Author Topic: SMAX - The Will to Power - mod  (Read 22632 times)

0 Members and 2 Guests are viewing this topic.

Re: SMAX - The Will to Power - mod
« Reply #825 on: August 29, 2020, 09:50:06 PM »
Quote
I have a wild idea. What if we pursue the idea of free citizens traveling across the empire at any time? Not necessarily at the base creation time. For example, let disgruntles citizen those otherwise turn into drones travel to better places in other bases those have excess of happiness and where extra citizen does not become a drone yet?
It would be inventing a mechanic for the sake of a single project which only a single faction would be able to build.

Plus it runs into the problem of the contribution mechanics in that a base should be a self-contained, independent entity. All effects should be evident from looking at the Psych tab and should not depend on situations in other bases.

Yeah. I feel the same already. Too complicated. 🤔
I guess, we'll go with your average base size approach for now.

A very intriguing idea, if you mean a mechanic that lets the player manually redistribute his citizens across his bases. You've reached your population cap? Send excess citizens to smaller bases. Nutrient failure? Evacuate half the base. Captured enemy territory? Move the inhabitants back to your own empire.

In scenario editor, you can manipulate population with - or +. Imagine, after you build PTS, you select a base window and press either - or + (doesn't matter which.) Another window opens, showing all of your bases and a field to enter a number. If you enter the number 3 and select Gaia's Landing, 3 citizens will be sent from whatever base you were looking at to Gaia's Landing.

Might add a check to keep this from breaking population caps. Might not.

AI would have a hard time handling this. Perhaps it could have a different effect for them, closer to what Nevill has in mind.

Yep. I was excited by this idea at first but it is completely new mechanics that would require tons of time to program it. 😒
And we also need to teach AI that, which is even more time. Not worth it.
Tim

Re: SMAX - The Will to Power - mod
« Reply #826 on: August 29, 2020, 10:10:16 PM »
Reworked my SE weights article. Made it more concise and to the point. Also added some contemporary WTP examples.
The formulas and numbers I came up with are disputable but I found them a good first iteration estimate to evaluate OP/UP SE models.

http://alphacentauri2.info/wiki/Social_Engineering_Mod
Tim

Offline Tayta Malikai

Re: SMAX - The Will to Power - mod
« Reply #827 on: August 29, 2020, 10:35:14 PM »
Oh huh, that was you who wrote that? I quite enjoyed it the first time round. I even tried using the suggested changes themselves, although they didn't really gel with me at the time.

Might be good to include the before and after SE configuration lines, so we can see how the numbers correspond to the changes, if you aren't already doing that.

Offline MercantileInterest

Re: SMAX - The Will to Power - mod
« Reply #828 on: August 29, 2020, 10:43:09 PM »
New tech cost function gives interesting choices. Very quick to research old techs you skipped on the way up but you might prefer to spend the time working towards more advanced abilities.

Noticed one of the AI's formers leveling a rocky site with a mineral resource instead of mining it. Strange behavior.

Re: SMAX - The Will to Power - mod
« Reply #829 on: August 30, 2020, 03:41:25 AM »
Oh huh, that was you who wrote that? I quite enjoyed it the first time round. I even tried using the suggested changes themselves, although they didn't really gel with me at the time.

Glad that you've enjoyed it.
That was long ago when text modding is only what I could work with. I found some of my initial assumptions quite questionable. However, I have stopped working on it completely as I didn't believe anyone was reading it at all.
😢

Might be good to include the before and after SE configuration lines, so we can see how the numbers correspond to the changes, if you aren't already doing that.

Well, I've included vanilla and v. 117. So they are essentially an absolute before and an absolute after. Anything else you would love to see there?
« Last Edit: August 30, 2020, 04:20:21 AM by Alpha Centauri Bear »
Tim

Re: SMAX - The Will to Power - mod
« Reply #830 on: August 30, 2020, 04:08:40 AM »
Actually, when we are on the topic you, guys, may help me with this a bit. I found that most important SE effects can be reduced to either mineral intake or energy intake. I do not worry about PROBE or non combat PLANET benefits. Their weights can be adjusted by eyeballing and playing experience.
The conversion between energy and minerals - that is what still bothering me. Even though it seems that I have established energy reserves to minerals conversion ratio it doesn't correctly describe relation between RESEARCH and INDUSTRY. With time I hardened my view on this that all major effects can be reduced to either RESEARCH and INDUSTRY (or combination of them). However it is quite difficult to devise clear relation between these two. I don't mean a direct ratio like 1 RESEARCH = 0.5 INDUSTRY. As you can see in the article, all my formulas include a lot of other factors. And that is completely fine. Effect values varies greatly with various factors. The purpose of this exercise is to understand how the relation between different effects vary based on these other parameters.

