80% science stupid?

[jeremiahrounds]

Hi. This is mainly a question about is the slider lossy or a direct percentage.

My thesis is this. In an optimum game there are exactly two positions to use for science: 100% and 0%.

The idea is this: if your finishing a learn science period with 0 coin then the best way to handle that is to bank your entire empires cost for the time it will take to train it at 100%. You still finish the research in the exact same time however you pay your empires bills from your savings instead of your future money. That way if there is a war or an emergency you have between 0 and what ever the savings was giving you flexibility to buy things and upgrade units.


Real numbers. Lets say you can have your science at 80% and break even while researching something that cost 4000 science in 10 turns. That means you actually have 80 science units worth of bills every turn. So the idea is for one turn you do 0% and you bank 480 beakers worth of coin. Then you go 100% and loose 80 beakers worth of coin every turn. You can do that for 6 turns. Then you go back to 100%

Why? Its an easy form of micro management that doesnt slow you down at all that creates coin to do stuff with if it comes up. Consider your savings to be 0 coin while your maintaining a 80% research rate at break even. If you do it as 100% and 0% then while your still learning just as fast and still technically flat broke you actually manage to have some slush fund in the bank for the occasion where you decide that an emergency upgrade to a unit or something is in order.


The actual use of buildings doesnt matter because it works out. HOWEVER the caveat might be that if you have a building coming midresearch period that either ups your beakers or coin then you would want to weight your bill paying or researching accordingly.

For example if 3 turns from now 2 cities will finish universities then at that time you would want to put research at 0. Get some coin while the opportunity cost isnt as high and then after the universities pop switch to 100% research. The reason is the opportunity cost of every coin generated is higher after the universities pop while their value remains the same.

Like wise if you had banks coming and the slush fund to deal with it. You would want to 100% your research and go broke and pay yourself back after the banks pop.

Does this sort of stuff matter? Typically not. Something to think about though.



Then instead of just being at break even

 

[Lord Chambers]

Someone made a post about "binary science" a couple weeks ago, and made two of the points you made.

Additionally, I recall that the cost of techs drop as more civs discover them, therefore by limiting your research to quick bursts you maximize the chance that you will benefit from other civs aquiring it while you bank.

Also, it reduces the chance of losing coins or beakers to rounding issues.

Here is the more thorough version, as well as several other good tips.

 

[Mahatmajon]

Unfortunately I think you're right. I want to make an SG with a restriction that the tax 'slider' can't be adjusted more than once every __ turns (probably 10) and that it cannot be adjusted more than __ % at a time (probably 20%).

Personally I still prefer doing things that 'feel' like I'm the leader of a civilization instead of things that are perfect game ideal, but I think you're absolutely correct that 0% / 100% is the way to go to maximize (short of actually calculating the optimum value each turn that gets you the best rounding).

 

[Dusty Monkey]

Mahatmajon:

I do not believe such a change will reflect reality.

Do you feel that such a change would significantly effect current gameplay play strategies?

 

[Mahatmajon]

Without trying to pull historical data I don't think a democracy has ever tripled its tax rate overnight and I doubt it has happened often in any government. That's the type of reality I was indicating.

No, I don't think that restricting large movements of the tax slider would affect gameplay much (especially at lower levels), but it could get quite interesting in some situations and at higher levels. If preparing for a war for example and making +1gpt at 80% sci you might have to turn science down before you attack if capturing a few cities will lower your gpt at 80% to -15+

 

[Warlord Sam]

Actually, once America decided to start "fighting" the Cold War, our tax rate more than tripled.

But its never gone down since

I do agree with what you're saying, in that I wouldn't see it as realistic, but just pointing out that it has happened

..sorry, I'm a history major

 

[Brave Jay]

I do this on occasion, however, you know that the beakers are not wasted, they go to the next research project. I like to maximize my beakers to the highest level financially possible, unless i go to war. then, when im at war i might go down to 0-30 percent and raise lots of cash for the war effort. I try not to do it for too long, but i just want to make sure that i never run out of cash during possible riots and war weariness in war time.

