Does giving away tech help your science?

Rufus T. Firefly · April 9, 2001, 05:07

Maybe I've been misreading things, but I think I keep seeing references to giving away techs "helping" your research rate. Now, I know it will help make the AI worshipful, and it will slow AI research down, but I can't see how it helps my research. In fact, lately I've been conducting this experiment: after building the UN (and saving the game), I set my science to zero, change my scientists to tax collectors, note the "discoveries every 'x' turns" number (which is the same as beakers needed for next advance, right?), identify the weakest civ, and give them everything I can -- usually 15-20 techs. Then I check "discoveries every 'x' turns" -- and it's the same. Then I play a few turns and check it again -- and it's still the same. So aside from making some puny civ worshipful, and slowing down still further the game's slowest researcher, what exactly did I accomplish?

------------------
Dig trenches, with our men being killed off like flies? There isn't time to dig trenches. We'll have to buy them ready made. Here, run out and get some trenches.
-- Rufus T. Firefly, the original rush-builder

Scouse Gits · April 9, 2001, 08:03

You may be trying this a little late in the game, but certainly in the early to middle game generosity of tech gifts to keep all the AIs more or less comparable with each other and quite close to you definitely reduces your beaker cost by 5 to 10% - some of the exponents have reported 15 to 20%, but I haven't seen this personally -- there is as yet no consensus on what exactly is required, but some tests by ?vik? or ?Tom DeMille? in the OCC fortnight threads suggested quite strongly that it was the differential between the sixth civ and yourself that made the difference -- this suggests that once two civs have been eliminated - you're on your own ...

------------------
Scouse Git[1] -- git1@scousers.net

"Staring at your screen in horror and disbelief when you open a saved game is one of the fun things of a succession game " - Hueij
"The Great Library must be built!"
"A short cut has to be challenging,
were it not so it would be 'the way'." - Paul Craven

Rufus T. Firefly · April 10, 2001, 00:25

Well, now I'm intrigued and want to test this. Tell me if this sounds like an adequate test;

1) Begin a game at a good OCC site (I have one saved). Cheat a lot of gold into existence.
2) Discover 20 techs, while rushbuilding Marco Polo and any improvement that improves science; DON'T trade techs, pop goody huts, or send out caravans; record beakers needed for each tech.
3) At 20 techs discovered, save. Discover 5 more techs; record beaker count.
4) Go back to saved game, and use Marco to identify the weakest civ. Give them my many techs. Discover 5 more techs and record beaker count.
5) Compare

Does this sound like a reasonable experiment, or just an unreasonable waste of my time? If the former, I can probably do it tomorrow.

------------------
Dig trenches, with our men being killed off like flies? There isn't time to dig trenches. We'll have to buy them ready made. Here, run out and get some trenches.
-- Rufus T. Firefly, the original rush-builder

Scouse Gits · April 10, 2001, 03:20

Can you can go back to a save in a previous game? It might be quicker and you could avoid using Cheat Mode, which may affect you tests.
---------
SG(2)
[This message has been edited by Scouse Gits (edited April 10, 2001).]

Scouse Gits · April 10, 2001, 05:05

remember that if the proposition above is valid it is the relative position to the 6th civ (not the 7th) that is, perhaps, key ...

------------------
Scouse Git[1] -- git1@scousers.net

"Staring at your screen in horror and disbelief when you open a saved game is one of the fun things of a succession game " - Hueij
"The Great Library must be built!"
"A short cut has to be challenging,
were it not so it would be 'the way'." - Paul Craven

Rufus T. Firefly · April 10, 2001, 05:14

quote:

Originally posted by Scouse Gits on 04-10-2001 03:20 AM
Can you can go back to a save in a previous game? It might be quicker and you could avoid using Cheat Mode, which may affect you tests.
---------
SG(2)
[This message has been edited by Scouse Gits (edited April 10, 2001).]

I'll look, but the problem is that I'm pretty lame about exploring, so early in the game I rarely have contact with a lot of civs. Here's a question: how late is too late? Measured in # of techs, not years, I mean.

