[C4:AC] Revised tech tree discussion

[Kuciwalker]

Originally posted by Rubin
Kuciwalker, it may very well be possible to do a Design Workshop using the SDK; however, combining the different unit graphics components within the game seems highly problematic. There may be an option to use the "UnitMeshGroups" system to combine the different modules, or perhaps something else. I don't know.

Ew, you're right. That's a big problem. Especially since we don't have access to the engine AIUI.

I think this could apply to "elevation terraforming" too. It may be possible to implement the mechanics of lowering and raising terrain, but the graphical representation seems... highly problematic.

I think it would be doable if you strictly limited the maxima and minima (well the latter not so much).

[Kuciwalker]

Hm, this sounds like what you were suggesting, but couldn't you treat it like a combination of two separate units a la a Warlord? Infantry guy with a laser turret in front, etc.

[Rubin]

IF we decide upon a set of graphics representing the most common unit designs and the few very special units designs (maybe around 80-100 units; each with a unique graphical representation) AND IF we are able to dynamically choose one of these representations based on a "multiple choice" query within the game; THEN we may have both the Design Workshop and a graphical representation of the units designed.

Unit designs within limited categories (for example infantry garrisons with an armor value of 3-6) share unit graphics. A system of promotions (which also affect the unit cost) are used to visually distinguish (in the text pop-ups) the units within the limited categories.

There are several issues regarding unit design. For example: Do we want/need a Design Workshop? Do we want to focus on the "weapon/armor/movement" design of SMAC/X or do we keep the "strength/moves" design of Civ4?

---

I think the default tier 1 technologies of SMACX make some sense.

Applied Physics: Laser weapons
Early warfare focus. Momentum.

Industrial Base: Synthmetal armor + Secret Project
Early defense. There are no base facilities here, but with a terraforming option to build mines this should be fine.

Social Psych: Recreation Commons
A rather useless facility during the very early stages of the game. Perhaps move this to tier 1.5.

Biogenetics: Recycling Tanks + Secret Project
A powerful early game facility. Builder.

Information Networks: Network Node
A distinct Discover technology and Discover facility.

Doctrine Mobility: Speeder chassis + Command Center
A Conquer/Explore hybrid. The facility would tie up vital production for an early game offensive; hence not a distinct momentum facility. Still a somewhat odd facility option here, I think.

Centauri Ecology: Terraformers + Secret Project
A hybrid technology and seems vital in the majority of playing styles.

Progenitor Psych: Aquafarm
A very odd tier 1 technology! The whole "progenitor" technology branch seems to be a hybrid branch with fewer but superior weapons and armors. Some interesting Secret Projects too. The branch seems to run parallel to the SMAC technology tree and could be a viable branch in itself with a multitude of different playing options.

What kind of changes to the tier 1 technologies do we want? (Or what kind of technologies do we want as tier 1?)

Edit: Added "flavor" image.

[ForesterSOF]

It is temtping to want to change the tech tree.

I would keep it as is for now since changes I want are just for my personal preferences and do not see them improving the tech balancing.

[Darrell01]

Just thought of another special ability that isn't in normal SMACX. Something like "Thought Emitter". Here's what it does. It doubles the chance of that unit encountering a mind worm when traveling through fungus, however, the unit cannot capture the mind worm. Perhaps you could make it so as long as the unit is next to the mind worm, it cannot be captured.

[Darrell01]

And while we are at it, since mind worms will be a predesigned unit, like everything else, you could make more specialized mind worms.
Hypno mind worms that are hard to kill, and they have a minus to psi attack.
Mind worms that are empath song that have a negative to defense.
Genetic atrocity mind worms that will half the population of a base if it wins. Maybe that one will have a minus to psi attack as well.
+1 movement ground mind worms (anti grav struts).
cloaked worms. These worms might be underground, tunneling worms or they move through fungus that is underground. This gives credibility to how a completely terraformed planet can still have worm problems.
aaa tracking worms.
police worms.
probe worms that instead of attacking the base, they just eat at the power lines, doing a drain energy credit probe attack if they come into contact with the base.
worms that can capture units or boats. They kinda burrow in the heads and instead of planting larvae, the body becomes a host. Perhaps units captured in this way become like battle ogres and cannot be repaired. The way you might measure capturability would be if they have negative planet rating. Great way to get zombies into the game.
disassociative wave worms.
Worms that plant fungus. Perhaps after pholus mutegen, they can plant forests as well.
Worms that give a psi bonus to other (perhaps all?) worms attacking near or on the same tile as them.
Worms that lower the moral of defending units. Perhaps it just lowers moral if they are near.
Worms that raise or lower terrain
Worms that "build" fungal towers

Anyway, perhaps these are types of worms you can stick on the "Explore" part of the tech tree.

