I have a lot to type here, and I have no idea if it's even worthwhile to do so. (Especially with what looks like news of a version of C4AC being ready by the end of the summer). I just hope I can be forgiven for this post if it causes more harm than good. Like through stretching the page or something.
Ever since I heard of the plans to mod Civ4 into SMAC/X, I have eagerly observed the progress and awaited the opportunity to find a way to discuss and/or contribute to the project. In November and December 2006, I
made myself known on Apolyton for a brief time regarding trying to perfect the tech tree. I had plans to reoganize the techs in standard Alien Crossfire and present my efforts. But for reasons unknown even to me, I couldn't get around to it. I'm hoping that won't happen a second time.
Like in my previous efforts, my two largest priorities are regulating the degrees of freedom in the tree and preventing dead-end techs. Make the tree too thick and stuff can be chased after out of order, or people might get confused by trying to scroll through too many options. Make it too narrow and there aren't enough choices, players can't really prioritize, etc. While my issue with dead-end techs is that if their immediate benefits aren't useful in a game, it can't rely on being useful later on. And I've been known to try to plan for tech trees of indefinite length, for which dead-end techs could potentially pile on until one has to scroll indefinitely to seek that one desired tech advance. I don't want something like that to happen.
After my last attempts at proposing a tech tree for C4AC, my considerations for what makes a tree of certain width switched over to how many choices one has on average. I set up a program that would research tech advances at random. (I chose at random because ideally the techs and their applications would be balanced enough that they have even chances of being the best option in any situation.) Anyway, I have the program go through a test tech tree anywhere from 10000 to a million times. Each time it'd take the number of options for the next tech to be research, convert it to a number proportional to the natural log, and add them together. Seeking a geometric mean of how many options are available after a certain number of tech advances have been researched, I then plotted the numbers into Excel, divided them by how many times the program went through the tree, and converted them back. For the tech trees of SMAC and Civ4, I've gotten charts like these:
What I'd ideally like to see is a tree that has its average number of choices start at 6 and stay at the 5-7 range until the late game. ...the narrowness of the tree always gravitates toward the number of "Future Tech" or "Transcendent Thought" techs available.
Recently I decided to try a go at seeing if I could create a tech tree based somewhat off Maniac's SManiac, somewhat off normal SMAC/X, and somewhat off a few of my own ideas, mostly renaming a bunch of the tech advances like Neurology and Cybernetics. I managed to shove in 94 advances, and yes, everything still ultimately leads to Transcendent Thought. My tech tree also relies on a few assumptions, like many factions would start off with three tier-1 advances each while some would start with a teir-1 advance and a tier-2 advance. This would be particularly important for the Gaians, Planet Cult, and Pirates. Sea Colony Pods I'd prefer in Doctrine: Flexibility (or equivalent) and Centauri Ecology in my current proposal tree requires Biogenetics or Doctrine: Mobility. (In a random collecting of techs in SMAC, Centauri Ecology averages as a later research than the other six. And I chose to have only six tier-1 techs. Also, I took Progenitor Psych out of the tree.)
Of course, I'm sure all of you have your own ideas of what techs should be in the game, in what order they should show up on average, and what advances they should probably be related to. With that information I could then jumble the advances together, use my program to make sure the tree isn't going haywire somewhere (like the very beginning), and use another modified version of the program to figure out how much research would go in before one gets a certain advance. Thus working to place the advances in a desirable order.
A few things to keep in mind:
Suppose you have techs A, B, C, D, and E. C requires A or B. D requires A. E requires A and B. It's easier to get C than D than E.
If B requires A and C requires A or B, C essentially doesn't require B at all.
If B requires A and C requires A and B, it's redundant to say C requires A. Every tech tree I've seen made by the professionals has made this mistake, even Civ4 with Nationalism -> Constitution -> Corporation -> Assembly Line -> Fascism.
Anyway, the tech tree I've come to pull together is as follows. Feel free to be struck with awe, disgust, both, neither, something in between, or something else entirely.
I think others may need to determine the applications of each advance. I might help with some ideas like a Network Satellite (research) or Graviton Park (happiness and industry). I suppose a major recommendation of mine is for each advance to have the same benefit, or for those benefits to gradually increase along with the progression of the game. So Optical Computers had better provide a benefit right away, whether it be a direct research boost, a second version of the network node, a new unit, or a combination of those. It'd also do best that each tech provide at least one benefit that doesn't rely on getting there first, lacking a superior unit, a certain victory condition, or having a tech advance that doesn't lead to it. So some changes would have to be made to techs like Biomachinery, Secrets of Creation, Threshold of Transcendence, and Secrets of the Manifolds.
By the way, here's a graph for average number of choices across my rough draft tech tree.
I suppose this is the point where I wait and see if any of my ideas have any use here? ^^;;;;
, 
, 
, 
, 
, 
, 
, &c.
wait till you see what we got cooking for ya.
