Scoutdog Scoutdog

Which techs do you research first?

Which techs do you research first?

I am currently attempting to create a mod that makes the AI reserch a bit more like a human player. Since I plan to upload the mod once this... stupid... GLITCH clears up, I would like input from other players on their research strategies. It would be great if some of the more dedicated players (or anyone, I'm not picky) would tell me a bit about their tech strategy.

UPDATE: The original question is not phrased very well. Basically, what I want is for players (the more expreianced the better) to tell me the basic order of their research progression so that I can tweak the AI values and make them research more intelligently. If you don't have a specific list, I could still use suggestions about what the computer researches badly so that I can correct it. Any info is appreciated, and when the mod comes out there will be karma for the contributors.

44,629 views 41 replies
Reply #26 Top

Quoting kpayton2001, reply 25
How are you gentlemen researching Xeno Research before Artificial Gravity and Planetary Improvements?  Xeno Research comes after these techs in the tech tree.

I'm referencing DA 2.01.  There have been some changes in TA, but I'm not familiar with them yet.

I believe Sole Soul is doing a TA one, but that'll take some time as there are unique trees per race.

Reply #28 Top

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

So... these are the CURRENT AI values? Or a graphical representation of the decision-making process? I'm not following you here...

Reply #29 Top

Both!

I figured that slamming a different value to anything in that file, alters how & when AIs research their stuff.

The "N" column to the right left side is their ID_Numbers as listed in the order they are in the xml; that's the only way to identify what they are in the Sandbox results.

 

Reply #30 Top

Now I see what this is! It's a direct visual output of the data in the XML itself. Took me FOREVER to realise that the little "hexagon" symbols were actually supposed to be tag signs! (What are > and < actually CALLED, anyway???)

Reply #31 Top

The thing is a simple snapshot of an XML-Marker program listing where i "reduced" the width of elements to keep everything i needed to insert the darker black numbers (5-10-15...etc). The yellowish tags are simple shortcut expanders of whatever sections selected.

Certainly easier than notepadding the stuff to find correspondances. Even the top tags can re-order these to say, highest costs above to the lowest.

When you have more than 215 (or more in TA) techs, anything is a lot simpler to analyse in trees.

Reply #32 Top

Bumping. Hopefully I will get a few more people to reply, so I can get a better sense of how most humans play. Also added some clarification to the OP.

Reply #33 Top

Might as well share a tricky "process" i've used to determine effectiveness of tech loops as used by AIs, ScoutDog.

But, it is a huge task to setup.

1- They mostly pick from the fastest (weeks as listed on the sidebar, btw) available.

2- Create a brand new mod folder.

3- Call it TechQuest.

4- Pick a Major, blank their tech list, zero the customization points and hand it over to an opponent status.

5- Do the same for yourself - BUT, as the custom race only.

6- Start a game in Cheat mode. Keep tech trading ON & No-Brokering OFF.

7- Tiny map, One vs One, no Minors.

8- Nullify all YOUR research resources.

9- Stack a bunch of cash.

10- Go to the trade screens often enough to keep track of that AI particular choices.

11- Buy out.

12- Keep Going until every "possible" steps has been recorded.

13- If you're lucky you now HAVE the whole track record for that AI, based on its RC values such as political, ethical alignment & abilities.

14- See the complexity?

15- Adjust ONLY the tech value for Xeno-Engineering 30+ in a temporary TT.xml.

16- New game(s).

17- Watch the evidence.

18- Repeat for everything you need to know.

 

That's how i figured some specific tech paths last summer. I'd still be at it to detect 215 DA-techs worth in highly variable parameters of 12 Majors or not -- and, i'd yet have a few years more ahead of myself.

B)

+1 Loading…
Reply #34 Top

Thnx for the tip. Will probably have to put the mod on hold while I try not to fail Freshman Lit&Comp, but over the weekend I could try this out.

Reply #35 Top

There ARE more important things in life, i'd agree. (No wonder i have so many diplomas in a drawer somewhere!)

