Research cost formula - how does it work?

[Misotu]

OK, I'm completely bemused.

Playing an MP game recently, my tech cost was ridiculously low by comparison with that of my pact mates, and yet we all had the same number of techs.

I am playing as the Gaians.

Now I know that tech cost does vary, according to the circumstances in which you acquire the tech. I think that I've observed the following to date:

Get a tech from an artifact - your tech cost rises immediately.

Get a tech from another faction (gift or theft) your current tech cost doesn't rise.

Does anyone have any information about how tech cost is calculated? It seems to vary such a lot - suddenly, I'm paying the same price per tech as my pact mates and yet for the last two or three techs, the cost was ridiculously low. There must be a factor I've missed, but I'm not sure what it is.

Anyone got any ideas?

[mark13]

I'm not by any means certain about this, but as far as I have been able to tell, tech cost is purely related to the amount of tech you had accumulated before you started to research the current tech.

i.e.

If you start researching Tech C, then acquire tech D and E, the cost for tech C stays the same. However, when you then research Tech F, the cost rises dramatically, taking into account D and E.

This is, pretty much, a stab in the dark based upon previous observations, and if anybody has any contradictory evidence, I, for one, would be very interested indeed.

[Chowlett]

Misotu,

Check the thread I just bumped. I have a spooky kind of feeling that I should have checked authors first though... I seem to recall a Misotu post in it...

[Misotu]

LOL. No, it's OK, there was no Misotu post

Made interesting reading, too. Then I decided to go back to the MP game that raised the query in the first place, to plot exactly what happened with tech cost ...

and I've deleted the historic game files to save space.

gnash gnash

Dammit.

I'm going to have to wait until I have another game that reproduces the circumstances. All I can say for sure, is that I discovered that getting a tech from an artifact did *not* increase my tech cost, even though it was the tech I was currently researching.

Thanks for the help anyway ...

[MariOne]

Funny how the same questions keep coming from time to time, or should I say over and over again .
Tech cost issue addressed twice both here and earlier in ACOL.

Here's the thread Chowlett bumped
Tech cost

There an ACOl thread was mentioned:
How does the game calculate research points?
but that was not the original one, which was
A couple of technical questions about research

my comments in the bumped thread, here my thanks for making me remember the ACOL one

[Misotu]

Thanks for this - it really helped. I'm not so bothered about the effects of different factions/different SE settings because I'm comfortable with those variations. What has confused me recently is one specific scenario on ACOL where I am pacted with 3 other human players against a very strong AI.

So, there's a lot of tech swapping going on. Now, I tend to hold off exchanging techs as long as possible to keep the cost of each tech down. As a result, my tech cost was very low indeed by comparison with that of the other human players.

Then, I accepted a load of tech - maybe 5 in one turn. My tech cost remained the same. Fine, that was what I expected. Finished researching a tech, expected the cost to jump but ... no! My tech cost remained low for another discovery and *then* jumped dramatically. Now, my tech cost is the same as the cost for the other players.

In terms of this particular aspect of research point calculation, I've just extracted this one snippet from one of the threads you posted the links for. Quote is by MichaelTheGreat:

**********************************************

There are a lot of inconsistencies. UoP, sometimes, at some levels, has an undocumented bonus - requiring less labs per tech. ***It also seems that when you trade or link multiple artifacts at once, the total number of labs required for the next tech increases less than if you get them one at a time.***

I have noticed repeatedly, in fact almost all the time, that there is up to a 35% spread in labs required for the next tech when factions have the exact same techs. The difference can vary anywhere within that 35% range, so there would seem to be a number of undocumented "features" in the calculation. Truly shocking.

********************************

This fits with what I think I observed in my recent game. First time it has ever happened to me, and it took me rather by surprise.

Thanks for the assistance, much appreciated - at least I know I'm not seeing things

[Simpson II]

Hum.. I just want to clear up some issues about research cost and research SE. All of this refers to Smac, though I doubt it will have changed for X.

First, the University does not get cheaper research costs thanks to it's +2 research. However, the techs you get at the start don't count towards your overall tech cost like other techs. Try giving a faction a tech, 50 ability so you get most of the tech tree right at the start. Their research cost will be the same as any ordinary faction. So this bonus should show up for e.g. pirates as well.
Secondly, all factions with negative research SE in their faction files get an increased research cost. The magnitude of the change is constant regardless of the size of the negative. For the first two techs the cost is doubled, then it goes down to +1/2, then +1/3, +1/4, etc. I'm sure you see the pattern. I don't think this happens when you use a SE choice with -2 research. It would be pretty small by that stage of a game, anyway.

It's interesting that cashing in artefacts increases your research rate less quickly than using them one at a time! I'll have to look at that this weekend. Likewise for trading. I've noticed before that the no. 1 has much increased tech costs even in MP. It sounds very silly to me, that was presumable meant to slow a human who is running away with the game vs the AI.. it's not terribly fair in MP.

