Tweet
By this I mean the minimum price in gold the AI will take for a resource or the maximum gold it will give for one.
Most of this has probably been guesses already but there were just some anomalies that really annoyed me and I just had to find out why. I haven't seen them posted anywhere.
Anyway:
Price of a strategic resource=xt(y+z*up((n+1)/4))
Sometimes:
Price of a luxury=xt(y+up(hf/2))
except when it isn't (and I don't know how either.)
Weren't they obvious?
I suppose you'll want the terms:
EDIT:
x:
Is 20 gold except at least once.
In a retroactive reply to DaveMcW, that's not exactly 1gpt.
t:
This is theNiceOne factor. It is civ dependent and is mostly affected by aggression and difficulty level. It always acts aginst the player so divide by his values when you are selling.
see:AI - human trade revisited
/EDIT
y:
Now this has always been an integer and never negative. I think its 0 when the resource is in abundance. I've seen 1 and 3.
up():
rounds up to the next integer. The only standard way of writing this I've seen is maths not programming so has silly symbols.
n:
Number of connected cities.
Those were obviously not important. The strategically important variables (as far as I can see) are hf and z.
hf:
Total number of happy faces you would get by importing this luxury. Obviously if you have no marketplaces, this is n.
z:
This is a sum of values for each unit,improvement,wonder or worker job the resource allows. Each of these so far is in the range 1-5. Everything from a previous era has a value of 1.
The analysis of whether you can actually build something is somewhat variable. But techs and things like apollo/manhattan are certainly whatever but other resources and weird requirements like coastal cities are not.
For example, for iron in the ancient age z=3, 1 from the Iron Works and 2 from swordsmen.
When you are in the medieval age before feudalism z drops to 2.
Right thats the formula. Now for some irrelevant comments:
1&2: I meant to make these seperate comments but accidently included them above.
3&4: I'll post these later.
Clarity, examples:who cares?
Strategic implications: mostly obvious.
Most of this has probably been guesses already but there were just some anomalies that really annoyed me and I just had to find out why. I haven't seen them posted anywhere.
Anyway:
Price of a strategic resource=xt(y+z*up((n+1)/4))
Sometimes:
Price of a luxury=xt(y+up(hf/2))
except when it isn't (and I don't know how either.)
Weren't they obvious?
I suppose you'll want the terms:
EDIT:
x:
Is 20 gold except at least once.
In a retroactive reply to DaveMcW, that's not exactly 1gpt.
t:
This is theNiceOne factor. It is civ dependent and is mostly affected by aggression and difficulty level. It always acts aginst the player so divide by his values when you are selling.
see:AI - human trade revisited
/EDIT
y:
Now this has always been an integer and never negative. I think its 0 when the resource is in abundance. I've seen 1 and 3.
up():
rounds up to the next integer. The only standard way of writing this I've seen is maths not programming so has silly symbols.
n:
Number of connected cities.
Those were obviously not important. The strategically important variables (as far as I can see) are hf and z.
hf:
Total number of happy faces you would get by importing this luxury. Obviously if you have no marketplaces, this is n.
z:
This is a sum of values for each unit,improvement,wonder or worker job the resource allows. Each of these so far is in the range 1-5. Everything from a previous era has a value of 1.
The analysis of whether you can actually build something is somewhat variable. But techs and things like apollo/manhattan are certainly whatever but other resources and weird requirements like coastal cities are not.
For example, for iron in the ancient age z=3, 1 from the Iron Works and 2 from swordsmen.
When you are in the medieval age before feudalism z drops to 2.
Right thats the formula. Now for some irrelevant comments:
1&2: I meant to make these seperate comments but accidently included them above.
3&4: I'll post these later.
Clarity, examples:who cares?
Strategic implications: mostly obvious.
Comment