v1.17 and A.I vs A.I tech "trading"

[kimmygibler]

Agreed, something must change. I have basically stopped researching techs for the first 2 ages. There is no point, as soon as one civ gets it, everyone else has it within a few turns. This was a bad flaw in the patch, but what can be done?

I think there needs to be some reason to value researched techs over acquired techs. Perhaps a culture bonus for developping your own alphabet/pottery/whatever rather than simply picking up the tech of another civ. As it stands, the amount of tech trading kills the game. Add to this, the lack of global trading screen and you are forced to click diplomacy screen with every civ on every turn.

[DaveV]

I agree on the military vs. peace imbalance. In my limited amount of playing, I've never had an AI break an alliance; my pattern is to wait for an AI to declare war (or declare myself, if I'm tired of waiting), buy alliances from the other AIs, and eliminate the AI with whom I'm at war. Select next target, repeat.

In my current (Emperor level) game, I've kept my research at zero, and I've been able to purchase AI techs to keep current. In this case, I think it's still to my advantage to eliminate the AIs, since losing one member of their team should slow their overall tech rate.

[Eric S]

There's one thing I've seen wrong with the theory that the AIs treat the human equally. I installed the 1.17f patch midway through a nasty game, huge pangia map with 16 players (something I'm never doing again until I have a 3-4Ghz PC) on Regent (I think, it's the third level).

And sure enough, the only Civs that aren't tech-equals at all times are the ones that are in a bad enough position that they can't afford anything. As soon as one AI sells something, they tech-whore it to all the other AIs, and when I notice that a new tech is available, I'll buy it off of someone. It takes some of the fun out of the game, as it homoginizes your opponents, but it's not anything critical.

However, *NOT ONCE* has anyone offered to sell me a tech before I initiated negotiations. It's not like I don't have anything to offer, as I've got extra Iron, Saltpeter, and tons of gold (you can still tech-whore with 1.17 if you do it right, at least at Regent level. I'm at 100% research with my treasury increasing by 500g per turn ).

[JeffNebraska]

Soren, if you're still reading this, which you're probably not, please know that this tech problem makes 1.17 unusable. It was working perfectly for me and I was very excited about the stacked movement and the improved automation options, but this problem just ruins the game.

Like many here and on civfanatics have said, it makes it USELESS to research techs. No matter how heavily one invests in that strategy, the AI will not only keep up, but will ALWAYS be ahead. I haven't seen a single post yet saying that someone has managed to gain and maintain a lead in techs over the AI on any level other than cheiftan.

It really is such a shame that one miscalculation can ruin an otherwise excellent patch. This turn of events shows what a delicate balance programmers are forced to maintain. Nonetheless, until you release a version of 1.17 that returns to the old tech system, I will not play with 1.17 again. Since I'm now spoiled about stacked movement and the shift-I worker command, I doubt I'll play at all until this issue is resolved.

Please focus all your resources on this problem. I imagine a 1.17g reinstituting the old tech system could be easily released, but if you wait for 1.18 to fix this problem, the quiet rumble of discontentment over this issue will blossom into outright civil disorder.

Good luck and Godspeed.

[Arrian]

Yeah, Dave, that makes sense. If you are going with 0% science to buy tech while warmongering, eliminating civs doesn't hurt at all.

But the builder route (note I don't say "peaceful builder" because even I pick a fight here and there) I think favors leaving civs alive. If you manage to grab the tech lead, keeping all 8 civs in the game should help you keep the lead. Also, in games with a higher number of civs, the loss of one has less impact.

Anyway, it seems that I am not alone in my frustration with the AI tech trading in 1.17. There are many other things I like a lot, including the obvious improvement in AI unit upgrading.

Again, my understanding of how tech devaluation works may be wrong. Soren, please correct me if it is.

-Arrian

Edit: Regarding gaining a tech lead on the AI above cheiftain. I have done it. I had a lead, lost it, and got it back. I finished 1-2 techs ahead of the remaining 2 civs. On Monarch. Of course, that was the game with SEVEN great leaders, where I built both the Pyramids and Great Library (leader #1), along with SunTzu, Sistine, Newton, Adam Smith's and industrial wonders, and captured Copernicus, Bach and Leo's. In a game such as that, I had damn well better be ahead in tech. Yet I strained to keep my lead.