So the big question here is what do you think relation between RESEARCH and INDUSTRY? Feel free to throw different ideas.
Tim

Re: SMAX - The Will to Power - mod
« Reply #831 on: August 30, 2020, 04:56:28 AM »
Just found this amusing.
😁
Tim

Re: SMAX - The Will to Power - mod
« Reply #832 on: August 30, 2020, 07:56:42 PM »
# Version 119

* [bug] Tile yield calculation didn't set base. That caused crash for tile yield computation with collector on it and adjacent mirror when base is not set sometimes. Strangely it didn't bite me earlier.

I just caught this today. Before it didn't bother me at all. So, apparently, this but manifests itself quite rarely and you may not even experience it once. It happens during AI turn. If it does - use this version for fix.
Tim

Offline MercantileInterest

Re: SMAX - The Will to Power - mod
« Reply #833 on: August 31, 2020, 03:45:39 AM »
AI likes to build land formers in sea bases. You having the same issue or did I somehow cause it by tinkering with the txt?

Sent locusts to bomb Lal's boreholes. He never attacks my locusts with his ground units but when the AI sends locusts to attack me, I can attack them with my ground units.

Re: SMAX - The Will to Power - mod
« Reply #834 on: August 31, 2020, 03:49:15 AM »
AI likes to build land formers in sea bases. You having the same issue or did I somehow cause it by tinkering with the txt?

I didn't play with production priorities. So - either vanilla or Thinker.

Sent locusts to bomb Lal's boreholes. He never attacks my locusts with his ground units but when the AI sends locusts to attack me, I can attack them with my ground units.

Locust uses gravship chassis so it is attackable by ground units. The fact that AI didn't attack yours is just poor AI programming as usual. It's pretty dumb. Even in Thinker it is just less dumber than in vanilla.
Tim

Offline lolada

Re: SMAX - The Will to Power - mod
« Reply #835 on: August 31, 2020, 08:52:26 AM »
Actually, when we are on the topic you, guys, may help me with this a bit. I found that most important SE effects can be reduced to either mineral intake or energy intake. I do not worry about PROBE or non combat PLANET benefits. Their weights can be adjusted by eyeballing and playing experience.
The conversion between energy and minerals - that is what still bothering me. Even though it seems that I have established energy reserves to minerals conversion ratio it doesn't correctly describe relation between RESEARCH and INDUSTRY. With time I hardened my view on this that all major effects can be reduced to either RESEARCH and INDUSTRY (or combination of them). However it is quite difficult to devise clear relation between these two. I don't mean a direct ratio like 1 RESEARCH = 0.5 INDUSTRY. As you can see in the article, all my formulas include a lot of other factors. And that is completely fine. Effect values varies greatly with various factors. The purpose of this exercise is to understand how the relation between different effects vary based on these other parameters.

So the big question here is what do you think relation between RESEARCH and INDUSTRY? Feel free to throw different ideas.

I find that in practice Industry > Research and its heavily skewed. Research is important up to a point, you have to get those key techs.. for example resource unlock, advanced terraforming, probe teams for example. But once you get that many techs are actually not important and unit/facilities spam takes over to win. You can research a lot but not have industry to build those expensive units and get overrun by military industrious AIs like Hive.

Another huge thing is Probe teams - AI builds tons of them, so does the player - and tons of techs get stolen. So factions with weaker research have excellent catch up mechanism. There's no good industry catch up mechanism.

One observation is that industry AIs are doing excellent (Hive, Drones) while research ones like University tend to get first to die. There are other factors, but in general it seems to work this way. One can argue that Zakharov does not die due to Research bonus.. but look at it this way - his bonus is not good enough to make up for his weaknesses. Meanwhile industry AIs are usually able to get enough techs and spam tons of secret projects, units and facilities to overpower other factions. Santiago for example tends to perform badly crippled by -Industry. Miriam with -Research is often very strong AI.

