I recently started a thread with similar topic in the modding forum :
http://forums.civfanatics.com/showthread.php?t=460497
New approach to Science / Research
1. Introduction
In every Civ-game research is based on population, specialists, buildings (and Research- / Trade-Agreements). If you disregard the monolithical Super-Science-City, usually the more population in fully developed cities a nation has the better. This means Science-Output will strongly scale with available landmass covered with cities.
Civ-games can be played on different map sizes. Bigger maps usually have more civilizations at start but usually most civs are also eliminated during game leaving the run-away-civs more land to settle and achieve enormous science production. Science progress on bigger maps therefore should be faster than on small maps.
To provide a comparable Science-Progress, research-speed might be scaled with map- or land-size. However a simple scaling making all techs more expensive will cause new problems. More expensive starting techs will slow down the early game until the player expands. Successfull expansion usually requires certain techs to exploit ressources or build happiness buildings which require techs. So the scaling value should slowly increase from antique techs to modern techs.
Scaling tech costs will make a game with OCC or small empire on a huge or giant map more difficult or eventually impossible to win.
In real life science is not a simple value based on population. If it would be so easy, india and china would have the same or bigger science output than europe and USA. Also in real live scientific progress is based on education and exchange of information. Having millions of individuals individually inventing some new tech will probably not be successfull since many modern techs require specialised knowledge and factory or laboratory environment for experiments. All these individuals might invent the first 5% of a new tech but would not be able to complete the 100% tech. So they should not count for research. For example 99,9 % of todays population is probably just not involved in development of newest computer chip technology. Development is done by specialists and scientists. So mass population just should not count.
Another limitating factor in real life is exchange of information and coordination.
A single person does not need to exchange information with itself but it is just a single person and may not live long enough to fully develop a new tech.
A small team of scientists already must invest some effort to exchange information between team members and to coordinate the team. So from 100% of the work time some part goes into the development and some part goes into project managing.
Having 2 or more teams working on the same tech increases the effort to exchange information between and coordinate teams / team members.
Probably for every project there is an optimum number of team members or teams where adding more team members / teams will result in no effect or a negative effect making the project more expensive and eventually will even delay finishing the project.
Also there are some projects which cannot be speed up by just adding more people.
"In nature it takes 9 months for a pregnant woman to give birth to a baby. A project manager is a person who believes that a team of 9 pregnant woman will give birth to a baby in just 1 month."
2. The new Science approach : Local Tech Projects
The new science model would change all techs to projects which each can only be locally researched in a single city. You can compare it with kind of a national wonder or space ship part or manhattan project which is constructed by a single city using the science points produced in the city. A Super-Science-City will complete tech-projects faster than a normal city. If you have 10 tech-projects available for research, you can assign research in 10 different cities parallel or you can queue them in your Super-Science-City. Key technologies probably should be researched in the more efficient cities to prevent bottle-necks.
Tech Projects will limit research to a maximum of 8-10 cities (depending on the tech tree) independent of the number of total cities in your empire. A scaling of tech-costs for huge maps or huge empires will no longer be neccessary.
There is also the possibility to have requirements and boni for research via ressources or local buildings.
Examples :
- A city with horses and stables might give a 25% bonus to development of knights/chivalry in this city.
- To research Fishing or Sailing, a city must be located at the coast. If the city also has a Fish-ressource in its borders, it gets a 25% bonus to research this tech.
- Or coastal cities get a 50% bonus to naval techs compared to inland cities.
- Banking might require a marketplace.
- Replacable Parts might require a factory.
3. Alternative approach
Another way to limit enourmous growth of science production when map-size is increased is to introduce new limited science buildings (similar to the National College).
In a centralistic, monolithical approach the Palace, National College, University of Oxford and other national wonders might each give a +100% science bonus to the capital making the capital a Super-Science-City and making other cities, especially smaller cities more and more irelevant for research. In a modded game with comparable changes to science boni my capital had science output of 100-1.000 Science Points a turn while other cities only contributed 1-100 points depending on size and buildings in the city. Small tundra-villages will only contribute a few beaker and can be completelly ignored unless there are thousands of them. (However with the huge capital-bonus a perfect starting location for the capital will be critical.)
In a similar approach a number of new minor National Wonders could be introduced, allowing each nation to build maybe 4-8 copies of each minor National Wonder. So for Science this would be small copies of National College, University of Oxford, etc. strengthening the core cities of a nation and making other cities less relevant for science. These 4-8 small national wonders might be combined with the stronger National Wonders or might be exclusive.