can anyone explain me ai to ai trade rate?

[sumthinelse]

Originally posted by Qitai
The main changes that makes research more viable in 1.29 is not the research cost. But, rather, the valuation of tech in 1.29 is double 1.21. A new tech valuation is now 160% of the research cost compared with 80% previously. And an old tech is valued at 70% of the research cost compared with 35% previously.

With the old trade cost, it is obvious buying tech is makes much more sense then researching tech. But given the new valuation, I have to start thinking that research might be better. Additionally, because of the change in valuation, AI trades so much slower now. This give human much more opportunity to actually sell any tech. Hence, research becomes a more viable strat. This does not, however, prevent players from being behind tech in Deity. Just that research has become a viable strat now.

Qitai, your post is the best insight into the tech-related 1.29f changes I have read. Where did you find this info?

Thanks!



Since it slows down the AI trading, this new way of evaluating the worth of techs may allow the human player to keep a tech lead (assuming the human plays well) at higher levels. If the human sells only "old" or "dead end" techs to the AI (for example if the human is researching nationalism, selling chivalry to another civ will not help the AI to catch up entirely) this may allow the human to keep a tech lead or keep near the tech leaders.

Let's see.... At deity the AI will accept about 60% (.625) of the gold it thinks that a tech is worth, but now the tech is worth twice as much, so at deity in 1.29f you would expect the AI tech trading to be as slow as (or maybe slower than) what the tech trading at chieftan used to be in patch 1.21f.

Does that make sense?

Of course the AI still has an advantage in the cost of research, but I would think this would make the game fairer and better.

I have not played a 1.29f game far enough to verify this, but it sounds good.

Thanks Catt, Qitai and everybody for clearing this up.

 

[Lt. 'Killer' M.]

same feeling here!

 

[sumthinelse]

Another thought: If building libraries and universities is worth doing now, that would make it a lot less likely for a human's city to culture flip!!!! . Firaxis should make this known to players. Did the readme.txt file get written before they knew what the changes were??????

 

[Lt. 'Killer' M.]

sumthinelse: I get flips despite ALWAYS building as much culture as possible....

you still on the move?

 

[sumthinelse]

Originally posted by Lt. 'Killer' M.
sumthinelse: I get flips despite ALWAYS building as much culture as possible....

you still on the move?

Do you build libraries and universities in every city? Maybe I have just been lucky.

I finished my plane ride but have some other things to do.....

 

[Qitai]

Originally posted by sumthinelse


Qitai, your post is the best insight into the tech-related 1.29f changes I have read. Where did you find this info?

Thanks!

Glad that you find the info useful.

Those are from observations and testing. I sometimes spend 1 hour when contacts are made, just to figure out the best deals (sequence of trade). I keep records too.

 

[Qitai]

Originally posted by mydisease
Freaky cross-posting, we posted almost the same thing.

Probably a stupid question, but what exactly is it that doubles tech valuation in 1.29f?

Not sure if I understand your question. An example.

At deity, on a large map, the new base (non-contact) research cost for Mysticism is

4*32*10/6 (tech base*map size*difficulty) = 213

With the old valuation, if only one civ knows this tech, it will only sell the tech to you if you pay an equivalent of approximately

80%*213 = 173 gold

With the new tech valuation, it will now cost

160%*213 = 346 gold

Hope this helps. I can explain more into this.

 

[Qitai]

Something to add about how AI to AI trades. At deity, AI would trade with other AI Mysticism at 346/160% = 216 gold between each other. Thus, if they have that much gold or the equivalent of that much gold in tech + gold + gold per turn, they will trade.

You can check this by checking that the gold and gold per turns actually changing hands.

 

[Lt. 'Killer' M.]

Originally posted by sumthinelse


Do you build libraries and universities in every city? Maybe I have just been lucky.

I finished my plane ride but have some other things to do.....

My build order usually is Temple, Library (Courthouse), Cathedral, Marketplace, University (if available), Barracks, Military (to garrison other cities)...... I normally get a culutre boost from Education, because I build Universities a little faster than the AI.

Go decorate your house

 

[mydisease]

Originally posted by Qitai


Not sure if I understand your question. An example.

At deity, on a large map, the new base (non-contact) research cost for Mysticism is

4*32*10/6 (tech base*map size*difficulty) = 213

With the old valuation, if only one civ knows this tech, it will only sell the tech to you if you pay an equivalent of approximately

80%*213 = 173 gold

With the new tech valuation, it will now cost

160%*213 = 346 gold

Hope this helps. I can explain more into this.

So what you are saying is that tech valuation=2*research cost where previously it was just tech valuation=research cost (obviously this is a simplification as valuation is affected by more than just research cost). Sorry, I thought it would be a more complex, specific change than this.

In this case, won't this change make Regent and below significantly easier. The AI will not trade tech as much and AI civs behind in tech will not be able to afford to keep up. Their position in the tech race will be related only to their postion in the power graph. Meanwhile the human will be able to push ahead further and further without having to worry about selling tech to prevent the AI from doing so. There will be less AI to AI trading so the civs behind in research will benefit even less from the drop in research costs due to the rapid selling of new techs amongst the AI. And with the AI to AI trade rates on Regent and below being insufficient to amalgamate this, the game becomes even easier to beat on Regent and below. Which begs the question, what on earth is the point of Chieftain and Warlord?

