the project if your tech advantage outweighs your production disadvantage.  (This will, among other things, make the midgame somewhat less energy-focused and more mineral-focused, i.e. more balanced.)

I'm still playing through a trial game.  Right now the factions are all 5-15 turns for discoveries in the early-mid game.  Amplified a lot by pacts, but infrastructure is relatively poor.  It will take some time and granted I'm not an expert player.  One curious thing about SMAC is just how much strategy there is in the game.  I'm starting to feel that perhaps very high tech steal costs won't be ideal.  The steal cost should scale with the lab cost that it would take to research.  Early game a probe team costs a lot, later game it doesn't.  That is, early on even the mineral cost makes it questionable to steal.  Trading of course is always ideal but if you fall behind, there needs to be a catchup mechanism (which is pacts and stealing).  Governor only helps the #2 player catch the #1 player.

I was planning on having it proportional to the target's energy...in retrospect, that might not be the best method.  I'll have to think about how to do it; perhaps it could depend on how many techs the target knows that you don't?

Satellite caps which get raised by late-game techs might work better, though that does encourage more smaller bases.

I wasn't thinking "caps" so much as "require late-game tech for sky hydroponics lab and make it and Nessus mining station more expensive, and earlier and cheaper orbital defense pods"; the idea would be that you can build a pod, use it to try to shoot down an enemy production satellite, and on average come out ahead.

I tried some games without booming, thinking it would be more interesting also, but it kind of wasn't.  It just ended up slowing down the game too much, everyone's bases stagnated around size 6-7.  Takes a lot of former time to make condensers to grow fast past that - and then you're better off just making more and more bases with PS.  Booming is what makes building vertically more worth it than horizontally.  Now you could eliminate booming but it would require a more gradual N cost curve.  Each worker costs more N, which means more and more former time.  I do agree this could reduce micromanagement some if done right since I know really good players will go through every base just to make sure it's +2N.

What I was thinking is increase the level at which booms happen, but +6 and higher nutrients (until you boom) give the same -1 nutrient/row each as lower ones do...so at +6 or higher GROWTH you'd still grow pretty quickly if you focus on nutrients.

Of course, if you aren't so nutrient-focused, you won't grow as fast...but I don't think having a few turns per growth would be such a bad thing.

If you haven't improved the terrain then the boom actually isn't going to help you much.  With Forests at 4 former-turns that means you need 4 formers per base to improve a tile per turn (2 if Super).  Much more than this however to do advanced terraforming.  Granted once bases hit their pop cap, then yes that means the formers can go elsewhere.  50/300 may have been a bit of an exaggeration but I find it's not too far off.  Reducing former time on everything may help with the massive number of formers required.  But at the same time, it may speed the game up too much - because advanced terraforming is so powerful.  May be worth trying the game with Condensor, Borehole, and TF/HF slightly weakened.  They have to be good enough to pursue though, so it's tricky.

Maybe I'd better ask: Around what turn are you hitting around 100 formers, and how many bases do you have at the time (and of what sizes)?

[Geo]

Wait, you can use colony pods to exceed the hab facility caps?  That seems to me like a sort of exploit...

At least in the SMAC game I'm now playing. See for yourself for the 'disappearing formers' thread I just started. There's a save attached and you have transports filled with colony pods to join other bases at your disposal.

and faction 'philopsophy'.

Meaning?
[/quote]

As a say democratic and enlightened leader I don't spare expenses to move recently conquered subjects from the slums and ruins of their former homes to the splendor of my faction's home bases. Just fluff thus.