What's a good rule of thumb to trade techs with AI?

[MarcusR123]

Hey community

I am replaying again this year Civ3. So addictive

I upped a notch the difficulty, so I had to start trading techs with the AI. That's what I could gather from many posts here and there.

The problem is that I do not know how. Is there a rule of thumb? Or it has to be evaluated strategically every trade?
Let me explain better with some examples:
How do I know that exchanging Mysticism for Warrior code is a good trade? WC is previous to Mysticism.
Or let's say Monarchy, is it worth trading it for a first-level tech like Pottery? Or do I need to at least get 2 first-level (e.g. Pottery and The Wheel) or one first-level (e.g Alphabet) and one second-level (e.g Iron Working) for the trade to be worth?
By levels I mean steps away from the beginning of the tech tree of each era.

Hope it makes sense what I'm asking.

Please let me know.

Cheers!

 

[vorlon_mi]

In the Strategy articles subform, there is an OLD article on the AI favorite techs https://forums.civfanatics.com/threads/ais-favorite-techs.123345/

What I've learned from experienced players here includes:

• In the first age, trade for techs that the AI likes to research. As you try to meet as many as you can (curraghs can be very helpful, on the right sort of map), trade or sell to get the early techs that you didn't start with.
• Try to research techs that the AI doesn't favor (see above) to trade/sell for the ones they do.
• In the later ages, sell techs to the AI for gold or gold-per-turn. That was counter-intuitive to me; the AI are weaker if they are cash-poor than if I sell them a tech. They also are likely to buy it from each other, so you might as well have their cash.

After the first age, I love to sell a tech which enables a wonder RIGHT AFTER I have built that wonder, e.g., Astronomy. Still gets a good price.
I can't speak to the tradeoffs for Pottery, Mysticism, or Mathematics. Prioritize techs to change your government out of Despotism, either Monarchy or Republic. Most of my games, it's Republic.

 

[tjs282]

Techs become cheaper as more Civs learn them. So if you've managed to acquire a monopoly- or near-monopoly tech to trade away, start by trading it for the most expensive tech(s) available that you don't know yet.

One way to figure out which tech would be the most expensive is to temporarily set your Science to 100% (if it wasn't already), and check with your Science Advisor (F6) which tech(s) would take you the most turns to learn. If anyone else has that tech(s) already, see if they will trade it, and -- if more than one AI has it -- how much each AI will give for it.

Sell first either to the most advanced AI, or to the highest bidder, whichever gives you the most advantage. If you do have to pay some extra gold, or GPT to get that first new tech(s), that's okay. If you need to offer GPT, reduce your Science and Lux sliders until the AI will agree to accept a GPT payment. Reducing both sliders to 0% will show you the highest possible GPT you could pay, but you should not pay more GPT than you can afford to spend over the next 20 turns -- either by getting everyone else's gold on subsequent deals and then deficit-spending, and/or by growing your Civ's population/economy over the next 20 turns.

If you sell to the most advanced AI, try also to get a tech(s) which one or only a few other AIs know. That way, you can add that tech(s) into the next deal you make. Rinse and repeat, either until you've collected all the AIs' techs and gold (if possible), or until you can't afford to make another deal without bankrupting yourself.

If you can get the CivAssistII app to work on your computer, its "Trade Options" tab can be really helpful in figuring out who has what, and which order to make your deals. I believe CrpMapSuite can also give this information.

And/or if you have a compatible .exe (GOG or Steam), the C3X patch-system can be set to add some neat extra functionality to the trade-table screen, e.g. allowing you to cycle through all the AIs directly, and autofilling the optimal gold-amounts when you click the "Gold: Lump sum" or "Gold: Per turn" options on either side of the table (lowest for you to pay, highest that the AI will give) so you don't have to waste any time "haggling" with your Foreign Advisor.

 

[MarcusR123]

In the Strategy articles subform, there is an OLD article on the AI favorite techs https://forums.civfanatics.com/threads/ais-favorite-techs.123345/

What I've learned from experienced players here includes:

• In the first age, trade for techs that the AI likes to research. As you try to meet as many as you can (curraghs can be very helpful, on the right sort of map), trade or sell to get the early techs that you didn't start with.
• Try to research techs that the AI doesn't favor (see above) to trade/sell for the ones they do.
• In the later ages, sell techs to the AI for gold or gold-per-turn. That was counter-intuitive to me; the AI are weaker if they are cash-poor than if I sell them a tech. They also are likely to buy it from each other, so you might as well have their cash.