 

[Lt. 'Killer' M.]

mydisease: i find that on Regent, AI that fall behind stays behind - period! I ahev a game running, huge map, where i tried out dome stuff on diplo, and I find that some civs with half my territory are researching Sanitation while I research Fission. The second best civ is 8 techs behind me, too

 

[mydisease]

As I thought would happen. Tech tree position will be related only to the workable territory size of any AI civ (as differences in territory type will cancel on average to give the same average gold per square). Generally, number of military units will be related to the same thing, so relative tech tree positions will be static. However, civs further up the tech tree will have better quality military, so will eventually beat civs that are behind as they have larger, more advanced militaries.

Is this a good thing however? It becomes very difficult for smaller civs to survive and stay up with the rest, leading to a less variable, more predictable and less challenging game.

 

[Lt. 'Killer' M.]

Yes, it is good, since now the ridiculous catching-up is gone, and yes it is bad, because you are forced to expand 8and fight) all the time..... Also, the AIs incapability to deal with corruption and the limit of OCN that they daren't cross as a result makes the human win all the time - if you manage never to fall behind too far you will get stronger and stronger by and by however stupid you are, just by keeping every city you take while the AI razes them....


But I wouldn't knwo how to solve that problem.

 

[mydisease]

Originally posted by Qitai
Something to add about how AI to AI trades. At deity, AI would trade with other AI Mysticism at 346/160% = 216 gold between each other. Thus, if they have that much gold or the equivalent of that much gold in tech + gold + gold per turn, they will trade.

You can check this by checking that the gold and gold per turns actually changing hands.

Can I just take this opportunity to say that 1GPT = 18 Gold, not 20 Gold as expected. Seems the AI wants interest on GPT.

 

[Qitai]

Well, it is not really true that Regent will suffer. Remember the base cost is still affected by the difficulty level. The base cost use in the fourmula is always the human's base research cost. So, at lower level, humans can research easier. Or even a bonus! And that will bring the trade cost down too.

As for the gpt. The observed exchange rate is 18 gold for 1 gpt as you have noted. Each item on the diplomancy has a valuation formula regardless of age. That is why things like world maps and contacts seems so valuable in ancient age, but is near useless in modern age. And techs can net you a few hundred gpt in modern times if the AI can afford it.

 

[Catt]

Originally posted by mydisease
As I thought would happen. Tech tree position will be related only to the workable territory size of any AI civ (as differences in territory type will cancel on average to give the same average gold per square). Generally, number of military units will be related to the same thing, so relative tech tree positions will be static. However, civs further up the tech tree will have better quality military, so will eventually beat civs that are behind as they have larger, more advanced militaries.

Is this a good thing however? It becomes very difficult for smaller civs to survive and stay up with the rest, leading to a less variable, more predictable and less challenging game.

Originally posted by Lt. 'Killer' M.
Yes, it is good, since now the ridiculous catching-up is gone, and yes it is bad, because you are forced to expand 8and fight) all the time..... Also, the AIs incapability to deal with corruption and the limit of OCN that they daren't cross as a result makes the human win all the time - if you manage never to fall behind too far you will get stronger and stronger by and by however stupid you are, just by keeping every city you take while the AI razes them....


But I wouldn't knwo how to solve that problem.

It's early yet in play-testing 1.29f, but I think one of the consequences of the tech rate changes will be to make it easier on the human player, even in come-from-behind situations; and while I acknowledge the possibility of tech stasis among civs depending on territorial control, I think that fear may be somewhat overblown.

My experience (including the vicarious experiences from others on these boards) is that human players are much, much better at maintaining a lead over AI opponents once said lead is achieved - especially as concerns territorial ownership leads. Under previous versions, however, significant territorial superiority by a human player would not assure technological superiority - with 4 or 5 viable but smaller AI civs, the large human player could find himself falling behind the technology curve.

OTOH, my experience indicates that AI players are not very good at all in maintaining a lead over human players (or other AI players) once a lead is achieved. This challenge is exacerbated by the AI civs proclivity to "go after" the leading civs (human or AI) in an attempt to knock the leader off its perch, and the fact that even a leading AI seems to have difficulty ending wars quickly. Particularly when the leading AI civ is non-religious, my expectation is that an AI civ with even a significant territorial advantage will be susceptible to losing the tech lead through governement changes to Communism / Monarchy during prolonged fighting, and due to the treasury modifications and build priorities (more gold on entertainment / more builds of military units) that are needed during times of heavy fighting.

In summary, I expect that a moderate sized human empire may be able to largely "keep up" or even lead the tech race over a larger AI empire, but the chances of a moderate-sized AI empire keeping up with a larger human empire will be very small indeed. Time will tell, of course - this is all just my speculation.

 

[Lt. 'Killer' M.]

Catt: yep. I see it all the time - even though I play nonsense to test AI diplo reactions I win.....

 

[sumthinelse]

Originally posted by Lt. 'Killer' M.
and yes it is bad, because you are forced to expand 8and fight) all the time

??? What is it about 1.29f that compels you to expand and fight? Don't you have to expand and fight in all the patches?

 

[DrJambo]

something to think about regards culture flipping even though it appears you have tons of culture improvements...

propaganda

if you or the ai successfully uses propaganda then the message that appears to inform you is exactly the same as if it didn't happen through propaganda.

 

[Lt. 'Killer' M.]

Originally posted by sumthinelse


??? What is it about 1.29f that compels you to expand and fight? Don't you have to expand and fight in all the patches?

Up to now, it was easier to play tiny empires and just buy the tech. Have you tried a OCC with 1.29????? Nocando, at least not like before.