------------------
Dig trenches, with our men being killed off like flies? There isn't time to dig trenches. We'll have to buy them ready made. Here, run out and get some trenches.
-- Rufus T. Firefly, the original rush-builder

East Street Trader · April 10, 2001, 06:18

Rufus

If perchance you have a save of the game you tried the give away in, it would be interesting to repeat the give away to the second weakest (assuming there was a full complement of A1 civs still alive).

It would test the proposition that part, or all, of the comparison which the game makes is between the human player's position and the 6th civ's position.

Scouse Gits · April 10, 2001, 08:18

Yes, EST, but ...

this does get complicated doesn't it -- if you gift civs 6 & 7 with an Encarta or two (and the hypothesis under test is true) your potential gain depends upon the relative positions originally of the 6th and 4th AI - the 4th having just become the critical 6th if you see what I mean ...

I think the 'best' test would be to measure the effect of gifting all to 1, 2, 3, 4, 5 and 6 civs preferably in ascending order of original knowledge and comparing your beaker costs -- only problem here is that in my experience the d*mn AI won't play the game and promptly refuses to speak to you just when you want to give him everything!!!

------------------
Scouse Git[1] -- git1@scousers.net

"Staring at your screen in horror and disbelief when you open a saved game is one of the fun things of a succession game " - Hueij
"The Great Library must be built!"
"A short cut has to be challenging,
were it not so it would be 'the way'." - Paul Craven

samson · April 10, 2001, 15:40

Dear Mr. Firefly,

The number of beakers required for a civ to research a given tech
is dependent on its number of acquired techs (starting techs don't
count) and the civ's position relative to the average of all civs'
acquired techs. If the civ has fewer acquired techs than the average
then that civ pays the minimum for its next tech. Otherwise it pays more.

For strategic purposes it is important to know the minimum beaker count
for the tech number you are researching. If you are paying more than
the minimum, you may be able to reduce your cost by giving away techs.
Since your cost is relative to the average of all civs, the more civs you
are in contact with, the easier it is to lower your cost.
In a seven-civ game I find I need at least three other civs in order to
reduce my cost in the early game, later on you need all of them.

Also, having a lot of starting techs is an enormous advantage.
They don't count towards your ranking, but when you give them away
to other civs they count towards theirs. This allows you to quickly
get 'behind' and reduce your cost in the early game.
On the other hand, if you have NO starting techs you will only pay the
minimum cost when you truly are behind. If you get ahead in the
research race you will be paying a premium even if you give everything
away.

Here are the minimum beaker counts for the first 19 acquired techs.
(Acquired techs are those that you research, obtain in trade,
or find in a hut).

These numbers are for Deity, 7 Civs, Medium Map (v2.42)

Tech# x Base = Minimum Cost
----- ---- ------------

1 10 10
2 9 18
3 10 30
4 10 40
5 11 55
6 11 66
7 12 84
8 12 96
9 12 108
10 12 120
11 12 132
12 13 156
13 13 169
14 14 196
15 14 210
16 15 240
17 15 255
18 15 270
19 15 285

From tech # 20 on up, the base for the minimum cost is 24. For example,
the minimum cost for tech # 20 is 480. The min cost for # 30 is 720.

You pay the mimimum if you are below average. If you are average
or above, a penalty is added to the base. This penalty increases sharply
the farther you are ahead. For example, at tech # 18, the penalty
can be as high as +8, meaning that you will pay 18 x 23 = 414.

At the start of each new research, I check the tech cost against the
minimum and if I'm paying too much, I give techs to the AI, one at a time
until my cost drops to the min. Actually, you can do this any time up to
the turn on which the research completes.

Remember too, that giving techs to other civs will slow their research.
This may be desirable or not, depending on your strategic objectives.

samson

[This message has been edited by samson (edited April 10, 2001).]

Rufus T. Firefly · April 11, 2001, 02:54

Samson: My specific question regards how your "beakers needed" count changes, if at all, if you giv techs to another civ. I actually have some very interesting news about that, which I will post shortly.

Meanwhile, while much of what you say is true, I would dispute the numbers in your chart (though I play ToT, not 2.42, so that may explain it). I've conducted the same test myself, through the first 20 techs acquired, and find that the multiplier for 2-6 is 11, not 9 and 10. Moreover, the multiplier for a specific tech could vary slightly from test to test; in one test, the multiplier for 13 techs was 13, while in another it was 14. Since I never traded or gifted techs, and never popped a goody hut, it's not at all clear what caused such variations.

