SMAX - The Will to Power - mod

[Alpha Centauri Bear]

1) Have cost fixed and level based like in thinker mod and have the cost of each level match development speed to extend the research beyond 300 turns

The rest of the game engine is not aware of fixed cost. It will be very difficult to integrate. This'll create more problems than it solves.

2) Make AI agree to trade only equal level techs as baseline, +- 1lvl depending on diplomacy standing, +- 1lvl depending on how much more or less traded techs are weighted according to AI weights.

Don't see a need. Why restrict it? AI is weighing trade benefit already and may reject worse offer. Level is not guarantee of need. You may desperately need level 1 tech and willing to give away level 5 for it and vice versa.

3) If trade would be not possible according to the above rule, AI should trade for money according to tech cost, adjusted by diplomacy, commerce and AI tech weights.

Even current rude rule of offering tech for 100 credits regardless of level does not have any devastating effect on game flow. Why bother with even smaller details?

4) Make probe steal tech action have a cost like base mind control, it should cost half the cost of research as baseline, modified by probe rating like mind control.

It does have cost already - vendetta! It's not monetary but not everything should be monetary in the game.

[Alpha Centauri Bear]

Researched tech cost some more. It is quite convoluted. I don't think I need to change it entirely as it doesn't break game strategically or something. If I am going to change it then at most I'll make it doesn't grow faster than 10% and change multiplier from map size square root to just map size.

[dino]

I think weapon and armor progression in your mod is not a good idea:
W: 1,2,3,4,5,6,8,10,13,16,20,24,30,
A:  1,2,  ,3,  ,4,5, 6 , 10,16,24,   ,    ,

Shortly after Tachyon Field introduction, weapon rating equals armor.
With 1/3 combat round chance, the chance of defeating a defender in a base with a tachyon field is very small. AI won't be capable of conquering any bases imo.

I'd go with something like this:
W: 1,2,3,4,5,6,8,10,12,16,20,24,30,
A:  1,2,  , 3,  ,4,6, 8, 10,12,16,   ,    ,

The combat resolution function is fully disassembled by Scient, so very easily two binary tweaks could be done ( or better yet pull the whole combat resolution function into the mod ):
1) Reduce perimeter defense and tachyon field bonuses to +75%
2) Remove stack collateral damage, from defeating a unit

Results ?
Defending a base would have roughly same chances as vanilla, but easily knocking down with a small task force and little to no loses, a continous stream of AI units approaching without any tactics your bases would become no longer possible. I'm positive that AI would benefit from these changes a lot, while still giving a player some room to gain an advantage with clever tactics.

[Alpha Centauri Bear]

Do you suggest 30:16 end game W:A ratio *together* with weakening base defenses? When you say it helps AI - do you mean it helps *attacking* AI or both AI sides? If attacking only then you just tilt the balance in favor of more aggressive AI factions. This is an exact thing I am trying to  neutralize with new weapon and armor design.
I don't see much problem for more economically advanced AI faction to fight less advanced one. You are right that TF triples defense. That just makes attacker lose triple number of units more against same strength defenders in bases. Assuming they can supply three times more units to the front line they will advance without problems. And this is only the worst scenario for attacker. Outnumbering opponent means ruling the field between bases impeding enemy communications, and achieving artillery superiority. With artillery on their side, attacker halves base defenders HPs changing initial 1:3 odds to 2:3. And then with discovery of strongest weapon to 10:12. It's almost equal losses! Now you want to weaken defenders even more. If we do this we are back to attacker advantage and indestructible army nightmare.

I agree that all this math is merely an estimation and only play testing gives real picture. What I've observed so far is that even same strength defenders behind PD do not stop dedicated invader AI deploying enough units to battle. Thankfully that doesn't happen swiftly anymore as in vanilla when AI with slightest advantage ate neighbors for breakfast and grew out of proportions in a blink of an eye. In this mod I still see AI advancing occasionally but at slow enough rate when snowballing can be stopped.

Please feel free to share your play testing results. I'm also happy to provide modified version of the mod for you to test your theories. I am quite keep of such research too.

[dino]

If you have 1/3 *combat round* chance, then the chance of the whole combat is much smaller than that, since after each damage taken the odds tilt even more in favor of the defender.
I assumed the odds of each round are adjusted for damage taken in the previous one. It'd 1/3 only if odds were always the same as in the first combat resolution round.
If that is the case with the combat function, please let me know, it's years since I've looked into it, I may misremember it.

The goal of the design presented is to make actively defending against AI ( by attacking incoming forces ) more difficult, without changing AI vs AI dynamics much.
Keep in mind that AI vs AI defender can replace/repair looses much faster and with thinker most factions remain of roughly equal power.
In vanilla AI mid game can conquer anything, *mostly* only after it gains a reactor advantage.
If you take reactor advantage out *and* tilt defenders odds so heavily as you did, in my opinion AI won't be capable of conquering anything.

