Micromanagement is alive and well in Civ 4!

cabert said:
You did not, but saying you get a bonus because you research a tech AIs already know implies you research it, isn't it so?
If you don't, you get no bonus at all!
Click to expand...

Let me clarify then. To get to an interesting tech that no AI will have, you usually need some prerequisite techs first. Until you do get the interesting tech, you may have nothing to trade with the AIs, which means that you'll have to research the prerequisites yourself.

Assume that you're going after tech C with prerequisites A and B. Assume that A and B are not interesting except for getting to C, therefore you don't care when you get A and B, as long as you get C as soon as possible.

Using the binary science rate, you can run at 0% science until you have enough money in bank to run at 100% science all the way to C. During the time you're at 0% science, some AIs may reach A and maybe even B before you do. All the better, as this means getting a bonus on them and getting them for cheaper when you start researching them yourself!
 
I like to do this during peacetime when there's access to stone/marble. Just build useless wonders like the Hagia Sophia to near-completion, cancel and forget about it. Once the AI completes it, you get the gold return for the incomplete Hagia Sophia. Since you're building at double production due to stone/marble, you'll get a lot more gold return than by producing wealth.
 
Interesting article and humerous discussion (well, almost). My basic thought when I started to read this was that the rounding down really hurts when it splits your commerce into beakers and gold. However, after looking at the game and reading further, am I right in that this is NOT the case. It only hurts if you consider science / gold multipliers (libraries, banks, etc)?

See below for an example where I have 19 commerce that is split into 9 beakers and 10 gold. So, with no multipliers, I get full value for my 19 commerce. Running 5 turns of 100% science and 5 turns of 100% gold will give me exactly 95 beakers and 95 gold. Running 10 turns of 50%/50% will give me 90 beakers and 100 gold. I don't lose anything, it is just the distribution isn't exactly 50/50.

Is this correct or am I missing something else?



Obviously, throwing in a library (+25% beakers) will mean that I should try to set my base beaker output to a multiple of 4 for maximum benefit. If that is so, then the strength of this approach is not 100% / 0% (ie binary) but making sure that your base is a multiple of 4. Again - have I got this right?

That is, apart from this item ...
Zombie69 said:
Using the binary science rate, you can run at 0% science until you have enough money in bank to run at 100% science all the way to C. During the time you're at 0% science, some AIs may reach A and maybe even B before you do. All the better, as this means getting a bonus on them and getting them for cheaper when you start researching them yourself!
Click to expand...
 
You are right that the tech rate percent does not lose commerce, any fractions will be recovered in gold - this was not the point being made in this thread. As you realized bonus buildings such as libraries cause fractions which do result in loss and this is the subject at hand. One way to limit this loss is to keep commerce to certain multiples but this is not always achievable or desirable. The binary science rate will limit losses to these fractions without having to achieve precise multiples.
Your example has no bonus buildings so will not on its own benefit from binary rates. (Note there are certain hidden bonuses in the game even in these simple situations depending on the tech and contacts). Also your example appears to have no maintenance cost in which case the most efficient research is 100% always.
The binary rate is for situations when you cannot run continuously at 100% - then running 0% then 100% is better than a flat percentage because of saved fractions from libraries etc.
 
junior7 said:
The binary science rate will limit losses to these fractions without having to achieve precise multiples.
Click to expand...
I don't think that this is correct. Consider the following where I am only looking at beakers, ignoring maintance and tech bonuses.

  • 0% beaker multiple - no loss of commerce as any fraction goes to gold
  • 25% beaker multiple - it is possible to lose fractional beakers at all science rates except 0% and 80%. You lose beakers if you have 25 commerce @100% science - this gives 31.25 beakers which gets rounded down to 31. If you run at 80% science, then the formula is beakers = commerce * 0.8 * (1.25) = commerce.
  • 50% beaker multiple - it is possible to lose fractional beakers at all science rates except 0%
  • 75% beaker multiple - it is possible to lose fractional beakers at all science rates except 0%
  • 100% beaker multiple - it is possible to lose fractional beakers at all science rates except 0%, 50% and 100%
  • 150% beaker multiple - it is possible to lose fractional beakers at all science rates except 0%, 40% and 80%
It is possible to lose fractional beakers at all science with other science multipler percentages. It seems there are some special cases (25% multipliers, 100% multipliers, 150% multipliers, etc).

If you have a nation that is full of libraries and no other multipliers, then the optimal science setting would seem to be 0% and 80%. Still binary but flipping between these two. This will give no loss of fractional beakers. If you are still in a positive cash flow at 80%, then stuff the fractional beaker loss and run at 90% or 100% - faster research times.
 