 

[maltz]

An alternative way is to sell low techs to inferior Civs... I find it very effective in games where they are many Civs. Some will fall behind and become your money source. It seems to me that all AIs get 10 (or multiples of it) free gold per turn in Emperor - Deity.

If they are pleased with you things are even easier - they will give you all their cash for free - except Mansa Musa. What a #$@%#

 

[1889]

I was trying to check these numbers and found that it may be even better than you suggest.

I set the slider to 0% and watched my science progress. My project increased by 1 beaker each turn even though no gold is being spent on science. This free science may be a built in pity point by the designers or due to the reduced cost due to other nations research.

Now when I set science to 100% my financial adviser shows that 12 gold go to research, but when I watch the progress it increases by 15 each turn!

So this binary 0/100% research maximizes this free research.

 

[maltz]

Now when I set science to 100% my financial adviser shows that 12 gold go to research, but when I watch the progress it increases by 15 each turn!

Let me guess. You built a library.

The library bonus also exists if you maintain 80% all the way, though.

 

[1889]

Good guess, but no. I'm only a few turns into the game. I have no buildings and my worker just got started so I don't even have any trade routes yet.

 

[Zombie69]

I wouldn't say stupid, but it's certainly suboptimal : http://forums.civfanatics.com/showthread.php?t=159109

I've been using binary science ever since Civ 1.

 

[Zombie69]

1889, what happened is that you're getting a bonus either from having prerequisites for the tech, for other civs knowing it, or most likely both these factors combined.

See the "Research explained" article.

The free research you get is always 1 no matter what your research rate is.

 

[DementedAvenger]

This isn't optimal if you have more research-enhancing buildings than gold-enhancing buildings, which is often the case, in which case it's actually very bad for your overall research.

 

[Zombie69]

It's optimal even if all you have is research buildings. I'd like it if you could explain why you think that's not the case so that i could explain where you're going wrong.

 

[DementedAvenger]

It's optimal even if all you have is research buildings. I'd like it if you could explain why you think that's not the case so that i could explain where you're going wrong.

You're right, after doing the math myself I see I made a bad assumption.


Let's say you are making 100 raw commerce per turn. Let's say you have 100% extra research and 50% extra gold from buildings.

100% research: 200 beakers / turn, 0 gold per turn


50% research: 100 beakers / turn, 75 gold / turn


0% research: 0 beakers / turn, 150 gold per turn

Let's say we need 3,000 beakers and are losing 75 gold per turn at 100% research.

If we were to use 50% research, it would take exactly 30 turns.

If we were to use binary research, we'd need 15 turns of 100% research, and to make up the 15*75 gold lost, we'd first need to have 15 turns of +75 gold at 0% research, for a total of 30 turns.


So yep, you guys are right, my bad.

 

[naf4ever]

Someone made a post about "binary science" a couple weeks ago, and made two of the points you made.

Additionally, I recall that the cost of techs drop as more civs discover them, therefore by limiting your research to quick bursts you maximize the chance that you will benefit from other civs aquiring it while you bank.

Also, it reduces the chance of losing coins or beakers to rounding issues.

Here is the more thorough version, as well as several other good tips.

You dont lose anything to rounding issues, typically you only gain. I think there is a minimum number of gold you can produce, say 1, even if 90% science would dictate you should be receiving none. So in fact i believe its possible to actually gain commerce but taking advantage of this.

Also the original posters idea fails to mention the opportunity cost of producing banks and markets and then barely ever using them. Versus having at least 10% in gold and utilizing the bonus from these buildings more and taking advantage of favorable rounding situations related to them. I dont believe 100% science is optimal because the bonus from anything that gives +% to gold then then drops to ZERO yet the efficiency of your libraries and universities stays constant. That is a big loss.

