Permanent Alliances effect on tech cost

[bestje]

This doesn't seem to be covered anywhere else on the forums so I thought I would post it here
After playing a few test games in BtS it seems that the increase cost of teching in a PA is by a multiple of 1.36 recurring

And the working for that:
increase in beaker cost for techs:
rocketry increases from 9900 to 13500 - 13500/9900 = 1.363636 etc
radio from 11880 to 16200
fusion from 15840 to 21600

the increase is a consistent multiple of 1.3636 recurring, this was on marathon speed

and on quick speed
rocketry increases from 2211 to 3015
radio from 2653 to 3618
fusion from 3537 to 4824

the increase remains the same multiple as previously so I expect that it will remain consistent for all techs and game speeds

Implications of the increased cost in beakers: it isn't worth making a PA for the additional beakers unless your PA partner is producing enough to offset the extra cost.

also the figure of 40 turns is approximately consistent at all game speeds, I haven't really tested it but it was about 40 turns at marathon so again I assume that this is a constant

 

[Kesshi]

bestje,

There is an old thread named Technology Research Explained.... In that one it says that the number of teams tech modifier is +(0.5 * # of teammates).

Also, I looked in the GlobalDefines.xml and it says this:

	<Define>
		<DefineName>TECH_COST_EXTRA_TEAM_MEMBER_MODIFIER</DefineName>
		<iDefineIntVal>50</iDefineIntVal>
	</Define>

Also the CvTeam.cpp file has these lines

iCost *= std::max(0, ((GC.getDefineINT("TECH_COST_EXTRA_TEAM_MEMBER_MODIFIER") * (getNumMembers() - 1)) + 100));
iCost /= 100;

I believe this means that the tech cost should be 50% more per team member added, but a lot of that is cryptic to me and I'll admit that I don't understand CPP very well. Have you done any tests on adding in a 3rd or 4th teammate?

Edit: Oh wait, I think I see it now.

1500 is 33% more than 1000. That's where you're getting your 36% from. I bet the game is rounding the "33%" funny and you're ending up with 36%.

 

[moyang]

http://www.civfanatics.com/civ4/strategy/tech_research.php
explained here.

 

[bestje]

@Kesshi and moyang, thanks for the link, surprised my search didn't turn it up though, oh well.

I think the error is in the fact i have got the beaker values in game, so whilst still producing a fixed increase the proportion of that increase is also dependant on the difficulty.

This is the formula for cost of a tech;

Thus the formula to calculate a Technology's cost is:

Technology Cost = FLOOR (FLOOR (FLOOR (Base Cost * (Difficulty Modifier + 0.5 * # of teammates)) * Map Modifier) * Speed Modifier)

dealing specifically with how it can change mid- game:

Because the tech cost in game already includes some of the modifiers (map size, speed) these don't need to be considered in calculating the tech cost change for a PA

the only variable is the difficulty modifier + 0.5 (PAs can only add 1 team member)

so to calculate the cost of a tech post PA what you would have to do is divide the in game tech cost by the difficulty modifier and then multiply that by the difficulty modifier plus 0.5

therefore adding a team mate mid game creates a fixed increase in the cost of a tech at any difficulty level, but the proportion of this increase changes.

Have found the list of difficulty modifiers in this article http://forums.civfanatics.com/showthread.php?p=3475140#post3475140

The difficulty level modifier for the human player is 1 at deity, immortal or emperor level, 0.95 at monarch level, 0.9 at prince level, 0.8 at noble level, 0.7 at warlord level, 0.6 at chieftain level and finally 0.5 at settler level.
The difficulty level modifier for the AI is 0.6 at deity level, 0.8 at immortal level, 0.85 at emperor level, 0.9 at monarch level, 0.95 at prince level and 1 at the levels below prince.

I'm doing something wrong here still as my maths doesn't add up, will have another look tommorrow

but unless i'm much mistaken then the increase relating to signing a PA will cause a much smaller relative increase in the beakers required at higher difficulties.

 

[Robert Smith]

Does the PA increase beakers for AI-AI PA or is only between AI-Human?

 

[Kesshi]

Robert Smith,

It is the same all around. Though individual tech bonuses would apply. That is, Deity AIs still receive Deity level bonuses towards techs.

 

[bestje]

Have played around with these figures a bit more and I still can't seem to get it to fit the formula

however the multiplier is consistent at each level:

deity: x1.3266
emporer: x1.375
prince: x1.3636
warlord: x1.5

I'm wondering if the tech cost is affected by the AIs difficulty modifier as well? this would explain why the cost changes at deity level when according to the formula it should remain the same as at emporer. unless I misunderstood something anyway

 

[Yamps]

From SGOTM 7 MW (Vanilla Deity OCC PA game):

Rules for Permanent Alliance (PA)

We can have one and only one partner
Neither we nor the partner can break the PA
We tell the partner what the partner shall research
If we DoW, the partner will DoW as well
The relationship from other AI will be the average of said AIs relation to us and out partner
It will take 40 turns of mutual war against a third AI before a PA can be signed
OR
It will take 40 turns of Defensive Pact with an AI before a PA can be signed
OR
Shared war and DP count together, but DP counter is decremented while you are at war, so you can top of shared war by a DP, but not the other way around.
Our GP can be used in the partner cities only by settling them as superspecialists.
(Any other use requieres our gifting the GP and hoping the AI will know what to do with them.)

PA research:

There is only one bucket for tech progress per team per tech. Both PA partners put their beakers into these buckets. That's independent of difficulty level.
If a PA is formed the beaker cost of the techs change (you can see this change in the tech screen). There is a penalty of a factor of 1.5 for 2 partners.
But the difficulty level factor changes to the value of the average of difficulties of the 2 partners :
floor((deity+noble)/2) = monarch.
So in the end we have 1.15*1.5 instead of 1.3 factor.