You are misinterpreting, again. I did not say binary science rate removes losses from fractions, I said it limited losses to fractions ie. reduces it. You cannot ignore maintenance since this determines the maximum tech rate you can afford to run.
The beakers formula with a library is actually - floor(floor(commerce*rate)*125%). So any tech rate may lose or not fractional beakers from library eg. 23 commerce -
floor(23 * 90%) = 20 loses no fractions when library bonus applied
floor(23 * 80%) = 18 does lose fractions when library bonus applied
You cannot avoid fractions from the tech rate but, as was said, that is converted to gold. The binary rate is to limit the loss from science bonus (library etc.) fractions which also cannot be avoided and are real losses (not converted to anything else).
 
junior7 said:
The beakers formula with a library is actually - floor(floor(commerce*rate)*125%).
Click to expand...
Ahh - that changes it. I didn't have the internal floor in my calculations. Running through the usual beaker bonus, 100% does, generally, limit the fractional losses. There are some rare situations where it doesn't ...

25 commerce with 75% beaker bonus, running 100% loses 0.74 beakers while running 60% losses 0.25 beakers

Yes, I did say it was rare.

Now, accepting this binary science proposition - is it *worth* running this the whole game to minimize these lost fractions? I can see that it is in the early game (tech costs are low, commerce generation is low). Later in the game, both of these don't really count but you have lots of cities and these lost fractions are at a city level (is that right?). So, if you have 20 cities, you might get an extra 5-30 beakers per turn. I guess that will add up. I'll give it a try and see if I can MM my way through.
 
Glad to have you onboard!
Certainly, if you wish to run binary science rate, you should run it the entire game for maximum benefit. The lost fractions are at a city level so the maximum gain is 1 per city per turn which does add up. Happy MM'ing!
 
Roland Johansen said:
In my opinion, it is far more important to get some gold now, then to have my towns developed at the same time. In the beginning of the game, I tend to have a rather low research percentage because of all the high upkeep costs. Every coin of gold that I can spend to help my research along to the next building (courthouse, marketplace, etc.) that will help my empire is very valuable. So I tend to use the most valuable tiles first.
Click to expand...

I totally agree with you. The early game is more important than a perfectly timed simultaneous conversion to towns. Cottages should be developed into towns as quickly as possible. Waiting to turn them all into towns at the same time wastes a lot of potential beakers and commerce because towns are much more productive than hamlets and villages.

Re: the 100% - 0% strategy, I totally ignore gold reserves until I get universal suffrage. By that time the game is basically over anyway. I don't really want gold sitting around, because it means I could have had another tech more quickly.

Gold can always be picked up via pillaging and tech trading & giving away techs for cash, I don't run commerce at 100% unless I am saving up to buy something ASAP.
 
jar2574 said:
Re: the 100% - 0% strategy, I totally ignore gold reserves until I get universal suffrage. By that time the game is basically over anyway. I don't really want gold sitting around, because it means I could have had another tech more quickly.
Click to expand...

Actually, the whole point of binary science is to get techs quicker. By not using it, you're getting them later than you could!

Having gold in reserve is just a nice side effect.
 
Zombie69 said:
Actually, the whole point of binary science is to get techs quicker. By not using it, you're getting them later than you could!
Click to expand...

As others have pointed out, once you have libraries and other improvements, the binary method is not always optimal. The appropriate rate will vary depending on the improvements you've built in various cities, and the relative output of those cities.

Requiring each city to maintain multiplies of 4 in order to make the binary method work at all times is counterproductive, because the benefits of developing towns ASAP outweigh the marginal benefits of the binary method. For example, requiring any city with the potential of creating 15 to create only 12 per turn would be counterproductive. Better to have 15 with some waste than 12 without.
 
The binary science is actually designed to limit the loss to fractions from libraries etc. (although it can benefit before also) so you do not need to run commerce in multiples of 4 to benefit. Binary is not optimum in every single position but it nearly always is, whatever the bonuses from buildings etc. An example I tried for another thread -

Example - 6 cities with 22,18,14,9,6,4 commerce, largest 3 with libraries, largest 1 with market also, with 39 maintenance per turn, researching 1 prereq unknown tech:
flat 50% gives (with 39g per turn) 100 BB in 2 turns, binary gives 104 BB in 2 turns (binary wins)
(Note BB is bonus beakers taking into account hidden bonuses from prereqs and meeting civs with tech you are researching).
I tried changing tech prereqs, knowing civs, maintenance cost, commerce in 6 cities and binary always won with the numbers I tried.
 
jar2574 said:
As others have pointed out, once you have libraries and other improvements, the binary method is not always optimal. The appropriate rate will vary depending on the improvements you've built in various cities, and the relative output of those cities.