[SirPleb]

I'd like to join the voices who are concerned about the new rate of tech advancement. I agree with the concerns already raised and want to add another perspective.

It seems to me that an important element of balance in the game is the number of turns involved in a finished game. The limit is 540 turns to 2050. I think that a finished game will feel well balanced when as it approaches this limit it is also approaching the end of the tech tree. (Within a fairly broad margin.) If the tech tree is completed dramatically faster then there is not sufficient time available to build most city improvements, and the game begins to lose balance.

With 1.16 I think that the timeline was generally well balanced. A player whose level of skill makes a Regent level game challenging would generally find that research approached the end of the tree after 400 or more turns. Sure some people complained that research was too slow (or too fast) but I think that this often just meant that they should boost their difficulty level.

At Deity level with 1.16 it seemed generally true to me that the tech tree would be finished in 300 turns or less. This caused a bit of imbalance but was not a killer, it was fun to play. It was a constant challenge to try to fill in infrastructure as advancements became available.

"Tech whoring" was possible with 1.16. Not a technique I've played much because I prefer to slow the AI's rate of research. I'm not sure it was a bad thing. The people playing it seemed to be having fun. And I think it carried a penalty of a sort in that the player was choosing to accelerate the pace of the game. At Deity level, giving the AI's advances allows them to start their construction of improvements sooner, at their 6/10ths construction cost, which is a penalty for this technique.

With 1.17 the pace has dramatically accelerated. In many Deity games it might now be necessary to complete the tech tree in something like 200 turns. This will imbalance those games badly since they were already pushing the balance in that regard. (I.e. the side effect of reducing the total duration of a complete game from the 400 to 500 "optimum".)

If anything, I think that perhaps the pace of research at higher difficulty levels should have been made slower from 1.16, not faster. If the "tech whoring" problem needed fixing, a different approach should have been taken.

Hindsight is of course cheap. I doubt that a major change like the following could be made in a patch. But for the purposes of overall game balance I think it might have worked better to "flip" the logic on the higher difficulty levels. Instead of making everything cheaper for the AI, make everything more expensive for the human. That would keep the number of turns from compressing as one learns to play at higher difficulty levels.

[monkeyman]

So, would a simple fiddling with the value of a tech and that decay parameter with widespread knowledge of the tech so that tech trading is less common solve these problems.

It seems to me, if an AI is looking out for itself, as I as a player always do, it would not trade fission, or nationalism for less than an astounding amount of gold. In ITS OWN BEST INTEREST, it is far better to keep another (AI OR PLAYER) civilization 5 turns away from an advance than to net a mere 50 gold. No contest. That sort of thing should be expected.

Looked at another way, the US and USSR invented and manufactured Nuclear Subs 50 years ago. Today, who has that capabilitiy? US, Russia, China, UK, France(?). (there may be more, but not many) My point, a mere handful. I mean, we are talking about HIGH technology. Not like stealing the formula for gunpowder. Tech disparities can and do happen. The scaling factor for already discovered techs should be more gentle.

Also, how about a toggle switch to regulate {AI free-for-all} vs {AI team up vs. player}? Also fun would be permanent alliance toggles...

just a thought or two,

-MM

[Ironikinit]

I found it totally possible to maintain a tech lead over the AI in both the small game and my current standard game since the patch. In both cases, the lead came after I completed the Theory of Evolution. Monarch level, if you're wondering.

I like the current system more than I liked the AI trading on my turn. It makes it more fun because the AI is more competitive. I don't get every industrial era wonder as a matter of course like I did with 1.16. Unlike some, I still research in the ancient period. To me it's all the same, I can spend the money on research or I can spend it buying the techs.

I have become more warlike, though, and can't win with diplomacy in my current game since I extorted a lot of tech.

[wervdon]

Well I've probally played 8 full games post patch now, and I have to say that I have mixed feelings about the AI tech trading. On the one hand, it makes it more of a challenge, on the other its not as much fun if your out to try to get a tech lead of half a age or more and keep it.

I do think there's a easy solution though that'd satisfy most people and makes intuitive sense. Just make the frequency of AI-AI trading difficulty level dependent. At the moment, I literally cannot see a difference between their trading on chieftan vs regent (I dont play any higher than regent).

