Deficit Research

[Teelman]

I'm sure this has been asked before but I have yet to find the answers. I read the strategy article on deficit research but it only seemed to confuse me more. Therefore, I asked these simple questions with what will most likely yield complex answers.

(1) If I am using defecit research, when do you start using it? (ie which time period/after which technology etc?)

(2) How long do you use it? Do you save up enough gold to research the tech that you want at 100% research then move on the next one? Rinse and repeat? Or do you only use it under specific circumstances (like several turns before oxford comes in etc) and research at your max rate without losing money the rest of the time?

(3) Why is deficit research considered more effective than just researching at your highest % rate possible w/o losing money for the whole game?

I am an emperor player with winning percentage 30-40% of time. Any advice is welcome!

 

[UWHabs]

Very complex. The main times it's most useful is when you're building a bunch of libraries or universities or Oxford or something to bump your research up. It's more efficient, since the main idea is that your gold multipliers are relatively the same, but your science modifiers go up after you build those libraries/universities/Oxford/Academy/whatever.

So, in a simple case, you have one city, 20 commerce. A library is supposed to come in in 5 turns. So, you can run at 50% science with 0 extra gold, and you will get 10 beakers for the first 5 turns, and 12.5 for the next bit after that. Alternately, you can run at 0% slider for 5 turns, accumulate 50 gold, and then pump your slider up to 100% for the next 5 turns after that.

So, in the first case, after 10 turns, you have 0 gold, 5*10+5*12.5 beakers = 112.5 beakers. But in the second case, you get 0 science the first 5 turns, but for the next 5 turns, you get 20*1.25 = 25 beakers per turn. So after 10 turns, you still have 0 gold, but now you have 5*25 beakers = 125 beakers.

So, by deficit/binary researching, you've effectively gotten some extra beakers. Now, it won't always be this much, but the idea holds that if you will somehow in short order get more science multipliers, you should try to maximize the beakers that go through those multipliers.

And that's not even counting the good thing about having some extra gold around for those random events that are like, "pay X gold and get a free military instructor" or "pay X gold to rebuild the forge" or whatever. Or even having the case to trade for techs.

 

[UncleJJ]

It makes sense early in the game when you first build libraries or monasteries and again when you build universities. Basically save gold anytime when you know the research modifiers will increase (due to a buildings completing or a civics change). That is the only time it is better than just running the slider as high as you can. The effects are usually small.

Another form of deficit research is when you can raise gold by some means and then fund a period of heavy research. There are many ways to raise gold, including : begging from the AI, failing to build a wonder, sending a GM on a trade mission, selling a tech for gold.

You can even build Wealth in some cities and fund research although I guess that's not a defict unless you lose gold, but it is more efficient than simply building Research if the research modifiers averaged over your empire are better than the gold modifiers

 

[lymond]

I found have the extra gold has made trading a thousand times more effective for me which is just as good as research. Gold is your friend. With beaker modifiers in and cottages growing you are still researching at 0% but at a much slower rate. However, after accumulating some gold and then popping back up to 100% -bam - your tech will be done soon and now you have some gold to work with to trade that tech around, balance out trades or simply buy techs. Also, a pot o' gold helps support any unit pumping you may be doing.

I definitely do these right after libraries are in and I've have most of my initial worker tech in. I usually use Alpha or Aesth as my initial placeholder depending on strategy.

A clear important benefit of this approach is supporting your expansion while stabilizing your economy. I'm rarely in a situations in which I can run 100% research while REXing or conquering, especially at higher levels.

Furthermore, once you have gold modifiers in appropriate cities like markets, banks, grocers etc., plus courthourseswhen you drop the slider to 0 and boom! you collect gold at a very fast rate.

Later, with Universal Suffrage running you can buy units and buildings like nobody's business which is great for low prod cities, new cities or to build a nice army.

Basically, I think this is a more effective approach that trying to run research at 100% or high as possible with a slow trickling rate of gold. Granted there are other ways of amassing gold as mentioned above but these are not always consistent or available. Timing all this is something that I still working with but have some general approaches as explained above. Certainly after reaching certain tech goals or having a solid tech lead, it's a good time to boost your coffers a while.

Gold is awesome and I've just started to realize this not long ago.

 

[cabert]

I'm a big fan of real deficit research, where the deficit is paid by the AI by one of the following methods :
- $ for old techs or maps
- $ for nothing (aka begging)
- pillaging their land/taking their cities

I just love currency for this, and I go for it asap.

 

[lilnev]

Another situation is when you might change your mind on which tech to research next. This is particularly the case for race techs (if I can be first to Music for the GA, I'm interested. But if an AI is going to beat me I'd rather have a pile of gold than half of Music done), and if you expect tech-trading opportunities to become available (if an AI becomes willing to trade Machinery to me, maybe I'll research Engineering next and prepare for war....).

The downside to having a pile of cash is that AIs tend to demand it as tribute.

 

[lymond]

Curious about begging for gold - how does that work? I haven't really tried that one before. What's the mechancs of it and how much can you usually asked for. I've certain demanded gold, but never just asked for it.

@lilnev - I generally don't think they will ask for a tribute unless your power rating is low, although that can be the case early in high level games when I'm just trying to keep up with techs and economy before mass pumping units.

 

[Izmir Stinger]

Isn't deficit research just when your income is in the red? I thought the term used for bouncing your slider between 0% and 100% was called binary research.