Requiring each city to maintain multiplies of 4 in order to make the binary method work at all times is counterproductive, because the benefits of developing towns ASAP outweigh the marginal benefits of the binary method. For example, requiring any city with the potential of creating 15 to create only 12 per turn would be counterproductive. Better to have 15 with some waste than 12 without.
Click to expand...

You don't have to use multiples of 4. Of course it helps, but it's not necessary. Also, with a 25% bonus, 4n+1 is still pretty good for limitting losses, and isn't hard to get even when trying to get your towns asap.

Even when not even trying at all to get multiples of 4, cases where binary science isn't optimal are extremely rare, especially once you get a few cities.
 
Zombie69 said:
You don't have to use multiples of 4. Of course it helps, but it's not necessary. Also, with a 25% bonus, 4n+1 is still pretty good for limitting losses, and isn't hard to get even when trying to get your towns asap.
Click to expand...
What is the definition of 'limiting losses'. My understanding is ...

base beakers = floor(commerce * science rate)
base culture = floor(commerce * culture rate)
base gold = commerce - base beakers - base culture

So any loss of beakers here just goes to gold and isn't really lost. I don't know the formula for how the pre-req tech bonus gets added so I will ignore this for the present (can someone supply the formula?).

beakers to research = floor(base beakers * city science multipliers)

The way I would define lost beakers is base beakers * city science multipliers less floor(base beakers * city science multipliers). This is the fraction of beakers to research that gets rounded away. Is that your definition too?
 
ruff_hi said:
The way I would define lost beakers is base beakers * city science multipliers less floor(base beakers * city science multipliers). This is the fraction of beakers to research that gets rounded away. Is that your definition too?
Click to expand...

Yes, that's it.
 
ruff_hi - See 'Technology Research Explained...' by Requies for the full research formula. Basically the prereq bonus, other than the 6 starting techs, gets 20% bonus for all the required techs (if any) plus 20% for each optional tech. You also get a bonus for knowing other civs who have tech you are researching (30% * known_civs / total_civs). These bonuses also produce fractions which are lost and binary science limits this.
 
Ok, looked at that post and found this formula ...

Beakers applied to Technology = FLOOR (FLOOR ((Total Base Beakers + 1) * KCwT modifer) * Prerequisites modifier)
Click to expand...

The Total Base Beakers (TBB) is the number of beakers generated by your cities. The two modifiers are 'known civs with tech' and the pre-requisite modifier. I can understand how binary science affects the TBB but I fail completely to see how binary science has any impact on the two modifiers and fractional beaker losses.

Thinking further :hmm: ... if you run at 0% for a few rounds and then 100% for the balance (binary science proposition 101), then you give the other civs a few more rounds to discover the tech and you get a bigger benefit from that modifier - ok, I can accept that. However, I don't see how it has any impact on the pre-requisite modifier.
 
WOW!

You are nuts! (In a good way) :goodjob:
 
You are right - Binary science does not gain from prereq bonus fractions. It can gain but so can flat rate, it depends on the numbers. I was thinking binary gains on the 0% turns but this generates prereq fractions due to the bonus beaker you always get.
As you say, binary mainly gains for Total Base Beakers plus other civs discovering tech while you are researching it.
 
junior7 said:
As you say, binary mainly gains for Total Base Beakers plus other civs discovering tech while you are researching it.
Click to expand...

Plus the opportunity value of deciding, during the 0% turns, that you changed your mind about what you want to research next. Plus the fact that you have more cash on hand, on average (because you spend some turns at 0% piling up cash, and then some turns at 100% spending it down), and sometimes this "temporary" excess cash is useful in making trades that you otherwise couldn't afford (since you aren't allowed to trade your way into debt).
 
You must log in or register to reply here.

Similar threads

Phosphoraptor
(7-09) Hospital and Medical Lab Rework
Replies
18
Views
466
rkkn
R
S
(7-20) Sabotage Production Spy Mission Redesign [Spy Update 9]
Replies
4
Views
222
AndreyK
A
AcaMetis
Nobles' Club 352: Joao II of Portugal
Replies
16
Views
755
antimony
A
Y
(7-26) Supply Cap Penalty rework
Replies
6
Views
275
Recursive
Recursive
Bibor
SMACX strategy update 2024
Replies
3
Views
273
Senethro
Senethro
Top Bottom