Just do something like this:
chieftan: AI requires 50% more for its AI vs AI trades, only contacts each other 1/5 as much as now.
warlord: AI contacts each other 1/3 as much as now, and asks 25% more than now.
regent: Basically 1.16 trade levels
above: Basically 1.17 trade levels

Basically, im just saying on the "easy" difficulties maybe the AI trading should be hampered somewhat. 1.16 should be considered "fair", and what we have now "hard". Or some similar scheme. What we have in 1.17 is certainely great for a challenge (and sometimes thats a good thing), but in general I dont think everyone playing wants a challenge like that in every single game....sometimes you just want to dominate for a while and have some fun.


Anyways, sorry if someone suggested this already, I didnt actually read every post in this thread first.

[Pius Popprasch]

Originally posted by SirPleb
Hindsight is of course cheap. I doubt that a major change like the following could be made in a patch. But for the purposes of overall game balance I think it might have worked better to "flip" the logic on the higher difficulty levels. Instead of making everything cheaper for the AI, make everything more expensive for the human. That would keep the number of turns from compressing as one learns to play at higher difficulty levels.

I don't think that would be difficult to implement. There could be a level dependent factor for the beakers required by the human player. Before starting my first game I assumed that it would be handled like that. I thought that research would be very cheap under Chieftain. Like it is now Chieftain players have a hard time because research goes really slow. And Chieftain is a beginner setup.

[Windwalker]

I don't know about y'all, but I've managed to get a tech lead on regent in the middle ages, and then keep it for the rest of the game. I think the tech trading "problem" only really comes into play at higher difficulty levels, where the AI has built-in advantages (I think Regent is the only difficulty level where AI has exactly the same production and research capabilities as player, right?). So maybe the aggressive tech trading is made to give players a harder challenge at higher levels. If that is the case, maybe y'all playing on the higher levels should drop down a level, and see if the game is about the right challenge for you...

About the arguments, I agree that the AI should be hesitant to trade techs too cheaply, because the deal would help the other civ immensely, and not benefit the tech leader all too much. The AI civs should look out for their own interests and not trade unless it benefits them and doesn't benefit another civ too much. They should never trade military techs (chivalry, military tradition, etc.) to neighbors (unless it's for A LOT, and even then maybe not... no sense trading for suicide), and they should not trade wonder techs if they are building that wonder (and are not halfway done already). The AI civs need to look out for themselves a little more and not just trade amongst themselves for the sake of trading... The whole deal with trading a tech for 2 gold is not very realistic and it's not a very good gameplay mechanic, IMO. The AI civs should not want a backward civ to catch up, even if it's a fellow AI civ

- Windwalker

[Analyst Redux]

"So maybe the aggressive tech trading is made to give players a harder challenge at higher levels. If that is the case, maybe y'all playing on the higher levels should drop down a level, and see if the game is about the right challenge for you..."--Windwalker

Higher upthread, someone posted this same thought, though in a rather more insulting manner. Suggesting that some people are now going to have to drop down a level to find the "right challenge" is rather a significant misreading of the vast majority of posts in this thread.

It is no particular challenge to beat this game on the highest levels, just as it was no particular challenge to beat any of its predecessor titles on the highest levels. The AI in this game is at its absolute worst in prosecuting war. At the highest levels, the AI can be goaded into war at the drop of a hat. Even with immense production advantages, the AI is not now able, and has never been able, to deal with a decently executed warmongering strategy. Being able to win on the highest levels isn't the point.

Now, if that isn't the point, what is? The points are: (i) being able to rationally choose any strategy other than warmongering on the highest levels; and (ii) retaining some sense on the highest levels that it is a game of 8 competing civilizations, and not a game of one human vs. a seven-headed poly-civ opponent.

With the new system of "aggressive" peer-to-peer AI tech trading, point (ii) is absolutely, utterly and completely lost. Not only does losing point (ii) substantially detract from immersion, and other game enjoyment factors, but it makes this product performance indistinct from any number of (supposedly) lesser products where this kind of AI behavior is routinely employed to cover up the gross weakness of the AI programming.

