Researching Efficiently in Civ 5

UncleJJ

Deity
Joined
Jun 13, 2006
Messages
2,226
Location
London
Efficient Research in Civ 5

This is a strategy article outlining my ideas on how to tune research in Civ 5 so that very few beakers (science points) are wasted and how it is possible to grab a tech a turn earlier. Unlike Civ 4, Civ 5 does not overflow unused beakers from a previous tech into the next tech. Many players consider this a major design flaw or even an oversight and want it fixed in a patch. But this is not mentioned on the list of things that will be fixed in the next patch so we will have to live with this situation for at least some time. So whatever you think about this “feature” it is worth knowing how you can overcome or limit its effects. That’s what this article is about.

The Problem
Many people are convinced, and keep writing on this forum, that simply running scientists will boost their research rate, but this is not really true. The key thing these people are missing about researching a tech with beakers is that there is no beaker overflow. That means, in the middle game for instance, if you are producing 150 beakers per turn and your tech costs 500 beakers you will take 4 turns to research it and having produced 600 beakers for a 500 beaker tech will waste 100 beakers. These excess beakers do not get added to the next tech as they did in Civ 4. In that case some of the scientists you are running are wasted and generate zero useful beakers (although their GPPs might be useful). In general it would be better to employ the citizens on other tasks (working a mine, farm or TP) or change the type of specialist. The scientist is definitely not the best specialist, in this situation, he's the worst.

If the tech you're researching does cost 500 beakers and your research output is around 150 beakers / turn you will take 4 turns to research it unless you can boost beaker output to 167 and then research it in 3 turns with only 1 beaker wasted in overflow. If you can’t do that then you should reduce all optional sources of beakers (specialists and beaker tiles) until you have 125 beakers / turn, if you don't want to waste beakers and have other options with the citizens (generate more food, gold, hammers or culture). This what I mean by tuning your research, minimizing the lost beakers and maximising overall economic output (food, gold, hammers and beakers)

Methods of Research
There are three basic ways to get your research done:

a) Research Agreements (post Philosophy) are probably the most cost effective way to get an expensive new tech (once techs start costing over 500 beakers). Pay 150 to 300 gold and receive a random tech in 30 turns as long as you don't go to war. Currently this is bugged such that declaring war ends RA and gives the tech instead of cancelling it. Human players can exploit this by making a RA and then declaring war soon after to speed up the tech.

b) Free techs can be gained by sacrificing a GS. You can also get free techs from 2 wonders and SP Scientific Revolution gives 2 free techs. Free GS can be gained from wonders and randomly from CS with the SP Educated Elite.

c) Research the tech with beakers. Beakers come from your pop, from specialists and from some tiles and CS, with appropriate SPs. This is explained more fully below. Beakers fall into two categories, those that provide a minimum or base load, and those that are voluntary and can be changed or tuned on a short term basis up to a maximum output. Population in your cities and puppets and the amount of beakers from CS can’t be changed easily but do tend to increase through the game as cities grow and build multiplier buildings and as new cities are acquired. Specialists can be run (or not) in cities with appropriate buildings. Some tiles produce beakers and these can be worked (or not) depending on need. So specialists and tiles are what we will mostly be considering when tuning research on a turn by turn basis.

Beaker sources
There are several ways to generate the beakers used for research

a) Beakers come from your own pop, and that of puppets boosted by research multipliers. Each pop you control in your cities and any puppets gives 1 beaker and the library boosts this to 1.5 base beakers per pop.
b) A few tiles give some beakers; jungles give 2 beakers (with a university in city), late game TP gives 1 beaker (with the SP Free Thought) and a GS settled on a tile as an Academy gives 5 beakers.
c) City states give circa 8 to 13 beakers with the SP Scholasticism.
d) Scientist specialists give 3 beakers and up to 5 can be run in cities with all the research buildings.
e) All specialists give +2 beakers with the SP Secularism and scientists give 5 beakers.
All the sources of beakers are added and then research modifiers multiply these basic beakers. Universities, Observatories, National College and Public Schools each give a 50% boost and Research Labs give a 100% boost. This makes a 200% or even a 300% boost possible in some big cities in the late game.