After the first age, I love to sell a tech which enables a wonder RIGHT AFTER I have built that wonder, e.g., Astronomy. Still gets a good price.
I can't speak to the tradeoffs for Pottery, Mysticism, or Mathematics. Prioritize techs to change your government out of Despotism, either Monarchy or Republic. Most of my games, it's Republic.

Thank you for the tips!

Techs become cheaper as more Civs learn them. So if you've managed to acquire a monopoly- or near-monopoly tech to trade away, start by trading it for the most expensive tech(s) available that you don't know yet.

One way to figure out which tech would be the most expensive is to temporarily set your Science to 100% (if it wasn't already), and check with your Science Advisor (F6) which tech(s) would take you the most turns to learn. If anyone else has that tech(s) already, see if they will trade it, and -- if more than one AI has it -- how much each AI will give for it.

Sell first either to the most advanced AI, or to the highest bidder, whichever gives you the most advantage. If you do have to pay some extra gold, or GPT to get that first new tech(s), that's okay. If you need to offer GPT, reduce your Science and Lux sliders until the AI will agree to accept a GPT payment. Reducing both sliders to 0% will show you the highest possible GPT you could pay, but you should not pay more GPT than you can afford to spend over the next 20 turns -- either by getting everyone else's gold on subsequent deals and then deficit-spending, and/or by growing your Civ's population/economy over the next 20 turns.

If you sell to the most advanced AI, try also to get a tech(s) which one or only a few other AIs know. That way, you can add that tech(s) into the next deal you make. Rinse and repeat, either until you've collected all the AIs' techs and gold (if possible), or until you can't afford to make another deal without bankrupting yourself.

If you can get the CivAssistII app to work on your computer, its "Trade Options" tab can be really helpful in figuring out who has what, and which order to make your deals. I believe CrpMapSuite can also give this information.

And/or if you have a compatible .exe (GOG or Steam), the C3X patch-system can be set to add some neat extra functionality to the trade-table screen, e.g. allowing you to cycle through all the AIs directly, and autofilling the optimal gold-amounts when you click the "Gold: Lump sum" or "Gold: Per turn" options on either side of the table (lowest for you to pay, highest that the AI will give) so you don't have to waste any time "haggling" with your Foreign Advisor.

Thank you that is very helpful. Indeed I am thinking about installing C3X, I am not sure though, it has sooooo many changes that I fear I will be playing a game the original developers never intend to be played. Dunno.

 

[vmxa]

You can start by using just the bug fixes and some of the improvements to play, such as grouping worker task. The trade improvement saves a lot of annoying bartering.

 

[MarcusR123]

You can start by using just the bug fixes and some of the improvements to play, such as grouping worker task. The trade improvement saves a lot of annoying bartering.

Actually, you are right. Thanks! I installed it and the default config doesn't seem to change that much the original ideas. Just balances and fixes the most blatant problems.

Would you recommend other balancing mods to stack with C3X?

There's a trend I noticed for myself during the years, that later was confirmed by everyone in these forums, that is once you get Cavalry and later Tanks/Bombers/Artillery you really don't need much else to crush the AI. So my typical games with Persia go always like this: Immortals, crush the few first civs. Wait for Republic+Cavalry then crush the next ones. And finally with Tanks, crush the rest. It is not fun anymore, because there is no incentive to use the myriad of other units like fighters, naval units, TOW Infantry, Marines, Paratroppers, you name it. What's the point of using those or having a real diversified military.

Also, there's no point on using fortresses, barricades, etc. -- Really a lot of the game options are useless with the vanilla balance.

How in hell Tanks and artillery can cross mountains in few turns? They should be airlifted or carried by naval transports.

Cheers!

 

[justanick]

As a rule of thumb you should try to sell cheap techs first and more expensive ones later. Later can often mean during the same turn and thus not "truely" later. You should also buy cheap techs before you buy more expensive onces. And of course you should research cheap techs before researching more expensive ones. Harvest the low hanging fruit first.

Important exceptions have already been mentioned. So this is really just a rule of thumb. If you overprioritize it, then you risk researching what AI will research and this will not suit you.

Thread 'What will the AI research next?'

Feb 24, 2003

It has happened to all of us. We go full-speed to get a tech so we can sell it to all the AIs and get rich, only to find that some AI is also researching the same tech, gets it before us, and sells it to everyone. Wouldn't it be nice to be able to predict what technology the AI will research next?

Well, after reading this post, you can at least have an idea of the AI priorities.

The AI places a value on each technology it can research next. The values depend on the various things allowed by each technology, as well as on the turns needed by that AI civ to complete the research. Here...