That we also lose point (i) via this change is only a slightly more controvertable point. Soren has attempted to make the argument that "creative" strategies can allow you to keep up in the tech race employing a builder strategy against the AI on higher levels. But with all due respect to Soren, any number of posters here who've attempted employing those strategies say he's wrong. Any achievement of tech parity is fleeting, and the pace of tech advancement is crushing. Attempting to employ a builder strategy is pointlessly ineffective, compared to the strength of advantage simple-minded warmongering gains you.

So the point isn't that one can't win on upper levels. The point is that strategizing on upper levels in one-dimensional, game play is non-immersive, the AI's strategic crutches resemble that of inferior products, and the flow of game play (and especially the only rational winning strategy) is diametrically opposed to the concept of building a civilization, which last I checked, was what this game was supposed to be about.

[SSBLoveU]

The AI is very warlike. It is difficult to make a lasting peace.

[Zachriel]

Originally posted by Eric S
However, *NOT ONCE* has anyone offered to sell me a tech before I initiated negotiations.

That's strange. They start offering me trades from the first time I meet them and continue the rest of the game.

Originally posted by JeffNebraska
Like many here and on civfanatics have said, it makes it USELESS to research techs. No matter how heavily one invests in that strategy, the AI will not only keep up, but will ALWAYS be ahead.

I usually buy techs until I switch to Republic. Then I usually stay ahead quite nicely. It is important to me as I like to get a few Middle Ages wonders. I usually trade techs, so I don't have a problem if the AI does too. Leaving yourself out of the loop hurts only yourself.

Originally posted by monkeyman
Looked at another way, the US and USSR invented and manufactured Nuclear Subs 50 years ago. Today, who has that capabilitiy? US, Russia, China, UK, France(?).

Though only a few nations have nuclear subs, many have the technology. They are just not very useful in today's world. For instance, I'm sure Norway or Israel could build one, but why would they?

World nuclear sub club:

Britain 13
France 11
China 6
U.S. 101
Russia 110, er 109
India is building one

[Panag]

The AI trades (again) techs during my turn , grrrrr , was this not supposed to be fixed ?

have a nice day !

[kimmygibler]

There is no use researching now. This is now my biggest problem with the game. Some incentive has to be given to research over trade.

[Ironikinit]

Again, I have to say that yes, I can get a tech lead, and that requires research. Tech lead = research. Hello?

Haven't noticed the AI trading on my turn, either. Not since installing 1.17, anyway.

[Lancer]

Though the AI does seem to get a sci advantage through trade that isn't open to me I can usually manage to beat the sci I need out of one of them. However it can turn into an early world war which resembles more of a fight for survival without techs to trade to gain allies like in 1.16f. If things go well it's a nice option to be able to catch up on sci through peace agreements, but must be a much tougher game for pacifist players since 1.17f.

In my opinion the AI advantage was a good idea... but went slightly too far in 1.17f.

[monkeyman]

Originally posted by Zachriel

Though only a few nations have nuclear subs, many have the technology. They are just not very useful in today's world. For instance, I'm sure Norway or Israel could build one, but why would they?

World nuclear sub club:

Britain 13
France 11
China 6
U.S. 101
Russia 110, er 109
India is building one

Could Israel or Norway really build nuclear subs that are anywhere near as effective as those of the US or Russia? Is the Indian nuclear sub going to be anything but a feather in their cap? I have my doubts. Another example is the question of Israel and its air force. How much of that is actually manufactured in Israel? How much are they buying? In fact, technology transfer is very complicated and must take into account secondary (applied) technologies which reduce the cost of utilizing more advanced applied and theoretical technologies.

Secondarily is the question of equivalent technologies. Just because a country can produce jet fighters does not mean that those jet fighters are going to be competetive with the jets of the most modern countries. If a country produces a fighter equivalent to those flown by the US during the Korean conflict, they will have NO chance against a US F-16 for example. Also see my question about nuclear subs. I think we saw some of this in the Gulf war re the anti-air missile tech that Iraq was using being almost totally ineffective.

So technology transfer has to take this into account, and thus in this simple CIVIII tech model tech transfer SHOULD be slower if we are going to nod to reality.

FFT

-MM

[m_m_x]

some off topic fact
israel has nuclear subs....

[Zachriel]

Originally posted by monkeyman
Could Israel or Norway really build nuclear subs that are anywhere near as effective as those of the US or Russia? Is the Indian nuclear sub going to be anything but a feather in their cap? I have my doubts. -MM

Any sub with a nuclear missile is certainly no mere window dressing.