Limits of tuning with Scientists
The easiest method of tuning research in the early and middle game, before Secularism, is by running an extra scientist or not for a few turns. Each city can contribute a degree of tunability depending on its buildings. Not all scientists are equal in their ability to tune research. Since each of the 4 basic research building requires the previous one to be built we can form a simple list that takes account of additional specialist slots and increasing multipliers.
a) A library adds 2 scientist slots that can tune up to 6 beakers / turn
b) A university adds 1 scientist slot, +50% and can tune up to 13.5 beakers / turn
c) A public school adds 1 scientist slot, +50% and can tune up to 24 beakers / turn
d) A research lab adds 1 scientist slot, +100% and can tune up to 45 beakers / turn
So a city with a research lab provides 7.5 times the tunability of a city with only a library despite having only 2.5 times as many scientist slots. Adding the National College or an observatory to a city adds another +50% multiplier but does not give extra slots so this increases the tuneable output by 1.5 beakers for each scientist. The ultimate Civ 5 research city with National College and observatory can produce 60 beakers per turn from scientists before Secularism.

Simple Method of tuning research
When you have a new tech to research take a look at how many beakers it takes and the number of turns the game suggests it will take. Notice many techs available at the same time cost exactly the same total beakers, so once this is worked out for one tech the others have the same solution. For example Electricity, Replaceable Parts, Railroad and Dynamite all cost 2090 beakers (standard size map, normal speed)

So say we have 6 fairly big cities and 4 puppets plus 5 CS (giving us 70 beakers with Scholasticism) but we do not have Secularism and with no scientists or other specialists this empire generates a minimum of 610 beakers. If we attempt to research Electricity at 2090 beakers the game will show we need 4 turns. Clearly we could reduce the time to 3 turns if we can raise our beakers to 697 beakers. How many scientists would we need to run? We don’t have research labs yet. 4 of our cities have public schools and the other 2 have universities. One city (with public school) has the National College and another has an observatory. Using the list above we can see each city with a university can generate up to 13.5 beakers, each with a public school up to 24 beakers and the National College or observatory up to 30 beakers. So we can boost research by a maximum of 134 beakers and we only need 87. How we decide to distribute the scientists will depend on food and what else the cities involved are doing but it is possible and worthwhile to increase our tech rate by 33% rather than waste 350 beakers and take 4 turns. Obviously the state of scientist GPPs in each city is another consideration, we may want to speed up or slow down the generation rate depending on what GPs we want to generate next.

Note if we have enough gold, then rush buying a public school in one of the cities with a university will do two things. It will boost the base beakers produced in that city (e.g. a size 10 city will gain 7.5) and also make any scientists running there better adding up to 10.5 beakers.

Tuning with Tiles
Cities with a university working jungle tiles get 2 beakers. Settling a GS to make an Academy adds 5 beakers to the tile. A TP on the jungle tile adds 2 gold and makes it quite a useful tile (for the first time in any version of Civ) yielding 2 food, 2 gold and 2 beakers. With the SP Free Thought all TPs give +1 beaker and in some cities with TP spam this can be a significant contributor. Selecting Science Focus is an easy way to select the best tiles that yield beakers. Focussing on other citizen allocations (e.g. gold) may reduce the amount of science produced by tiles.

Secularism
This SP affects the way specialists can be used to generate beakers. No longer do scientists have a monopoly. Every specialist including unemployed citizens generates +2 beakers and that adds greatly to the tunability of every city. The ability to run several merchants, engineers and artists instead of scientists and still generate enough beakers to meet the requirement makes the decision of how to boost research much more complex and affects other areas of the game. Using artists to add culture, for instance, can increase the rate of adding new tiles for the city and speed up the time taken to obtain another SP (which can be useful even if only 1 turn is saved). If that can be done and still research the next tech in the same number of turns then that is worth considering. Running engineers can generate engineer GPPs and a GE can make building a wonder much easier in a city with low production. And so on, there are many considerations.

It is hard to make general guidelines about how much each city can add in terms of tunability. Now it becomes possible to generate very large amounts of beakers from non-scientists that also do useful things and generate other types of GPP. It is also possible to starve cities to temporarily boost the science output. However it is interesting to observe the relative power of different cities depending on their research multipliers
a) library 10 (2 scientists @ 5) + 2 per other specialist
b) university 22.5 (3 scientists @ 7.5) + 3 per other specialist
c) public school 40 (4 scientists @ 10) + 4 per other specialist
d) research lab 75 (5 scientists @ 15) + 6 per other specialist

Any of those cities with the National College or an observatory will also add another + 2.5 per scientist and +1 per other specialist. Notice that swapping between running a scientist and a merchant in a city with a research lab reduces output by 9 beakers and in a library city it would only be 3 beakers.