------------------
Dig trenches, with our men being killed off like flies? There isn't time to dig trenches. We'll have to buy them ready made. Here, run out and get some trenches.
-- Rufus T. Firefly, the original rush-builder

Rufus T. Firefly · April 11, 2001, 04:05

Well, now I feel like a real Apolyton: I'm about to challenge conventional wisdom and contribute a bit of knowledge.

NEWS: The "sixth civ" hypothesis regarding gifting techs is incorrect, except in cases of coincidence, and the potential change in beakers needed can be much greater than the 5-10% reported!

Last night I opened 2 saved games and conducted some tests. Both games were ToT, Diety/7/raging. The first game was a large map, the second a medium map.

Test 1
In the first game, it was 1905, I was researching Fusion, and the next discovery was going to require a total of 2400 beakers. The game stats for all civs were:

Civilization	# of Techs Possessed	Position on Power Graph	Techs Received from Me	Change in # of Beakers I Need
Vikings	54	4	26	none
Babylonians	56	7	25	none
Aztecs	71	6	11	none
Russians	67	3	15	none
Chinese	73	5	10	none
Carthaginians	73	1	11	-300 (-12.5%)
Sioux (Me)	78	2	N/A	N/A

Test 1a: Give the number of techs listed in column 4, moving through the civs as listed on the table. I started with the Vikings, because they had the least # of techs -- no change. Then the Babs, because they were weakest overall -- no change. Then the Aztecs, because by that time they were #6 in terms of number of techs and #6 on the power graph -- no change. And no change from then on until I gifted the Carthaginians.
Test 1b: Give techs only to the Carthaginians. Same change as above.
Test 1c: Give techs only to the Chinese, who have the same number of techs as Carthage. No change.
Test 1d: Give techs to Carthage one at a time, checking beaker-needed count after each tech. Result: 1-2 techs given = no change; 3-6 techs given = change -150 beakers (-6.25%); 7+ techs given = change as in table above.
Conclusion: Gifting techs to the sixth civ was irrelevant. The only civ that mattered was Carthage, #1 on the power graph and tied for #2 in techs. Now the question is: is this because Carthage was Supreme, or because they were ahead of me on the power graph, or because they were right behind me in tech count? Read on!

Test 2
In the second game, an OCC game, it was 1894, I was researching Miniaturization, and the next discovery was going to require a total of 2052 beakers. Game stats:

Civilization	# of Techs Possessed	Position on Power Graph	Techs Received from Me	Change in # of Beakers I Need
Germans	52	1	10	none
Greeks	55	2	8	none
Spanish	53	3	N/A	N/A
Indians	44	4	N/A	N/A
Persians	47	5	N/A	N/A
Russians	33	6	28	-684 (-33.33%)
Lydians (Me)	60	7	N/A	N/A

Test 2a: Give techs to the Germans, because they are supreme; no change.
Test 2b: Give techs to the Greeks, because they are the next civ behind me in number of techs. No change.
Test 2c: Give techs to the Russians, because they are the next civ ahead on me on the power graph. Voila!
Test 2d: Give techs to Russians one at a time, checking beaker-needed count after each tech. Result:
1-2 techs given = no change
3-5 techs given = -57 beakers (-2.78%)
6-8 techs given = -171 beakers (-8.33%)
9-11 techs given = -228 beakers (-11.11%)
12-14 techs given = -285 beakers (-13.89%)
15-17 techs given = -342 beakers (-16.67%)
18-20 techs given = -456 beakers (-22.22%)
21-23 techs given = -513 beakers (-25%)
24-26 techs given = -570 beakers (-27.78%)
27+ techs given = -684 beakers (-33.33%)

Conclusions: (1) The only time giving techs away would seem to matter is when you can give them to the next civ above you on the power graph. This would generally mean the 6th civ in OCC, but not elsewhere. It would also seem to imply that, if you are Supreme, you gain nothing by giving away techs. (2) The more techs you give, the more you improve, but you must give at least 3 techs to effect any change.
Unanswered questions: What if there are fewer than 7 civs remaining? What if the techs given are extorted from you instead of given? What if techs are traded instead of given? The testing continues...