 

[BlueRenner]

I'm finding myself massively unimpressed by this strategy.

Even if you're just bopping around the Science slider, trying to take advantage of the the round-down nature of the system, the biggest bonus you can ever realize is 1 extra beaker per turn.

When at 0% science, you get this free beaker.

When at any other science setting, you have the sum of all the science produced by all your cities (fractions count here!) and have that total rounded down. So, if with this enormous amount of micromanagement, you can recover that singular lost beaker.

By the end of the game you might have accumulated enough extra beakers to research Meditation. Go you. This could concievably put you a single turn ahead in the tech race at the veeeeery end of the game, assuming it gets that far. Pardon my failure to applaud.

This is before even factoring in the effects of Inflation on those piles of gold you presumably have laying about. This might effectively wipe out any gain, depending on how high you run it up before allowing 100% science to whittle it down.

Otherwise, you are just pushing the same beakers and gold around -- you don't gain anything by running 70% vs 100% vs 80% vs 40%, if you are running a deficit you have to make up later. I mean, DementedAvenger hits on it above: researching that tech takes 30 turns at %50, and 30 turns at a %0 - %100 hybrid -- the exact same time.

Beyond that, it works for all combinations. Using his setup:

90% research: 180 Beakers/turn: 16.66 turns-to-tech, 999.6 gold debt
999.6 debt: +75 Gold/Turn: 13.32 turns-to-repay
16.66 turns + 13.32 turns = ~30 turns

80% research: 160 Beakers/turn: 18.75 turns-to-tech, 843.75 gold debt
843.75 debt: +75 Gold/Turn: 11.25 turns-to-repay
18.75 turns + 11.25 turns = 30 turns

70% research: 140 Beakers/turn: 21.43 turns-to-tech, 642.9 gold debt
642.9 debt: +75 Gold/Turn: 8.57 turns-to-repay
21.43 turns + 8.57 turns = ~30 turns

60% research: 120 Beakers/turn: 25 turns-to-tech, 375 gold debt
375 debt: +75 Gold/Turn: 5 turns-to-repay
25 turns + 5 turns = 30 turns

If you want, I can get really fancy and show you the effects of switching around the science slider in the middle of the routine from a non-binary %X0 - %0 situation, but just take my word on it -- the numbers work out exactly the same.

Beyond that singular beaker you might save by calculating the factions for every city you own, there is no point to dancing around the Science slider. You neither create nor destroy anything. You're just pushing one singular commodity around.

- Bill

 

[rddc05]

Without trying to pull historical data I don't think a democracy has ever tripled its tax rate overnight and I doubt it has happened often in any government. That's the type of reality I was indicating.

Actually, it seems that the "tax rate" is set for you in that the amount of commerce you pull in is the economic resource that the government has to work with at base. You then decide how much of those economic resources go to science, liquid assets, and culture. Of course, there are plenty of game mechanics that otherwise make no sense (how doed Universal Sufferage allow cash buying for example?), but that's the way I view the slider anyhow.

 

[dalessi12]

Another possible problem with this strategy, and I admit it is sort of a stretch, but I find I have it on higher levels, is that the more gold you have, the more another civ can ask for. I like to keep my balance low and slightly increasing, when it is mid game, high difficulty (when not using Suffrage). If I "bank" 600 gold for a rainy day (under this method), then after I ratchet up research and run deficit (to catch up), Huayna will pop in and demand 490 gold from me. Whereas if I ran lean and mean the whole time, he asks for less. I usually run research at whatever the highest level is where I am in the green (mid-game), if 70% has me -15gpt, and 60% has me +50 gpt, then it's 60%. If I am in "upgrade units" mode, and upgrading a frigate to destroyer is 250 gold, and I don't feel a threat is imminent, I wait 5 turns, upgrade one, start over, never having more than the 250 gold.

anyway, just thought I would throw that out there as I have found there is sometimes a good reason to keep my treasury low, all things being equal.