A more general solution
Once we have Secularism and Free Thought it is much harder to make a simple decision about how it is best to reduce the time to research the next tech by 1 turn. With the many other types of specialist here are now many options and these have consequences. Different cities can be run in different ways.

Moving the example used above into a later age we can look at the options available now. Again assume 6 cities, 4 puppets and 5 CS and we’re now researching Robotics (3685 beakers) while building the Apollo Program for a Science Victory. With no specialists our cities, puppets and CS are now generating 1000 beakers and it would take 4 turns to research the tech with 315 beakers wasted (about 9%). If we can generate 229 more beakers per turn using specialists and working any tiles with jungles or TPs not worked at present then we can reduce the time taken to 3 turns. That will allow our other cities to start on building Space Factories one turn earlier and could shave 1 turn off the Space Victory. Let’s see if we can do that.

First go to our cities and check if selecting Science focus will add many beakers by working better tiles. Assume this adds 20 beakers at the loss of some gold and production. We now need 209 more beakers from specialists.

We have 2 research labs already built in the cities with the National College and another city. The city with the observatory is building a research lab but it won’t finish in time for this tech. The last two cities still only have a university. So we have a wide variety of specialists and conditions to pick from.

Here is what we could have, in beakers, for each combination of (scientist, other specialist) for each city:

city1) Nat College + research lab (17.5, 7)
city2) research lab (15, 6)
city3) observatory + public school (12.5, 5)
city4) public school (10, 4)
city5) university (7.5, 3)
city6) university (7.5, 3)

A lot of options and it’s interesting to note how much more powerful city 1 is than either city 5 or 6 for tuning purposes. Ordinary specialists are nearly as powerful in city 1 as scientists are in the other cities.

So let's use some specialists as a first approximation to getting the extra beakers we need to reduce the number of turns to research Robotics from 4 to 3. Say this configuration of specialists makes sense from the point of view of production, GPPs, gold and spare food for growth.

city1 with 4scientists and 4 other generates 98 beakers
city2 with 4 scientists and 2 other generates 72 beakers
city3 with 0 scientists and 4 other generates 20 beakers
city4 with 2 scientists and 5 other generates 40 beakers
city 5 with 2 scientists and 2 other generates 21 beakers
city 6 with 0 scientists and 2 other generates 6 beakers

This totals 257 and is far more than the 209 we need. So we can tune down closer to the target.

We can reduce wasted beakers by either removing scientists or other specialists to work more tiles or by swapping a scientist with another specialist. I won’t bother to detail the actual final set up of scientists and other specialists, but hopefully it's clear the objective can easily be reached in many ways. With Secularism generating a few extra beakers is not harmful as long as you don’t want to use the citizens for other tasks that help the economy more. This is particularly true if you have the Freedom SP (0.5 happy per specialist) and Civil Society SP (1 food per specialist) combined with the Statue of Liberty (1 hammer per specialist), then running extra specialists has an impact on happiness, food and hammers as well as generating beakers.

However, we have achieved our primary objective of reducing the time to research Robotics from 4 turns to 3 and thereby reduced the time to a Space Victory by 1 turn, at least potentially.

Conclusion
By tuning the number of specialists it is possible to speed up the time taken to research key techs. It is also possible to avoid wasted beakers and use the citizens for more useful purposes when they’re not needed to generate beakers.

If the current overflow problem is patched then the wasted beaker aspect of this article will be superseded. However, there is still merit in understanding how to boost research on a temporary basis to grab a few techs a turn or two early. So hopefully the techniques and observations made here will still have some value even if a patch fixes this problem.

Thanks for reading

Abbreviations
Beakers: are science points, a hang-over from Civ 4 where the symbol for science is a beaker.
GS: great scientist
GM: great merchant
GE: great Engineer
GP: great person
GPP: great person points
CS: City State
SP: Social Policy
RA: Research Agreement
TP: trading post
 
Obviously the simplest thing would be if they'd just re-implement beaker overflow. I don't understand why they didn't do this. They fixed it in Civ4, why did they step back again? It encourages way too much BS micromanagement.
 