------------------
Dig trenches, with our men being killed off like flies? There isn't time to dig trenches. We'll have to buy them ready made. Here, run out and get some trenches.
-- Rufus T. Firefly, the original rush-builder
[This message has been edited by Rufus T. Firefly (edited April 11, 2001).]

Scouse Gits · April 11, 2001, 07:25

Good Catch

Really good work Rufus - thank you very much - and you were right - the 6th Civ hypothesis did originate in OCC ...
the fog of war recedes just a little ...

------------------
Scouse Git[1] -- git1@scousers.net

"Staring at your screen in horror and disbelief when you open a saved game is one of the fun things of a succession game " - Hueij
"The Great Library must be built!"
"A short cut has to be challenging,
were it not so it would be 'the way'." - Paul Craven

Rufus T. Firefly · April 11, 2001, 08:24

Thanks SG! One more thing (for now): In both games, I started with 4 extra techs; in game 1, then, Fusion was the 75th tech I was seeking to acquire, and in game 2 Miniaturization was the 57th tech I was seeking to acquire. Thus, the amount the beakers are reduced by appears to be a multiple of the tech number of the current tech under research, which means what's actually being reduced is the multiplier. In Game 1, then, giving away techs drops the beaker multiplier from 32 (32x75=2400) to 30 to 28. In Game 2, the multiplier is dropped from 36 (36x57=2052) to 34, then 33, 32, 31, 30, 29, 28, 27, 26, and finally 24. This seems too sketchy to draw further conclusions from, but it is interesting that in each case the first and last drops were 2xtech number, and the intervening drops (for game 2) were 1x. What's not clear is why, in game 1, the drops occurred at 3 and 7 only, while in game 2 they occurred at 3, 6, 9 etc. Hmm....
------------------
Dig trenches, with our men being killed off like flies? There isn't time to dig trenches. We'll have to buy them ready made. Here, run out and get some trenches.
-- Rufus T. Firefly, the original rush-builder
[This message has been edited by Rufus T. Firefly (edited April 11, 2001).]

samson · April 11, 2001, 10:48

Rufus,

The reason my beaker chart differs from yours is that mine is for
the MINIMUM beakers required. Your multiplier is higher for some
of the early techs because your civ was not getting the lowest rate.
That is also what caused the variations you saw. You were paying
different rates because of different rankings.
The importance of knowing the minimum cost for a specific tech being
researched is to know how much you are overpaying for that tech.

There are also MAXIMUM costs that you will pay if you are the top-ranked
civ. The maximum cost gets higher relative to the minimum as the tech#
increases.

I knew from looking at your numbers that both of your games started
with 4 techs because of the multipliers involved. In the first game,
for the 75th tech you were paying 2400 which is +600 over the minimum
of 1800 (75*24). However, you were only able to lower your cost by 300.
In the second game, as the lowest power-ranked civ, you were able to
get down to the base 24 multiplier and pay the minimum.

I think you are onto something with the powergraph connection.
My testing was mainly with OCC games and my assumption of an 'average'
tech ranking appears to be incorrect.

Since tech is part of the powergraph equation, perhaps you can lower
your powergraph ranking in the first game by giving techs to
the other civs. You won't get an immediate beaker reduction from this
but when the civs are re-ranked after the next oedo year,
you should drop and, maybe, so will your beaker requirements.

The two things required for a Minimum Cost of Research Strategy are to know
what the minimum IS (and when you are paying too much) and how to lower it.
Your powergraph idea looks promising for refining the second aspect.

Thanks for sharing.

samson

P.S.
The drop in beaker cost by 3 techs given might be related to the fact
that 2.67 techs are one powergraph point.

[This message has been edited by samson (edited April 11, 2001).]

East Street Trader · April 11, 2001, 10:53

Magnificent.

Just magnificent.

Well caught indeed.

And explained with great clarity, to boot.

The only question remaining is whether the Firefly phenomenon is the more worthy of this year's Apolyton oscar or oedo's xxxo.