Weekend? what weekends. I start at 11:30AM Fridays and stops at 8:00AM on Mondays, extenuated enough for working almost round-the-clocks. Insomnia, maybe but that's exactly what the responsabilities are.

Pay is good though. :erk:

Reply #36 Top

Bumping, possibly for the last time.......

Reply #37 Top

Yep, it's looking like nobody wants to give away their precious "personal strategy" plans!

Have you tried the above monitoring attempt within cheat mode?

Cuz, it releaved a *LOT* to me a few months ago. It took me a whole week to summarize the essential, but with some patience the algorithms became much more evident through simple deduction.

Reply #38 Top

I usually research everything that gives to abilities, diplomacy is also very important since I use it to swap every tech I can get my hands on and I use alot of it on the minor races since they are not a threat to me. What surprises me though is the a.i is amazingly fast in the tech tree even when they don't seem to tech broker much.. I wonder how often or at all if they tech trade with other ai's..

Reply #39 Top

Wow,

Quoting Zyxpsilon, reply 33
Might as well share a tricky "process" i've used to determine effectiveness of tech loops as used by AIs, ScoutDog.

But, it is a huge task to setup.

1- They mostly pick from the fastest (weeks as listed on the sidebar, btw) available.

2- Create a brand new mod folder.

3- Call it TechQuest.

4- Pick a Major, blank their tech list, zero the customization points and hand it over to an opponent status.

5- Do the same for yourself - BUT, as the custom race only.

6- Start a game in Cheat mode. Keep tech trading ON & No-Brokering OFF.

7- Tiny map, One vs One, no Minors.

8- Nullify all YOUR research resources.

9- Stack a bunch of cash.

10- Go to the trade screens often enough to keep track of that AI particular choices.

11- Buy out.

12- Keep Going until every "possible" steps has been recorded.

13- If you're lucky you now HAVE the whole track record for that AI, based on its RC values such as political, ethical alignment & abilities.

14- See the complexity?

15- Adjust ONLY the tech value for Xeno-Engineering 30+ in a temporary TT.xml.

16- New game(s).

17- Watch the evidence.

18- Repeat for everything you need to know.

 

That's how i figured some specific tech paths last summer. I'd still be at it to detect 215 DA-techs worth in highly variable parameters of 12 Majors or not -- and, i'd yet have a few years more ahead of myself.

 

Wow, you really have a lot of time on your hands. There's a much easier way to figure this out. Simply raise your espionage on a specific civilization to the appropriate level (high, I believe) and go to the Foreign Stats->Report page. The report will list, in sequence, the technologies researched (or rather acquired, I should say) by an enemy civilization. By removing, tech capturing, brokering, and trading you can easily see exactly what each civilization researches in sequence.

 

I mean, I actually haven't tested this theory, but I'm pretty sure if you look at the order in which the technologies appear in the Espionage Report you'll at least understand why I thought this to be true.

Reply #40 Top

Figures. I leave the forum for a while, and then this thread shows some activity...... testing out my AI valsued with the above method now.

Reply #41 Top

I do not usually follow a "standard" research path. What and how I research depends on what races I am facing and what I can pick up through trading.

I usually play a custom race and I have Xeno Researcg and Stellar Cartography as starting techs. The first tech I actively research is usually sensors. I generally start a game by building a couple of Scouts. Once I have Sensors, I can upgrade the scouts to a "Surveyor" class I designed, which are able to explore anomalies. I will then worl on propulsion so that I can upgrade the Surveyors to the faster "Surveyor II" (also my own). After that I will concentrate on military rsearch until the times for researching the "Trade" techs drop to a reasonable level. I try not to get stuck researching things that are going to take more than about 10 turns or so. And I try to get medium and large ship-building techs as quickly as possible, too. There really isn't much that can be built on a small hull that will not be blown into little bits the first time it encounters - well, anything! 8C

As I said, I do not have a specific research path I follow, but more of a general strategy. The actual research path will differ from game to game, depending on other game factors.