Also, Teelman, there was a time when floor functions (i.e. rounding down to the nearest integer) in the calculations for beaker production could cause a loss of up to one beaker per city per turn, which could be avoided by always either having your slider at 0% or 100%. A patch (3.13?) modified the formula so that you can loose at most 1 beaker per turn for your whole empire, so that aspect of the advantage of binary research has greatly diminished. Still though, if the game is 500 turns long that's a whole tech lost to rounding right there.

Waiting for better science multipliers to come on line is reason enough to do it, though.

 

[Windsor]

Waiting for better multipliers is a good reason. There is also the fact that you get a discount on techs based on how many civs who already know it. But this is not my main reason, not at all. Trading and knowledge is.

Knowledge, you start teching music and then you realize Ragnar is going to declare on you. If you had saved the cash you could have gotten feudalism, now you have the option of finishing music or switching to the more important tech but you will not get it before Ragnar declares since you invested some in music.

Trading, this is the most important thing about binary research and it's relevant the entire game, not just when building libraries and oxford. Say you have a city with a banana, and three dyes. Two of those dyes are cowered in jungle. You have BW, math and sailing. What do you research? You start on calendar right? And some turns before you finish it all the other AIs have it, but now the best thing is still to finish it and then start on IW.

So, if you instead turn off research(with calendar as a placeholder). You would see that all the AIs got calendar. So you decide on currency and trade for IW, calendar and some gold. Voila binary research have just increased your science-rate way more than the few percentages those libraries acount for.

 

[cabert]

Curious about begging for gold - how does that work? I haven't really tried that one before. What's the mechancs of it and how much can you usually asked for. I've certain demanded gold, but never just asked for it.

the mechanic is simple, but you're never sure you'll get what you ask for.

If an AI is pleased or friendly, you can ask for anything, no diplo penalty to fear.
However, he'll only give something for every turn where you didn't ask.
And if you asked recently (5% to forget a demand each turn) he won't give anything at all.
And you can't ask what he'd be ready to give .
So often, I'll ask for all the money they have as soon as they get pleased, then wait for a while (20 turns min, but longer isn't necessarily wrong), and ask for some more.

If you keep good relations, it can be a good money source.

 

[Izmir Stinger]

Sometimes they will give a surprising amount. Once I accidentally clicked "Can you spare this" when I meant to click on "What do you want for this" when trying to find out what Mansu Musa wanted for Rifling. He just gave it to me.

Since that game I set myself reminders to beg for something every 30 turns and ask for what would most benefit me rather than what I think they will realistically part with. I've had many pleasant surprises doing this. Few games go by now where I don't get at least one tech this way. It is usually some cheap back fill tech like Meditation, but still...

 

[Syiss_]

Very useful, begging. I still don't do it nearly as often as I should, but it can be VERY helpful, and its surprising how much you can get from the AI's by doing it. Techs I dont often get, but AI's stockpile a lot of gold at times so its not hard to get enough money from your buddies to upgrade your units after you finish researching a tech. Much better than dropping your research and waiting several turns to save up the gold for it.

 

[lilnev]

It is usually some cheap back fill tech like Meditation, but still...

Does this count against WFYABTA? I would never accept Meditation in a trade, even if they're willing to throw it in as a freebie.

 

[Izmir Stinger]

I think people are too paranoid about WFYABTA. Its only the leaders with a limit of 5 you got to worry about. Unless one of them is likely to be a valuble trade partner, I ignore it. Sometimes I'll hit the guys with a 10 limit, but only every once and a while. WFYABTA is not a serious enough consideration to justify blanket statements like "Never trade for Archery" etc.

As to you question, I don't know but I suspect it does.

 

[CornPlanter]

I think people are too paranoid about WFYABTA.

I've only got WFYABTA twice in my life. Never care about it.

 

[TheMeInTeam]

I think people are too paranoid about WFYABTA. Its only the leaders with a limit of 5 you got to worry about. Unless one of them is likely to be a valuble trade partner, I ignore it. Sometimes I'll hit the guys with a 10 limit, but only every once and a while. WFYABTA is not a serious enough consideration to justify blanket statements like "Never trade for Archery" etc.

As to you question, I don't know but I suspect it does.

Deity can really press WFYABTA badly. You get fewer trades until it triggers and probably need it more.

 

[lilnev]

I have some AIs in WFYABTA almost every game. Maybe large fractal maps contribute -- often fairly high connectivity, and thus a lot of trades; also, WFY gets incremented for every leader you've met by the time of the trade, so it's a big issue on pangaea-ish maps, less so on continents-style maps, where there will be a 'second round' of trade partners who have no knowledge of your early trades.

But, every time a trading partner reds out on me, I think back to which throwaway tech is now costing me the ability to trade for Sci Meth, or whatever.

 

[Izmir Stinger]

also, WFY gets incremented for every leader you've met by the time of the trade, so it's a big issue on pangaea-ish maps, less so on continents-style maps, where there will be a 'second round' of trade partners who have no knowledge of your early trades.

Ah yes. My favorite map type is the classic Continents script, which may contribute to my lax attitude towards it, and my cottage spamming ways mean I often get a resounding tech lead moving into Industrial times, so if I haven't hit it by then I wont. However, I do concede that your strategy for dealing with WFYABTA is very much dependent on how many civs are in contact and how early and my attitude wont be appropriate on all maps.

Do you agree that is Mansa Musa is on the map it can usually be ignored entirely?