Which techs do you research first?

This is tricky for it involves miscellaneous amounts of pre-techs selected or not with customization points... but if were to start completely blank (which i do often for modding purposes) here's how i'd proceed;

Xeno-Engineering

Xeno-Research

Hyperdrive

Stellar Cartography

Planetary Improvements

New Propulsion Technics

Ion Drive

Galactic Warfare

Xeno Communications

Universal Translator

Basic Logistics

General Life Support

Xeno Entertainment

Diplomatic Relations

Space Militarization

Trade

Xeno Economics

Soil Enhancement

Xeno Industrial theory

Space Weapons

Interstellar Governments

Alliances

Republic

Sensors

Xeno Biology

Extreme Colonization

--- and, whatever type of closest Extreme planets still available.

(wow, up to 26! This is the Core basic but essential set, in strict order for a number of reasons that can be summarized as -- Develop Infrastructures, Contact, boost Economy (Trade/Republic combo), Fast colony ships, pick some Extremes first.)

--- Then, some Defenses, a few Weapons, Research stuff, BANG:Xeno Ethics, etc.

You forgot Impulse Drive.

What the hell is wrong with you?

Tight or Loose clusters?

But... what the hell is right with you? -- Let's all read about your most optimal path!

Scattered.

Depends on version...DL, DA, or TA?  OP doesn't specify, but I'd assume TA.  Then tech tree comes into question.  I still don't have a favorite, but I'm leaning heavily towards Torian.

Over the next week, I'll make a list for each civ, although the quality will vary somewhat due to my not playing them all equally.

Then insert Impulse Drive right before the first extreme of five done -- for DA, but i wouldn't even try a TA set of twelves unless i had Netriak's Galactopedia in a tight secondary window to refer to.

Besides, i don't have enough experience yet over that much complexity, too busy modder that i am.

Also, does anyone know if the AI will take the situation into account when researching techs? i.e. does it know to develop its weapons and logistics when a war is going on and research economics when it is in the red? And does it do the same thing for galaxy conditions (more extreme planets = more colonization techs & so on).

The AI will only research a given extreme planet tech if there are any of that type on the map.  I'm not sure if it prioritizes distance from its empire/homeworld or if it just counts them up, but in DA 1.8 it really, really, really liked to research the extreme colonization techs, at least in my experience.  It also wouldn't sell them for less than your firstborn and the firstborn of your firstborn.

I'm working on putting together the various things I've written down in my time playing GC2 and I still expect to have something for you some time this week.

It responds decently well to a war; no idea how it responds to debt.  It classically responds horribly to plague/disease, though, which is my main complaint-it doesn't react fast enough.

I have already made sure that in the mod, those techs have the absolute highest value possible, so that it will research them as soon as the infection hits.

Through observations, i concluded that AIs follow a clear pre-defined path of research according to many factors; Abilities, Victory goals (as in prioritize Conquest rather than Influence or not, etc).

They have a focus over everything from first tech to the last. Optimally balanced as to provide the greatest challenge at any given time. Another reason why i insisted on Ethics in my initial list. Pivotal to the point where they MUST research it to gain what they need.

Much of the gameplay situations push them away from that path if only to grab a key 'node' away from the human player or other Ais. Diplomacy & Trade has to play a role too, judging from the *AI value* tags inserted in the xml based tech tree, even more so in TA with the *Willingness to Trade* quite ingenious trigger.

Actually, this is what makes these algorithms sooooo competitive even though the variables are extremely high.

Presume an AI wants to keep the advantages of their current PolParty?

Rarely do they actually fail the Vote, but in specific circumstances they DO take that decision, not because of Unhappiness but to switch gears.

That's how deep the reasonings go, AFAIC.

Have you playtested it yet?  There may be more to it than that.  Like I said, it reacts slowly.  It's easy to tell when a war starts getting bigger, but the plague and disease are one-off events...they don't get steadily worse with each turn.  (Okay, the disease does, but that's beside the point; if you let it get that far you're dead anyway), so they're inherently harder for the AI to cope with.

That's my perspective, anyway.

Considering that both of them have had a very high research and/or trade value for quite some time, and the AI did not noticeably improve its reaction time to them, I posited an explanation that the AI simply couldn't "see" where they were on the tech tree, as has happened to real players numerous times since they tend to "spawn" below everything.

They're still researching power plants in TA.  In the first year no less.  Hell, in the first six months.  That's a Bad Idea.

No. I am currently just tryijng to determine the path most l;ike a human player. I haven't even written the mod yet, much less tested it.

I'm not trying to determine if they are efficient, but only IF there is a strict definitive list of techs numbered #1 to the end (or not, btw) common enough to serve as a basic template where too far deviations on specific nodes can force them into a category switch. DA, not TA i should insist.

Say, Get Laser II but don't forget PDCombo while throwing in a few Miniaturizations % and at least second Warp Level *IF* your Military Ability is +20 and  *IF* you are at war with more than two enemies. Otherwise, skip a single of the above priority items for Democracy **IF** your Economic Ability started with Zero and *IF* three or more opponents already have it.

That's how much calculations the current AIs can do. Bazillions worth per second.

Bad ideas within six months of gameplay? The only time factor that really counts for them is *Present*, as they reset, re-evaluate everything, and decide upon each start of turns.

Edit -Deleted.

Re-reading, I see that you're trying to do this for an AI.  In that case, my approach wouldn't work. 