[TimeTraveler]

I had better post this while it still makes sense for me to do so.

I got the Alpha Centauri Planetary Pack a few months after the release of Alien Crossfire. Perhaps a few months after that, I found the alpha.txt file and that I could edit the technology tree to my liking (to a degree). Since then, I've been obsessed beyond all reason with designing the perfect technology tree.

At the beginning, I had conceptions like each tech having two free-order direct prerequisites, everything ultimately leading to Transcendent Thought, and the number of techs in each level be based on a mathematical function. I was greatly disappointed in the limitations to modding the game, so my dreams of modding a game turned to CTP2, then GalCiv, and now Civ4. Not that I've managed to edit any game. I keep re-planning what I'd ultimately do.

One major issue I wanted to address was to design a tree of a certain level of narrowness. I had my own ideas to that. At first it was the average number of techs per level. Then it was the total amount of prerequisites to each tech, producing a sum that was compared to the maximum possible. SMAC scores at, I think, 1600 out of 2926, or 54.68%.

Then it was the total number of orders in which one can acquire every tech. I tried to create a C++ program that would do that for me, and the results just got larger and larger and there was no way I could figure how to calculate it other than either brute force or through a massive equation with a number of variables I'd have to brute force.

I then went after going with every possible one-instant state in terms of owning or not owning a tech. At worst, the number of options would be 2^n where n is the number of techs. That number could be greatly reduced by chains of advances in the tree. I recall Civ4 got some real decent scores for that. However, I ultimately realized that when limiting the search to certain numbers of advances, I couldn't account for differing tech costs.

I eventually went back to basing the narrowness of a tree on the costs of engaging in a beeline from scratch. By this time I decided that I wanted a tree where the narrowness was more or less constant throughout the game, which I also wanted to extend ad infinitum. For Civ4, I compared the cost of a tech to the minimum cost of a beeline. I also noted that the advances and their beelines rose in a way not too far from exponential. I decided that the costs of the techs of a tree (and consequently an empire's research rate) should rise exponentially (as opposed to a polynomial or power function), and that the cost-of-beeline to cost-of-tech ratio for any advance should be close to a certain constant. (That constant would then be divided by the average usefulness of that tech in relation to average in that part of the tree, but I'd prefer to instead see advances match each other in usefulness.) For Civ4, that ratio is around 9:2. If SMAC's advances rose in cost like the ones in Civ4, its ratio would be near 11:2. I also couldn't help but notice that each ratio shoots up for individual techs near the end of the game. After all, Transcendent Thought requires everything (or everything but String Resonance), and Future Tech in Civ4 requires Composites and Genetics, two of the lengthiest advances in the game to reach.

I still need a way to account for OR-type prerequisites in the current model. Suppose there were two tech trees. In one, each advance would require the one three earlier in the list or the one four earlier in the list. In the other, each advance would require the one four earlier on the list only. In my current model, those two would be even in terms of narrowness, but clearly the former is way more free than the latter. One thought I had was taking the reciprocal of each tech cost and beelining down to each advance. This would produce two numbers. It would also strike the issue of what to do with both advances that require and lead to a lot (Pre-Sentient Algorithms and Chemistry) and advances that require and lead to very little (Centauri Empathy and Archery). I may have to model or even invent all sorts of mathematics for that.

Take into account that I have to combine the narrowness and wideness of the tech tree with the realism of technological and societal progress, as well as how I'd like each tech to always eventually lead to something else (and preferably something less than 100 times as expensive). As has been mentioned already, I also have to ensure that the advances (or branches) are balanced in usefulness. As in adding all the units, buildings, civics, terrain improvements, special abilities, and instant bonuses. And trying like crazy to get everything balanced out as if I figured out Sentient Econometrics for real.

Now depending on what game is to be modded, the advances to be placed into the tree may preferably be chosen and arranged around the condition of the factions in general. In our case of C4:AC, the condition is that of colonists landing on a planet with psychic native life, and starting with little more than the basics of their survival. So I'd keep most of the advances of SMAC in the tree to be designed here. However, I would like to see those advances rearranged. Like I'd place Secrets of the Human Brain after Mind/Machine Interface and Neural Grafting. Neural interface exists in our world but the fundamentals of consciousness have yet to be confirmed, at least to my knowledge.

Hopefully I mentioned everything I need to for this post. Any questions or comments?