------
S2, insane builder.

[Misotu]

Simpson II - are you sure that the increased tech cost for no 1 applies to MP? I've been taking a look at some of my games, and I'm not seeing it. I think it's actually possible that the game distinguishes between human and AI ???

Well OK. I'm not swearing to it. But if there is a penalty for being the tech leader in MP games, I'm sure it's nothing like as significant as the tech penalty in an SP game.

[Simpson II]

Misotu,

Yes, I think it does. In my game with Paul where he is way ahead of me in the powergraph, I remember that even when I was (momentarily )ahead in tech his research costs were higher than mine. He is the Uni and I am the Gaians, so certainly I wasn't getting any faction-specific research cost cuts above and beyond his! At a guess, this isn't happening in your MP games because most were over very early. I first noticed this effect sometime after 2200. I know that in Civ at least, the designers liked to make certain evening-out factors kick in only after a certain date, so this could well be what is happening. I'll check some of my old save games when I get back to my own PC this weekend, but I'm pretty sure of it.

Mind you, the costs weren't so low as the AI's, but then the AI gets massive cheats with regards to labs, regardless of whether you are ahead or behind. I was the one making use of that..

[Misotu]

That's an interesting thought about it kicking in after a certain MY, because most of my MP games don't reach 2200. You've reminded me about one game against a friend that went on until over 2300. Unfortunately I only have a few save games left, but I went back to take a look.

In this game, the Angels were in the lead position on all measures - pop, tech, milit etc

The Angels' tech cost increased consistently by 44 or 45 labs each breakthrough (rounding explains slight variations)

The Drones' lab cost did not increase consistently. It varied between 39 and 41 labs per tech gained, so it certainly seems that the game favours factions who are not in the #1 position.

But what really clinched it was the data for three particular years:

2302
Angels' lab cost/techs discovered: 3410/77
Drones: 2498/60

2303
Angels: 3454/80 (3 new techs, 2 from Uni Translator)
Drones: 2471!/61 (one new tech)

2304
Angels: 3587/81 (cost of Uni Trans techs now factored in)
Drones: 2511/62 (one new tech)

Yes indeedy. The tech cost for the Drones actually FELL by 13 labs when the Angels discovered the Universal Translator and made three techs in one turn, even though the Drones themselves discovered a tech in that turn and should have experienced an increase in cost.

You'll note that, although the Angels' research rate only increased by the standard 44 points in 2303, their 2304 tech cost was recalculated to factor in the Universal Translator techs.

Unfortunately, none of the AIs discovered new techs after I got the Uni Translator, so I was unable to compare the game's treatment of AI and Human factions.

What puzzles me, though, is the point of this. When you're talking a cost of 3000+ labs per tech, then it really doesn't matter much if the Drones' tech cost rises by 39 labs while mine rises by 45. It simply isn't significant late-game. But *early* game would be a different story. I wonder ...

[mark13]

Oh well, at least it ties in with the common Firaxis penchant of creating ridiculous formulae for totally insignificant components, presumably for SMACaholics to unravel two years later. Keeps the interest up I suppose.....

[Fitz]

Misotu, that makes sense to me. IIRC, there is documentation that bases several formulas in the datalinks (not tech sots though) that are calculated based on Total Techs discovered. I always read that as total techs discovered by ANY faction. If the tech cost formula is something like:

(TechDiscoveredbyFaction)/(TotalTechsInGame * TotalTechsDiscoveredInGame)

then the cost would go down if lots of techs were discovered, but not by you.

I'm not sure that formula says exactly what I want it to, but I'm just trying to get a point across. I'm kind of suprised that no one has though that not only is the formula based on the ratio of techs discovered to techs available, but also on techs discovered to techs discovered by everyone.

Does that make any sense?

[Misotu]

Yes. But I have no idea about the true formula. Maths was never my strong point.

I'm pretty keen to discover what the early MP game effect is though ... I need to collect a few more save games and investigate.

[Basil]

According to the Prima guide, a simplified version of the formula is as follows:

Research Cost = [11 + (4*Difficulty) + (1.2*Techs) - (Turns/8) - (.2*MostTechs)] * Techs * Worldsize

Where Difficulty ranges from 0 (Citizen) to 5 (Transcend)
Techs is the number of techs you have
Turns is the number of turns that have passed
MostTechs is the number of techs the faction with the most techs has
Worldsize varies from .6 (tiny) to 1.6 (huge), etc.

It notes that there are additional modifications due to SE research settings. Also, if you are a lot behind the tech leader you will find research even cheaper than is indicated by the formula.

[mark13]

A simplified version of the formula?!?!?!?!

Yet another of the well-known Firzxis formulae. However, is there, as stated by various previous posts, an advantage for a positive research SE setting, other than the +20% accumulated?