Player can shift the trend and use research quite well - most new players really like to play University. Still the easiest game i had are with factions like Hive and Drones they are just great.

I would not underestimate Probe and Planet rating they matter a lot especially for AI. Player can ignore these and with good play play around the minuses, but AI gets ruined. Worms often ruin -planet factions and Probe teams are significant investment. Its not just random worms around - which can deal huge damage in right circumstances - but also later fungus blooms and worm spawns which ruin terrain and then kill defenders and base facilities.
Having high probe rating really helps and negative is really really bad. One of the reasons why University is so pathetic in the mod is due to AI probe team spam.. Uni is just easy to subvert and steal tech from.


Offline lolada

Re: SMAX - The Will to Power - mod
« Reply #836 on: August 31, 2020, 09:30:37 AM »
Quote
Reworked my SE weights article. Made it more concise and to the point. Also added some contemporary WTP examples.
The formulas and numbers I came up with are disputable but I found them a good first iteration estimate to evaluate OP/UP SE models.

http://alphacentauri2.info/wiki/Social_Engineering_Mod


Great work  ;b; this is really impressive. I took somewhat detailed look in prior version as well. Its huge amount of information and its all very complicated so its hard comment generally. We could discuss interesting parts. I am a bit currently out of a loop with new SE changes (you changed some things recently). Gonna comment a few things I found interesting last time i looked at this.

SUPPORT AND POLICE rating

This pairing is interesting. Support is really good stat early to mid game and since its non-linear its potentially OP if stacked to max and played in military playstyle. What striked me about pair is that Police is actually quite useless without enough Support. Police requires investment to build units and thats already asking for a lot. Then if you don't have support its -1 mineral flat drain and thats a big no no. Higher police means -2, -3  mineral drain and thats even worse.

So I think +Police rating is kind of crap if not combined with enough +Support. One could try to combine it, but then there's opportunity cost of not picking something else. Support on the other hand on its own is really good and as you pointed out quite a few times, its better than Industry for quite a while.

Growth
I think you had 2.0 rating all the way in prior article ? Not sure, its now rated 2.0 1.4 0.7 over time.. i guess in general that's good, but it bugged me earlier. I would emphasize that ratings could be judged differently for player and AI. So what's the deal.

For one, AI can grow much more than player on Transcend (drones issues).. so player does not need to rate growth highly for long period of game. Its actully a bit binary for the player if optimized..  Secure room for the growth (for example build Holo Theaters everywhere.. grow quickly.. turn back SE to something else).

Same goes for Probe and Planet ratings. AI really needs these to be neutral or positive or its very bad for them. Player can handle it.

Some random comments

Quote
Apparently, Fundamentalist is a complete waste of a SE slot, which is also corresponds to community opinion.

- In general, but its great for early rush/war as temporary pick. Can be game decider.

Quote
Power seems to be under-powered as well especially later in the game when SUPPORT value deteriorates.


Not sure if you changed it again after you buffed it with change to -1 Industry (from -2), but I remember it was quite good at that point. Having very high morale units is no joke, its really important in WTP. So once you have strong economy and are ready for war Power is excellent pick up.




« Last Edit: August 31, 2020, 09:59:21 AM by lolada »

Re: SMAX - The Will to Power - mod
« Reply #837 on: August 31, 2020, 08:24:37 PM »
I find that in practice Industry > Research and its heavily skewed. Research is important up to a point, you have to get those key techs.. for example resource unlock, advanced terraforming, probe teams for example. But once you get that many techs are actually not important and unit/facilities spam takes over to win. You can research a lot but not have industry to build those expensive units and get overrun by military industrious AIs like Hive.

Another huge thing is Probe teams - AI builds tons of them, so does the player - and tons of techs get stolen. So factions with weaker research have excellent catch up mechanism. There's no good industry catch up mechanism.

One observation is that industry AIs are doing excellent (Hive, Drones) while research ones like University tend to get first to die. There are other factors, but in general it seems to work this way. One can argue that Zakharov does not die due to Research bonus.. but look at it this way - his bonus is not good enough to make up for his weaknesses. Meanwhile industry AIs are usually able to get enough techs and spam tons of secret projects, units and facilities to overpower other factions. Santiago for example tends to perform badly crippled by -Industry. Miriam with -Research is often very strong AI.