Does AI use artillery to purposefully drop defenses to 50% ? With Tachyon field odds are already more in favor of the defender than in the early game.
If you calculate the odds, my proposal doesn't change base conquering odds at all from vanilla, only makes aggressor faction units take less damage in the open, from human defender attacks.
In my experience AIs with thinker are in permanent stalemate, I'd rather like to see more decisive actions between them, not less.

I'd love to have combat resolution function disassembled and it's source code as a part of the mod to tweak.
Or, you could just make 4 versions with binary tweaks: with/without collateral damage from loosing a unit in the stack and with 50%, or 75% facility bonuses.
With access to such tweaks I'd be interested in doing playtesting and experiments.
Some tweaks to the existing source code is the extend of my ability, if it comes to coding, so I can't make it happen myself

[Alpha Centauri Bear]

If you have 1/3 *combat round* chance, then the chance of the whole combat is much smaller than that, since after each damage taken the odds tilt even more in favor of the defender.

You are absolutely right. The chance of a 1/3 weaker unit win a singe duel is zero. However, we are talking about an end game assault saturated with units and multiple attacks against a base per turn. It that scenario we care less about individual unit winning chances but more of mutual HP depletion. Each defender will be hit by multiple attacker. Some of them may die but they eventually kill a defender. Let's say AI 1 have 4 defenders behind TF and AI 2 attacks a base with 20+ attackers of about same strength as defenders all at the same turn. As a result AI 1 will lose all 4 defenders when AI 2 will lose 12 attackers on average. That is all war math we need: proportion of losses matches combat odds.
This being said the actual SMACX combat odds are different from those displayed! But this is another story and potential fix target.

Back to the discussion. Battle win chances matter only when lone units meat in wilderness. In large scale war where each side can sacrifice as many units in a single turn to defeat target unit only proportion of losses matters. In this regard slight increase in base defense does not negate invasion completely but only how little the invader can bite with the size of their current forces. I never tasked myself with fine tuning this. Feel free to do that research and I'll listen to your suggestions.

[Alpha Centauri Bear]

I assumed the odds of each round are adjusted for damage taken in the previous one. It'd 1/3 only if odds were always the same as in the first combat resolution round.
If that is the case with the combat function, please let me know, it's years since I've looked into it, I may misremember it.

That is correct. Odds are calculated "taking damage into account" - this is an exact phase on odds dialog. However, the proportion of HP losses stays the same. For example, same strength units have 1:1 odds. Now if defender is half damaged already, odds displays 2:1 for attacker. However, the proportion of HP losses is still 1:1. Since defender has only 5 HPs that means attacker will also lose 5 HPs on average.

All combined the HP losses are proportional to unit strength. When you extrapolate it to large armies that translates to number of units lost since each dead unit loses 10 HP each.

[Alpha Centauri Bear]

The goal of the design presented is to make actively defending against AI ( by attacking incoming forces ) more difficult, without changing AI vs AI dynamics much.
Keep in mind that AI vs AI defender can replace/repair looses much faster and with thinker most factions remain of roughly equal power.
In vanilla AI mid game can conquer anything, *mostly* only after it gains a reactor advantage.
If you take reactor advantage out *and* tilt defenders odds so heavily as you did, in my opinion AI won't be capable of conquering anything.

In vanilla actively defending against AI is already impossibly difficult. When my neighbor amasses forces and invades my territory this is a certain death. Switching all bases to building combat units doesn't help. The only plausible way to defend is to amass striker units and invade enemy territory before they do. The one who does first strike wins. There is absolutely no viable "defend on your territory" strategy in vanilla. Why do you want to make defense even more difficult? In my mod I see that I really can defend with minimal efforts against some barging forces. Of course, stronger forces would break my defense but at least I can defend to some extent while in vanilla any wandering attacker is a threat of losing base.

Let me repeat once more. Please play test it and share your experience. Theorizing is often quite off.

[Alpha Centauri Bear]

Does AI use artillery to purposefully drop defenses to 50% ? With Tachyon field odds are already more in favor of the defender than in the early game.
If you calculate the odds, my proposal doesn't change base conquering odds at all from vanilla, only makes aggressor faction units take less damage in the open, from human defender attacks.
In my experience AIs with thinker are in permanent stalemate, I'd rather like to see more decisive actions between them, not less.

I'd love to have combat resolution function disassembled and it's source code as a part of the mod to tweak.
Or, you could just make 4 versions with binary tweaks: with/without collateral damage from loosing a unit in the stack and with 50%, or 75% facility bonuses.
With access to such tweaks I'd be interested in doing playtesting and experiments.
Some tweaks to the existing source code is the extend of my ability, if it comes to coding, so I can't make it happen myself

Not purposefully. I never saw them amassing artillery fire against a base but they do use it in quantities they have. I guess we can tweak the percentage of artillery built by AI to aid them in that.

Collateral damage: you mean disable it completely? It is already disabled in bases. Do you want to disable it everywhere?

Facility bonuses: do you mean absolute values or on top of intrinsic base defense? I.e. just list these defensive bonuses:

• Intrinsic base defense (exposed in alphax.txt, default 25%)
• PD (hardcoded, currently 100%)
• TF (hardcoded, currently 100%)

PD and TF are cumulative. Either defense facility in base disables its intrinsic defense bonus.