In any event we now have two MAGNIFICENT nominations.

Rufus T. Firefly · April 12, 2001, 08:07

Something else I just noticed: the numbers generated here contradict Caesar's Beaker Numbers in the Great Library; my beaker count at 57 and at 75 were lower than Caesar's, even before I started lowering them further with tech giveaways. Caesar's numbers at levels 3-5 are also lower than samson's. There's clearly much work to be done.

samson · April 12, 2001, 09:25

Rufus,

You're right.
I just played a game where I got even lower beaker numbers on some
of the earliest techs. More work to do on this.

Regarding your theory about powergraph position ... I toyed with it a bit.
In one instance, I was the top civ on the graph and was able to reduce
my beaker count by giving techs to the lowest civ on the graph.

Another observation ... sometimes giving techs results in no immediate reduction,
but the beaker requirements decrease at the start of the next turn.
In one situation I found that I could reduce the beaker count
on the next turn by giving techs to any combination of four civs.
Giving to three of fewer civs didn't effect it, but four did.

I'm feeling more confused about this. Except for the one constant ...
giving techs does help your science.

samson

Rufus T. Firefly · April 12, 2001, 14:22

quote:

Originally posted by samson on 04-12-2001 09:25 AM
Another observation ... sometimes giving techs results in no immediate reduction,
but the beaker requirements decrease at the start of the next turn.
In one situation I found that I could reduce the beaker count
on the next turn by giving techs to any combination of four civs.
Giving to three of fewer civs didn't effect it, but four did.
samson

Are you sure the fourth civ wasn't simply, coincidentally, the appropriate power-graph civ? I ask because, in my first example above, I gave a boatload of techs to 5 civs without producing any effect. Just wondering.

------------------
Dig trenches, with our men being killed off like flies? There isn't time to dig trenches. We'll have to buy them ready made. Here, run out and get some trenches.
-- Rufus T. Firefly, the original rush-builder

Mixam · April 14, 2001, 07:40

How exactly can you check the power graph? I can't seem to find it except when i quit a game. I have the classic verison. Thanks

------------------
Shhh... Just Take It!
Shhh... No-one Has To Know!

Rufus T. Firefly · April 14, 2001, 09:25

Well, the easiest way is to save and retire, but you can do it if you have contact with every civ or all but one civ (where the info about all the other civs would allow you to infer info about the one that's missing). You just have to visit them, one by one; in their antechambers, to the left of their pictures, will be icons of 0-6 weapons (arrows, rifles, missiles, and something for early civs that always look to me like wooden spoons). The number of weapons corresponds with the position on the power graph; 6 rifles = first, no rifles = last (with 7 civs; with fewer, it still starts at 6 icons but only goes down as far as necessary). You can always find your position on the power graph by consulting the foreign minister, who will tell you, "Sire, our power is ____...". There are seven possible adjectives he can use, and they correspond to the seven possible places on the power graph: supreme, mighty, strong, mediocre, inadequate, weak, and pathetic.

------------------
Dig trenches, with our men being killed off like flies? There isn't time to dig trenches. We'll have to buy them ready made. Here, run out and get some trenches.
-- Rufus T. Firefly, the original rush-builder

samson · April 17, 2001, 00:20

Rufus,

Sorry to take so long in replying to your question.
Yes, I'm sure. I tried giving techs to each civ alone, no effect.
I tried various two and three civ combinations, no effect.
Any combination of four civs worked, as did five civs.
(At that point in the game I had no contact with the sixth civ).

I think it's important to reiterate that nothing I did reduced the tech multiplier immediately.
It only reduced it at the start of the next turn, which brings up another point.
The tech multiplier is sometimes increased. This only seems to happen at the start of a turn.
Lately I've been noting the turn year when this occurrs.
So far it has always been one turn after an 'oedo year'. I read somewhere
that Powergraph rankings are recalculated in oedo years. So this may
have something to do with that.

When the tech multiplier is increased at the start of a turn, nothing I have done
has been able to reduce it immediately. But giving techs away can reduce
the multiplier at the start of the next turn.
Sometimes, even when there has been no start-of-turn increase, giving techs
only reduces the multiplier on the next turn.