Obviously the simplest thing would be if they'd just re-implement beaker overflow. I don't understand why they didn't do this. They fixed it in Civ4, why did they step back again? It encourages way too much BS micromanagement.

Instant beaker overflow can also be abused. Depending on how it is implemented, it can be abused to reward micromanagement just as much as not having overflow at all.

I think the best way would be to throw all overflow into a bucket, but to have that bucket leak back to you slowly...say capped at 5% of your total research rate.

For example:

You generate 20 overflow while researching at 100 beakers/turn. Those twenty beakers get returned over the course of 4 turns (4 = 20 / (100 * 5%)) as a simple +5 beakers/turn added to your global total, like a CS science bonus.

If you get crazy overflow, like when you research Pottery after Dynamite, then you'll get every last beaker back eventually, but it may take up to 20 turns, (or longer...since you are likely to keep overflowing several techs in a row with that setup)
 
Instant beaker overflow can also be abused. Depending on how it is implemented, it can be abused to reward micromanagement just as much as not having overflow at all.
How so? This makes no sense to me.

Supposing you had instant 100% overflow, then 1 beaker is as good as another. I don't see any MM incentives. More science is always useful, but always has an opportunity cost, and these are continuous functions.

MM incentives come from discontinuities in cost-benefit functions - like that caused by a lack of overflow.
 
Good article, but I won't be doing this. It's just too tedious, and I don't care enough. I don't play multiplayer where this would really be needed though. In single player, you can get away without doing this.
 
This is a well thought out article. Unfortunately (for this article) I think we can all assume that beaker overflow will be patched into the game, but until then this is a good guide for those wanted to micromanage as much as possible.
 
I appreciate the in-depth analysis, but honestly it seems like a lot of words were used just to say:
Adjust your beakers per turn to be as close as possible to a number that [tech cost] is divisible by. Do this by assigning or removing specialists and working or not working science tiles. Specialist slots, multiplier buildings, and some social policies allow greater flexibility.

So while the math was cool and all, I just don't think it has any practical use. Figuring all this out exactly on a turn-by-turn basis (that city grew, now I'm wasting an extra beaker per turn, reassign tiles. That city got a library, I can stop running a scientist now, etc.) seems like a very poor return for your time.

It is fun for me to read all the specifics, I guess I kinda wish the effort was put into something where it could all be used somehow, like how much culture a new city would need to produce to get back up to a pace of X turns per social policy- something where using the results will benefit the player a lot compared to just fiddling and approximating in-game. I really want to like this thread since it's a great analysis, but I just don't take anything practical from it at all.
 
So while the math was cool and all, I just don't think it has any practical use. Figuring all this out exactly on a turn-by-turn basis (that city grew, now I'm wasting an extra beaker per turn, reassign tiles. That city got a library, I can stop running a scientist now, etc.) seems like a very poor return for your time.
Of course it has practical use. If you do it incorrectly and end up one beaker off your research you basically waste a turn's science finishing off that single beaker. It's something that can be optimised, provides useful results when optimised, so why not optimise it? In some cases it can make all the difference, especially at war time and over the game it might save you hundreds, if not thousands of beakers. It's not like this is an RTS where you only have a limited number of things you can physically do in a short space of time, so I don't see how "return for your time" comes into it, unless we're talking about fun! Fun is the reason it should overflow...
 
Thank you for sharing the fruits of your study.

It is not micromanagement to focus on the factors vital to a particular strategy. If this is the best way to get a technology victory your method is not micromanagement. If it is not necessary for an Ancient Era domination victory it would be micromanagement in that game. It depends on your game strategy.

I will be using in my next game as I've not scored a technology victory yet.
 
This is your optimal science city. Grow the population in this city to as large possible.
Consider a science city that is next to a mountain and has some jungles nearby:

Library (+1 science for every 2 citizens)
University (+50% science and +2 science for each WORKED jungle tile)
Observatory (+50% science - city must be built next to a mountain)
Public school (+50% science)
Research lab (+100% science)
National college (+50% science)

Also include a granary, watermill, Lighthouse (on the coast), hospital, and medical lab for population growth. Use maritime city states (Siam multiplies city state bonuses). You can explode your population by allying with one or more and also settling in an area with high base food. Build a garden in this city to improve the rate the city generates great scientists. To further increase this rate, get the Hagia Sophia wonder. It doesn't matter which city builds it, but put this wonder on your priority list.