• alexman
• Replies: 89
• Forum: Civ3 Strategy Articles

 

[MarcusR123]

As a rule of thumb you should try to sell cheap techs first and more expensive ones later. Later can often mean during the same turn and thus not "truely" later. You should also buy cheap techs before you buy more expensive onces. And of course you should research cheap techs before researching more expensive ones. Harvest the low hanging fruit first.

Important exceptions have already been mentioned. So this is really just a rule of thumb. If you overprioritize it, then you risk researching what AI will research and this will not suit you.

Thread 'What will the AI research next?'

Feb 24, 2003

It has happened to all of us. We go full-speed to get a tech so we can sell it to all the AIs and get rich, only to find that some AI is also researching the same tech, gets it before us, and sells it to everyone. Wouldn't it be nice to be able to predict what technology the AI will research next?

Well, after reading this post, you can at least have an idea of the AI priorities.

The AI places a value on each technology it can research next. The values depend on the various things allowed by each technology, as well as on the turns needed by that AI civ to complete the research. Here...

• alexman
• Replies: 89
• Forum: Civ3 Strategy Articles

Thanks. My real question is what is a cheap tech and what is an expensive one? What determines its value? How close are they to the start of the era or how much the AI is willing to pay or the turns it takes to research them or... what exactly.

 

[justanick]

Thanks. My real question is what is a cheap tech and what is an expensive one? What determines its value? How close are they to the start of the era or how much the AI is willing to pay or the turns it takes to research them or... what exactly.

If you have not yet researched a tech, then the amount of turns it would take to research it is a good indicator.
If you want to buy a tech, then the amount of gold you would have to pay is a good indicator.
If you want to sell a tech, then the amount of gold you could get for it would be a good indicator if the AI has enough tradeable money. That can be a big if.

If only one civilization has researched a tech, then the trading value is still high. The trading value drops a lot if several potential sellers need to compete on the price. The step from 1 to 2 potential sellers is quite steep.

The research costs drop slightly for each known civ who already has it. In case of 8 civs it is 14/14, 13/14, 12/14, 11/14, 10/14, 9/14, 8/14 and 7/14 of the full price for zero to seven other knowns civs already having the tech.

M

Thread 'Research cost formula v1.29f'

Aug 11, 2002

This is the formula for research cost for v1.29f. I have attached an image of the formula as it is easier to visualize that way than in plain text. Also it is now simplified even more for your pleasure .

Research Cost =

- Research done so far

I'll also write it out for those who cannot view images.

Research Cost = [MM * [10*COST * (1 - N/[CL*1.75])]/(CF * 10)] - Research done so far

Research Cost and research done so far are in gold.

Square brackets indicate truncation /rounddown

MM = map modifier(tech rate on world...

• mydisease
• Replies: 33
• Forum: Civ3 - Strategy & Tips

 

[vmxa]

To me the games tend to go down in a similar fashion. This is not an issue to me. I like the grind of trying to pull out a win from a big hole. To that end I tend to play some form of AW or oversized maps or both. If you are tired of easy winning, play AWDG on a large pan map, very hard to win those games. If you want a lot of work use maps like LK's world map. It was recently posted.

Some of the settings for Flintlock's great tool, that I currently use.

enable_stack_bombard = true
enable_stack_unit_commands = true
enable_disorder_warning = true
show_detailed_city_production_info = true
enable_trade_screen_scroll = true
autofill_best_gold_amount_when_trading = true
skip_repeated_tile_improv_replacement_asks = true
group_units_on_right_click_menu = true
gray_out_units_on_menu_with_no_remaining_moves = true
put_movement_icons_on_units_on_menu = true
describe_states_of_units_on_menu = true

show_detailed_tile_info = true
show_golden_age_turns_remaining = true
show_zoc_attacks_from_mid_stack = true
cut_research_spending_to_avoid_bankruptcy = true
dont_pause_for_love_the_king_messages = true
reverse_specialist_order_with_shift = true
disable_worker_automation = false
indicate_non_upgradability_in_pedia = true

my fav apply_grid_ini_setting_on_game_load = true
enable_trade_net_x = true
ptimize_improvement_loops = true

More, but too lazy to cont.
Using the bombardment currently, but I think it is not always good for AI.
I use it as it occasionally adds some spice to the game. AI pulls up with a stack over 30 covered by 30 unis, it can be scary.
3 or 4 small stacks start hitting one of your towns or an army, very annoying.

fix_ai_army_composition = true
that one is a toss up as well. Can be dangerous, but mostly not.

I am still using r19, lol.