Concerning your point that techs in Civ3 transfer too fast; the U.S. exploded our first A-bomb in 1945, the Soviets exploded their first H-bomb in 1953.

I've found that technology transfers at different rates depending on the strategic situation. Yes, sometimes every Civ is at approximately the same level of technology. Other times, if the AIs are fighting they don't transfer techs at all. And once a Civ falls behind, it rarely catches up, as it has no money to buy techs and can't develop them independently.

[Zachriel]

Originally posted by m_m_x
some off topic fact
israel has nuclear subs....

I thought that all Israel's subs are diesel, either Dolphin-class or the older German Gal-class. ?

(Of course, they are certainly capable of launching nuclear missiles.)

[m_m_x]

lately we bought some new dolphin class subs from germany
they are already have been modified to lunch nuclear missiles...

[Zachriel]

Civs have a strong incentive to trade techs.

Case: Behind in techs.
This is obvious. If you are behind, you certainly hope that other Civs will provide you technology.

Case: Ahead in techs.
Looking at the strategic situation, you usually see that some Civs are stronger than others. The stronger Civs are your rivals. By supporting the independence of the smaller Civs, you are countering your rivals. Your rivals must expend resources to defend, or if they are attacking, they will need more force to accomplish the same goal.

So every Civ has an incentive to trade in order to keep the strategic situation balanced. The only exception is trade between the rival "superpowers."

[monkeyman]

Lots of good points, others on this list obviously know more about the order of battle of world powers than I, but just look at all the nations that DON'T have the technology. Also one must make the distinction between buying the product of the tech and buying the tech to produce the product. The game has no mechanism to distinguish, which is a flaw that was deliberately introduced to minimize exploits... ...perhaps some of these problems would go away if you COULD trade units again, but including AI algorithms which will better distinguish value of units and would disband lesser units in favor of better units when they are gifted, etc. This way, you could arm your favorite vassal state with cavalry without giving the the corresponding tech. Win-win situation IF an effective AI algorithm is present. For instance, if an AI is building fighters anyway, it should prioritize purchasing jet fighters. Building cavalry? >>seek tanks.

So I say, more cutthroat AI's, less tech transfer, more transfer of goods.

-MM

[Zachriel]

Originally posted by monkeyman
Lots of good points, others on this list obviously know more about the order of battle of world powers than I, but just look at all the nations that DON'T have the technology. Also one must make the distinction between buying the product of the tech and buying the tech to produce the product. The game has no mechanism to distinguish, which is a flaw that was deliberately introduced to minimize exploits... ...perhaps some of these problems would go away if you COULD trade units again, but including AI algorithms which will better distinguish value of units and would disband lesser units in favor of better units when they are gifted, etc. This way, you could arm your favorite vassal state with cavalry without giving the the corresponding tech. Win-win situation IF an effective AI algorithm is present. For instance, if an AI is building fighters anyway, it should prioritize purchasing jet fighters. Building cavalry? >>seek tanks.

So I say, more cutthroat AI's, less tech transfer, more transfer of goods.

-MM

You are right that technology is not evenly distributed. Many countries were purposefully limited in industrial and scientific development during the colonial period. And I agree that unit trading would be an easy and logical addition to the game.

Life is much more complicated than Civ3, though. For instance, I'm sure that they can make guns in the third world, but because they are manufactured far better and cheaper in the the West, local industries have a great deal of trouble staying competitive and in business. It's not that they don't know what gunpowder is, though. They have the technology, but not the infrastructure.

I have seen that situation many times in Civ, especially with the Romans or the French. They get behind, become weak vassals of the larger Civs, and never catch up. No tech, no money, no future.

[Selkirk]

Don't forget that on deity, the computer starts out with what? a dozen warriors? They have far more capacity to get goody huts and with all of their units exploring. Even the non-expansionists have better maps than the human player. They also have a much better chance of being first to contact other civs than the human player does. These first contacts are valuable. Not only to sell the contact, but remember they likely have disjoint maps at the point when contact is established.