[dino]

Yes, vanilla game favor an attacker in the open so much, that you can keep destroying a stream of units from a few times bigger AI opponent with almost no loses with a pack of cheap rovers and clever tactics abusing predictability of the AI. By removing collateral damage and increasing armor to weapon ratio you could make it impossible.

But after you increase armor ratio, you should imo proportionally decrease perimeter defense and tachyon field bonuses, or AI won't be able to deal with conquering bases, since it usually won't prepare an attack with artillery properly. While we are at artillery, I'd reduce max damage from arty to 50%, and max damage in the base to 70%, for the same reason: AI can't use it to the full potential, by nerfing it you even the playing field a bit, while still keeping artillery worthy of using.

These are my ideas to make it more difficult for a player, to fight against strong AI opponent, while keeping dynamics between AIs mostly unchanged.

[Alpha Centauri Bear]

Yes, vanilla game favor an attacker in the open so much, that you can keep destroying a stream of units from a few times bigger AI opponent with almost no loses with a pack of cheap rovers and clever tactics abusing predictability of the AI. By removing collateral damage and increasing armor to weapon ratio you could make it impossible.

But after you increase armor ratio, you should imo proportionally decrease perimeter defense and tachyon field bonuses, or AI won't be able to deal with conquering bases, since it usually won't prepare an attack with artillery properly. While we are at artillery, I'd reduce max damage from arty to 50%, and max damage in the base to 70%, for the same reason: AI can't use it to the full potential, by nerfing it you even the playing field a bit, while still keeping artillery worthy of using.

These are my ideas to make it more difficult for a player, to fight against strong AI opponent, while keeping dynamics between AIs mostly unchanged.

AI uses artillery and in large enough quantities just not concentrated fire. Tweaking build priority may naturally convert amount of artillery units into bombardment quality. However, concentrated fire is the major AI problem not only in artillery but in general.

I agree that increasing armor theoretically calls for decreasing other defense components if resulting defensive strength become too high. However, equally armored defenders behind TF is still not too high defense in my eyes. I want defense to be three times more effective in the end game. Huge end game production results in massive armies. Economical strength also greatly diverges toward the end of the game. It is not at all unusual for one faction to be 3-5 times more powerful. That results in 3-5 times more units with more advanced weapon. In such conditions triple defense is just bare minimum to slow down the invasion. Without it weaker factions would be all gone long before end game.

Again. This is my experience based on my play testing. I haven't seen TF in action yet but I noticed that PD in the early-mid game doesn't completely block invaders. They eventually break through. I'm afraid it'll worsen toward the end of the game. So TF is not a luxury.

I'll see if I can find and expose defense values anyway.

[dino]

As I said these ideas are not about ultimate and fair game design balance, like for a multiplayer game, but about nerfing game rules that player can put to *much* better use than AI.

Also by increasing defense so heavily you make it easier for the player to concentrate defenses in a base and defend without any looses, with enough defenders and these defense values even rover raids won't be necessary, because it'll be almost guaranteed that individual attacker will loose and only damage defender a bit, with enough units in the base not a single one will be destroyed and they'll quickly heal.

As you noticed AI can't into concentrating power, this is the whole point of my proposals, that player unlike AI can, so I am thinking about changes that would allow AI wear down player's defenses even when he concentrates power in a defending base.

I am also not worried about decisive victories in the mid and endgame between leading AI factions, with large empires arising, that could rival the player.

I'll see if I can find and expose defense values anyway.

I was browsing through Scient's database in IDA few years ago out of curiosity and remember reading through combat resolution function, both facility defense bonuses and the part applying collateral damage were disassembled with named variables. This is the main reason I hammer about these ideas. I don't remember if artillery combat was also there, but very likely it was.

[Alpha Centauri Bear]

changes that would allow AI wear down player's defenses even when he concentrates power in a defending base.

That is the root of all SMACX combat problems. You can theoretically tune combat odds to be very close to 1:1 in one scenario. However, number of scenarios and variations are huge. Due to other odds modifications you'll see them ranging from twice as low to twice as high all throughout the game. And even slightest (!!!) strength advantage above 20% (like 6:5) almost guarantees survival in a duel. So you see it is absolutely impossible to balance each and every duel properly. Most of the cases units on one side will just drop like flies due to balance currently tilted in favor of one of them.

Due to this impossibility I do not even bother to balance this. Here is what I am for: if somebody's units inevitably should drop like flies let it be attacker. I am completely fine with attacker units dying in large quantities not avenged. That means that there is a certain economical advantage threshold below which invasion is just a waste of resources as you correctly noticed. So one should achieve certain power before invading. That was my goal! Such threshold doesn't exists in vanilla.

[Alpha Centauri Bear]

Scient's database doesn't open in my IDA version. But don't worry. It is possible to find all this.

[dino]

It's interesting, that we see an opposite problem with AI

In my opinion without reactor advantage, it's not effective in the conquest enough and you think it's too effective already.