[Kuciwalker]

There are several issues regarding unit design. For example: Do we want/need a Design Workshop? Do we want to focus on the "weapon/armor/movement" design of SMAC/X or do we keep the "strength/moves" design of Civ4?

kinda random here, but perhaps the weapon choice would determine strength and the armor would determine "hardness" a la HOI? Different weapons would be better versus soft/hard e.g. guns good against soft, lasers against hard, gatling against soft, missiles against hard.

[N35t0r]

That would be a hassle to mod in, I'd say, as you'd have to add a second number at least... Just using the different modifiers like in unmodded civ would work better, IMHO...

[Littiz]

wow, those screenshots give me new hope

[Blake]

Centauri Ecology: Terraformers + Secret Project
A hybrid technology and seems vital in the majority of playing styles.

I just noticed this.
I think it's a necessity to either make the player start with a former or have formers buildable from the start, or both. Terrain improvement is too important.
It could still be made a powerful, "nearly essential" tech, for example a +50% bonus to formers work rate, or having the usefulness of forests tied to it. Just as long as it's not the only right choice for 1st research...

[Maniac]

You could make formers require no tech at all, and simply tie different terrain improvement to different techs. Eg forests to Centauri Ecology, mines to Industrial Base, solar collectors to Renewable Energy Sources, like SMAniaC does.

I'll state my opinion again that taking over the vanilla SMAC tech tree unchanged will lead to a bad game.

[Rubin]

Originally posted by Maniac
I'll state my opinion again that taking over the vanilla SMAC tech tree unchanged will lead to a bad game.

I think this is obvious at this point in the mod design. We are not doing a SMAC/X clone. We are doing a hybrid of Civ4 and SMAC/X. There are SMAC/X features that we probably cannot implement (i.e. elevation terraforming) and there are Civ4 features that are so good that it almost doesn't make sense if we leave them out.

I asked Darsnan about his thoughts on the SMAC/X technology tree and what kind of "major improvements" were required to balance the mod. He said that "[...]it seems that I reach a point where I have to say that it all depends on what units/ facilities/ SPs are going to be available for the mod[...]".

The base structure of the SMAC/X technology tree seems to have a minor impact on gameplay. It is the "contents" of each individual technology that need major revision in order to make a good game. Hence, I conclude that the changes required for the C4:AC project are 1) minor adjustments to the SMAC/X technology tree structure and 2) major adjustments to the technology contents.

Maniac, do you view this approach leading to a bad game?

[Maniac]

I see it as an improvement, but an insufficient improvement.

[Rubin]

Originally posted by Maniac
You could make formers require no tech at all, and simply tie different terrain improvement to different techs. Eg forests to Centauri Ecology, mines to Industrial Base, solar collectors to Renewable Energy Sources, like SMAniaC does.

Using this approach in combination with some of your other ideas could lead to something like this:

- Terraformers and "build road" have no technology prerequisites.
- Centauri Ecology gives +50% production for Terraformers and makes terraforming time +50% faster.
- The 3 basic (early) terraforming options are distributed this way:

Build mine: Applied Physics (tier 1, Conquer)
Build farm: Centauri Ecology (tier 1. Explore)
Build solar collector: Planetary Networks (tier 2, Discover)

The important issue here is that the different terraforming options are available in different technology branches.

- We apply 3-4 different terraforming approaches and tie these to specific branches. For example:

Heavy Industry (Build/Conquer)
Clean Industry (Discover/Explore)
Ecology/Fungus/Conservationist (Explore (Progenitor techs))

Each approach offers alternative terraforming options to the basic terraforming options. A tentative example:

• The build branch holds no basic terraforming options. Instead a Heavy Industry alternative is available relatively early. Discovering Industrial Economics (or perhaps a tier 3 technology) enables "Advanced Mines".
• Advanced mines get extra production (+1) in combination with production resources and extra energy (+1) in combination with energy resources--in addition to the basic mine modifiers. On non-resource tiles, the advanced mine functions like a basic mine (except for drawbacks).
• The drawbacks to advanced mines are: Longer terraforming time, highly increased ecodamage (high risk of having fungus "pop" on the improved tile or stirring up mindworms), and perhaps a food penalty. (Optional: Use of Adv. Mines could be tied to "ideologies" causing +1 drone to "ecology" ideologies).
• The result is: Superior production (with resources), average energy output (with resources), low food/slow growth, high ecodamage.

Other alternatives to the basic terraforming options could be:

- Facilities that increase output of forests/fungus
- Enable multiple improvements on the same tile (i.e. farm/solar)
- Wind turbines on hills increasing energy and production output
- Specific feature improvements (like lumbermills and watermills)

I hope this explanation makes sense. Any comments?