I think the tech trading is just fine. Deity SHOULD be hard and I suspect that it is fair. (although perhaps not fun for certain styles of play)


So far I have been able to keep up by:

Aggressively exploring and trading first contacts and maps. (Prior to .17f, I rarely explored farther than necessary to find resources and city spots and built ships only if absolutely necessary.)

Trading excess resources and luxuries for techs (instead of gold).

Buying the dead ends and optionals of the tech tree later at discount prices (horseback riding, republic, free artistry, espionage, etc.)

Beating the crap out of my neighbors at regular intervals.


Also, I have found that at some point in mid game things will start going badly for one or two of the computer players and all of the other computer players will pile on the same guy and declare war. There is no reason not to join in. You might even be able to sell a military alliance. Wait until the target empire has been razed from the main land mass and they are only on some little tiny island somewhere, then ask for peace. If you get them at the right time, they will give you a bunch of tech when you haven't even attacked any of their cities or units. That new capitol with two forested tundras will need all sorts of extra iron, horses, and saltpeter and rubber too.

[Aeson]

If you watch the AI use it's extra units, they really don't do anything useful with them. They will send stacks to target barbarian camps, but if barbarians are off, only 1 or sometimes 2 warriors (or the expansionists scout) will be out scouting. The extras just mull about in the culturally claimed territory, waiting for a Settler to escort. Unless the AI is an Expansionist civ, they also seem to avoid opening huts close to their cities.

I just played through the Ancient Era on a Deity game, with the Iroquois. It was a huge map, so my 15 or so Scouts kept me just about even on tech. It helped that 2 of the first 6 huts were Settlers, my second Deity opening with 2 Settlers from huts since installing the patch. I have the fixed version (no armies from huts) but it does seem that Settlers are more common, might just be luck though.

The AI seems much more moody with the patch, 6 of the 8 AI have declared war on me, some of them twice, and it's only 500BC. Oh well, Mounted Warriors still seem to be extremely powerful, the retreat roll doesn't effect them nearly as much as plain Horsemen. I've been buying up techs that everyone already has, or demanding them from my neighbors. I can get Middle Ages techs for 80-150 gold each once the other 8 Civs have them, the tech rate is pretty scary. Might be seeing launches before 1000AD on Deity. My tech slider has yet to get off of 0%, might stay that way the whole game, at least until the AI has been crippled enough by my MW's and their successors.

[ikol]

Originally posted by monkeyman
Looked at another way, the US and USSR invented and manufactured Nuclear Subs 50 years ago. Today, who has that capabilitiy? US, Russia, China, UK, France(?). (there may be more, but not many)

nuclear subs launched by the US in 1955. France's nuclear subs in 1965 or something like that. Not sure China has any though.

ikol

[ShuShu]

And I must admit I like this new model better. I play on emperor. I find I have manged to get and keep a tech lead, albeit only for a short time.

I definitely like the fact that the AI's do not give over their research capabilities to the human player with higher and higher per turn pay outs.

Please note, I NEVER START wars so I am getting ahead on tech without conquest.

In the early stages, I am way behind, but I wait to see if I win the Great Library race before I start paying. As soon as I lose the race (70%), I buy all that I can, spreading the money around and including a small per-turn incentive to stay friendly with me.

Until education becomes available, minimize research (one scientist is best). Once you find eductaion, it is time to start shifting over to research instead of purchase. This conversion is only possible in the middle ages because the AI strays from the main path giving you a lead to the Industrial Age.

Only sell techs for Resources and MPPs after war declared against you. If you sell the tech to one for this reason, sell it to all for any reason.

Resources and Luxuries are the key to trade now. The world is full of warrior tribes. There is always war and they like to raze cities. Your stong culture will expand to fill gaps. Have settlers ready to fill in gaps and claim resources with cultural brute force.

Any idea how much Horses are worth to an AI in a war?

AIs don't trade with each other when they are at war. The biggest trick to gaining a tech lead is STAY OUT OF WARs. Not only are you able to continue trading, but if a militaristic AI is not fighting you, it IS fighting someone else. If you are in a war, make sure you are not alone. Again, this greatly slows down the global research rate.

Finally, although I don't win every game I play, I don't feel the effort is hopeless. This is the first release where I have seen culture become a major player (peaceful resource acquisition).

I realise this is not in keeping with the mood of the thread, but I actually like the new release much better.