That is absolutely my impression too. RESEARCH is no match for INDUSTRY. I usually tend to approximately evaluate it as RESEARCH = 1/2 of INDUSTRY. Different evaluation techniques gave me a range of somewhere 0.3-0.7 for RESEARCH/INDUSTRY. As for other factor affecting this I also agree with you that it may become slightly less important as game progress. Especially because INDUSTRY is a completely proportional factor: 10% increase in it results in 10% more stuff built over the course of the game. Whereas 10% increase in RESEARCH doesn't give 10% more tech over other factions, even without stealing, due to the ever-growing tech cost. The player will be just 4-6 techs ahead all the time.

I would not underestimate Probe and Planet rating they matter a lot especially for AI. Player can ignore these and with good play play around the minuses, but AI gets ruined. Worms often ruin -planet factions and Probe teams are significant investment. Its not just random worms around - which can deal huge damage in right circumstances - but also later fungus blooms and worm spawns which ruin terrain and then kill defenders and base facilities.
Having high probe rating really helps and negative is really really bad. One of the reasons why University is so pathetic in the mod is due to AI probe team spam.. Uni is just easy to subvert and steal tech from.

Oh, no. I don't underestimate them. I am just saying it is hard to compare them to some economic effects like INDUSTRY but I don't worry about them as they have direct applications and can be evaluated from the "feeling" of it.
Tim

Re: SMAX - The Will to Power - mod
« Reply #838 on: August 31, 2020, 08:47:27 PM »
Growth
I think you had 2.0 rating all the way in prior article ? Not sure, its now rated 2.0 1.4 0.7 over time.. i guess in general that's good, but it bugged me earlier. I would emphasize that ratings could be judged differently for player and AI. So what's the deal.

For one, AI can grow much more than player on Transcend (drones issues).. so player does not need to rate growth highly for long period of game. Its actully a bit binary for the player if optimized..  Secure room for the growth (for example build Holo Theaters everywhere.. grow quickly.. turn back SE to something else).

It's difficult to evaluate precisely as it is a far future investment. All other effects are immediate. Yes, its weight definitely varies depending on evaluation method and point of view. If was usually higher than 1, though. Don't remember ever valuing it 0.7. Probably a mistake. I'd say 1.5-2.5 somewhere. Again, very depending on game stage. Quite useless at the end (except to buff the score).

My evaluation have nothing to do with their usage strategies. Just with overall bonus "fairness" they bring with them. Obviously, time and game state dictate the choice.

Quote
Apparently, Fundamentalist is a complete waste of a SE slot, which is also corresponds to community opinion.
- In general, but its great for early rush/war as temporary pick. Can be game decider.

Agree. But again, I am not saying it won't ever be picked. I just wanted to increase a proportion of time it picked to match other choices.

Quote
Power seems to be under-powered as well especially later in the game when SUPPORT value deteriorates.

Not sure if you changed it again after you buffed it with change to -1 Industry (from -2), but I remember it was quite good at that point. Having very high morale units is no joke, its really important in WTP. So once you have strong economy and are ready for war Power is excellent pick up.

That was vanilla analysis. In WTP it has one less INDUSTRY penalty. That seemed to do the trick.
Tim

Re: SMAX - The Will to Power - mod
« Reply #839 on: September 01, 2020, 03:56:16 AM »
# Version 121

* SE reworked a little.\
http://alphacentauri2.info/wiki/Social_Engineering_Mod

Inspired by recent conversations and feedback.
Tim

 

* 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)
-=-
19 (7%)
XP Compatibility patch
-=-
9 (3%)
Gog version for Windows
-=-
78 (29%)
Scient (unofficial) patch
-=-
29 (10%)
Kyrub's latest patch
-=-
15 (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
-=-
13 (4%)
Total Members Voted: 265
AC2 Wiki Logo

* Random quote

Those who join us need only give up only half of their humanity?the illogical, ill-tempered, and disordered half, commonly thought of as 'right-brain' functioning. In exchange, the 'left-brain' capacities are increased to undreamed potentials. The tendency of Biologicals to cling instead to their individual personalities can only be attributed to archaic evolutionary tendencies.
~Prime Function Aki Zeta-5 'Convergence'

* Select your theme

*