I suspect that there are two mechanisms involved in Beaker Cost assignment.
One is cyclical, every four turns, the other is an adjustment that is made
when techs are aquired, but this might be done only in non-oedo years.

happy researching,

samson

SlowThinker · April 17, 2001, 18:28

{advances}
{beakers count}
{}{SlowThinker}{end1}

------------------------------
This is a post with keywords. See The Great Library: a hierarchical structure" thread.
This thread has got to the top of the forum because of this post. It may be a very old thread.

Rufus T. Firefly · April 18, 2001, 03:22

I'm beginning to feel like those researchers a few years back who went public with their "discovery" of fusion.

It sounds like samson hasn't been able to duplicate my experience with giving to the next-highest civ on the power graph. Moreover, I recently failed to duplicate it myself: I'd never tried the give-away early in the game, so in my most recent game I did -- and in spite of being 5 techs ahead of the next nearest civ, and in spite of giving 11 techs to the next-highest civ, my tech rate didn't budge (it was at 224 while researching tech #14; I should have been able to drop it to 196). On the other hand, I haven't to been able to duplicate samson's experience with lowering tech rate by giving to multiple civs (though I haven't tried checking the rate at the beginning of the next turn), and my successes (see post with tables above) happened instantly and don't correspond in any way with odeo years. The more we know, the more we don't know.

Still, at least we know a couple of things: the sixth civ hypothesis is wrong, the beaker count in the Great Library is inaccurate at the upper levels, and the "penalty" for being ahead in techs can be greater than +8. A couple of things are suggestive, if by no means conclusive: I have yet to get the tech rate at higher levels down below the hypothesized base of 24 (see samson's first post to this thread), and I have yet to succeed in lowering my tech rate once a civ has been eliminated.

Ideas, anyone?

------------------
Dig trenches, with our men being killed off like flies? There isn't time to dig trenches. We'll have to buy them ready made. Here, run out and get some trenches.
-- Rufus T. Firefly, the original rush-builder
[This message has been edited by Rufus T. Firefly (edited April 18, 2001).]

samson · April 18, 2001, 09:28

Rufus,

The immediate reduction of the multiplier does seem to be triggered by giving techs
to one specific civ. However, what the significance of that civ is, or its relationship
to the human player is uncertain. In my last game I had two instances that refute
your hypothesis. In one case I was the lowest civ on the PG and gave techs
to the #4 civ and got a beaker reduction. In the other case, I was #5 and gave to #2.
The one-turn-delayed reduction may work differently, I only recall seeing it
after giving to multiple civs.

Increases in the multiplier at the beginning of a turn are quite troublesome.
If it occurs on the turn in which you expect to complete research, you can lose a turn.
Very annoying, especially when the beaker count reduces again the next turn.
One time I had a size 4 city starve down to size 3, and the tech multiplier INCREASED.
I'm talking about the required beaker count, not just the # of turns.
I replayed that piece several times and it was definitely the reduction of my population
that triggered the required beaker count increase! What the heck?
There's some twitchy stuff going on in there.

samson

debeest · April 19, 2001, 00:26

The data in the Great Library, posted by Caesar the Great in "Yet Another List of Data," are clearly not generalizable. He generated them by cheat-adding the advances at the beginning of a game, which meant that none of the other civs gained any, and he got farther and farther ahead. Those data may actually represent the MAXIMUM penalties that samson refers to. Interesting, but not all that useful in figuring out how much science you'll actually need or how to keep your science costs down.
[This message has been edited by debeest (edited April 19, 2001).]

Mixam · April 19, 2001, 15:56

The data in "yet another list of data" written by Caesar in the GL seems to have been done in cheat mode on the first turn of the game as i have duplicated this by testing. Of note is that you will get the same results no matter how many techs the other civs have. Also even if you kill all the civs off using the cheat menu the results stay the same. At least this proves that there is no change in beakers required until the next turn no matter what. This may not be true on Oedo years. I think i recall hearing some people say they have had their beaker requirement drop the same turn they gave away techs. I am trying to test whether or not this happens on Oedo years. I will make my results public once I am finished probably within the week.