The AI wouldn't be able to utilize my tech path without getting summarily stomped by the nearest AI or player.

It doesn't have to be perfectly detailed: if you say something like "reserch up to the colonizatuion techs", or "improve PD and missiles" I can sort of interpolate the techs.

All of this reminds me of a request i made to SD staff in the "Please Slam my Game" thread a few months ago;

A true Research Governor from where we could use some sort of external xml listings and simply automate technos gathering. Even my adventures of trying to enumerate these from SandBox mode results file were bound to fail since the parser resets everything in numerical order; the only trace i figured had to be available somewhere.

Here's an excerpt from a very old file;

<TechTree>0|2|3|4|28|53|254|262|</TechTree>

I purposely lost that game to find out... but, i didn't research number Zero(ImpossibleTech) upon starting (Xeno-Engineering#53 & GalacticWarfare#28 & TechVictory#215, etc - in fact) but to my dissapointment the saved result hadn't recorded the actual path and my own progress pace & choices. These numbers identify techs uniquely by reading the tech tree sequentially, slapping in the completed number of techs in the order mentioned above upon Victory, splash statistics, exit.

That way, we would have been able to share all of our stuff right here for Scoutdog project.

I'd really enjoy seeing MottiKhan secrets & tricks (and many more) too, though!

(PS; i still have the reference listing of exact IDs for DA techs... just PM me your eMail Scoutdog, and i'll send it all to you.)

Not really a secret.  It's just the research path I normally take.  I'd imagine it's fairly common among the other all lab players.  I'll repost it since it seems to be of some interest...

(Sorry about it being too detailed though.  This is just a copy/paste from a .txt file that I used to write it up the first time).

Xeno Research
Advanced Computing
Planetary Improvements
New Propulsion Techniques
Ion Drive
Impulse Drive
Impulse Drive II
Sensors
Universal Translator
Diplomatic Relations
Trade
Research Centers
Research Academies
Advanced Diplomacy
Expert Diplomacy
Majesty
Total Majesty
Xeno Ethics
Good And Evil
Concepts Of Malice
Galactic Warfare
Space Weapons
Beam Weapons
Laser
LaserII
LaserIII
LaserIV
LaserV
Plasma Weapons
Plasma Weapons II
Plasma Weapons III
Phasors
Psionic Beam
Xeno Entertainment
Extreme Entertainment
Starship Defenses
Armor Theory
Titanium Armor
Titanium Armor II
Xeno Economics
Xeno Trade Centers
Xeno Bank Construction
Galactic Stock Exchanges

Like I said before though.  If an AI tries to use it, they'll fall to any enemy within reach very quickly.  It'll bankrupt the economy quickly (by design). 

On the other hand, as a human, I've never lost a single game using this approach.  I've used it on maps ranging from tiny to gigantic without fail.

Edit -  and loose clusters. 

That's fine. I don't mind the detail, it's just that I think more peple will be willing to post here if they don't have to write down every single tech.

Scoutdog if you are looking for a place to upload your mod to that requires no moderator clearance please consider using the Galciv2Guide.com site.

All you need to do is register yourself and then go to the mods section and upload your mod and it will be instantly available to all.

Will look into it. As of now, the mod does not exist except as a number of replies to this forum and a few scratched notes on my Global History homework, but hopefully it will be nearly complete by the end of spring break.

I'd called this very special "path"; *Armed & Ready for anything! the Fastest way possible*

Laser V++ could have used Range, Logistic, Miniaturization & Medium Hulls inserted somewhere sooner, but as they say; if it ain't broke -- don't fix it!

Perhaps, but Basic Miniaturization normally comes free early on.  Medium hulls aren't neccesary.  I just use Psionic Beams on small hulls and cargo hulls if need be.  I get better hulls later, but the tech can be easily taken in trade.  Once you get Plasma Weapons III, the AI will bend over backwards to buy it from you.

I haven't built a medium hulled ship since my very early days of playing.  Tiny, Small, Cargo and Huge are the only ones I've used lately.

I also build very specialized ships for specific purposes.  In my current game, I designed a huge fast ship to take out pirates, but only built 4 of them.  One for each corner of the galaxy.  One of these can take out any number of pirates and just move on to the next cluster of baddies.  I also built tiny fast colony ships and constructors.  Lots and lots of them cheapies.

I don't research range at all.  I build a starbase or take the enemy's planets to increase my range.

That, as I see it, is the major problem with a mod like this: different people have different styles of playing. Means I will have to look at the AI and see what style it uses, then adjust the techs accordingly.

Gee, whadda'ya'know, just one more "modifying" situation that our hardcoded AIs must take into account to BECOME efficient enough to compete against the Human brain!

And yet, we can't get any Huge Hulls without having to go through some timely steps -- as does everything else, btw.

Range, soooo true... but lemme give you a secret of my own, i hand over techs for at least one scout per AI and soon after as many of their Miners as i can possibly hold onto to create remote mining operations that flip or not. 1mp here, another there.

So, without some Logistics - medium (or less) maps look like 5 wide gigantics to these scouts & miners.

Here's the basic principle behind what any AI does, tech values in the second last column to the right;

Alliance & some Logistics have 30!

6 more pages, want them?

How are you gentlemen researching Xeno Research before Artificial Gravity and Planetary Improvements?  Xeno Research comes after these techs in the tech tree.