Shoot for the great library and porcelain tower the free great scientists.

This city is likely going to have poor production, so accumulating some cash to rush build at least a few of these building can go a long way.
 
Let me understand.....

in a city with poor production you are able to build:

Library
University
Observatory
Public school
Research lab
National college
Granary
Watermill
Lighthouse
Hospital
Medical lab
Garden
Great Library
Porcelain Tower

?!?!??!?!?!?!

Ah ok .... if no coastal you're going to skip the lighthouse.....
 
I think it is worth noting here that research overflow was implemented in the last patch.
 
Let me understand.....

in a city with poor production you are able to build:

Library
University
Observatory
Public school
Research lab
National college
Granary
Watermill
Lighthouse
Hospital
Medical lab
Garden
Great Library
Porcelain Tower

?!?!??!?!?!?!

Ah ok .... if no coastal you're going to skip the lighthouse.....

No one said it has to have poor production. Moreover, you can purchase many of those buildings if you're taking in enough gold. In fact, if you have a few riverside hills, you can get good production, and when farmed will provide plus two food with Civil Service.

Often times, my best science city is my capital because with Tradition, it grows quickly and almost always has good production.

BTW, the wonders don't have to be built in the capital to get their benefit, and I don't always get them, but I usually at least try.
 
I’m a little late getting to Civ 5. Was this issue of beaker overflow dealt with post patch?
 
Many thanks for this great article and especially for the legend of abbreviations, which really makes it understandable for beginners like me. TY! Even as they fixed it meanwhile, the article contains lots of useful information for me!
 
over me head, but thanks for the read

however i have facts due to my last game

Fact- Build next to a Mountain and have Jungle nearby

Fact- Babylon

Fact- Use the first scientist to build an academy

Fact- Gold, Science. Two cities - one on the mainland (capital- Science), second on an isle (Gold)

all gold to research agreements. Great Library, Oracle, Porcelin thing. Tradition, Rationalism, trading posts

the evidence is clear- Science Master of the New Forever
 
over me head, but thanks for the read

however i have facts due to my last game

Fact- Build next to a Mountain and have Jungle nearby

Fact- Babylon

Fact- Use the first scientist to build an academy

Fact- Gold, Science. Two cities - one on the mainland (capital- Science), second on an isle (Gold)

all gold to research agreements. Great Library, Oracle, Porcelin thing. Tradition, Rationalism, trading posts

the evidence is clear- Science Master of the New Forever

That's absolutely terrible advice. Take your "facts" elsewhere.

1) Build next to luxury resources you don't have or can sell, on or next to river hills, or clusters of food resources.
2) Use the first GS, as well as all other GS, for a strategic tech acquisition.
3) Use gold to ally with maritime city-states that provide needed luxury resources, and to build monuments in new cities. Buy research agreements 3-4 at a time, and either fully research or block (exploit imo) cheap or unimportant techs before the agreements kick in.
4) Wonder progression: National College before GL so you can get it done with 1 library (before you expand)
5) Run science specialists full-time in capital for GS points; you'll have no problem with growth with the city states and lots of farms.
6) Aim for tech focus switch from econ to military by renaissance era; there will most likely be at least one warmonger who will pull ahead and you may need to take him down or grab a few weak capitals in order to keep up.

Credentials: research and practice (Emperor level currently).
 
@ Troytheface and Shaithis:

I don't agree with either of you. If you play as Babylon, the first GS can be used for either a free tech or academy. I've made it work both ways. You don't need to build the NC before you expand. You can expand first and build it after you get libraries in your first three to five cities. The beauty of civ games is that there are no absolutes. There may be a formula that works for you, but I've found that there is no right or wrong answer. If it works for you, it is the right answer. I can win on immortal with any leader. I can delay science to start an early war if the situation warrants, or I can delay expansion and go NC early. There are many ways to win the game. Pick yours, but don't make absolute statements of fact. They don't work in a strategy game that has many answers to the same question.
 
Top Bottom