How to Tech Faster?
- Thread starter GAAAARG
- Start date
- Joined
- Mar 26, 2007
- Messages
- 7,920
Nothing can be done about those losses.
Xunin is absolutely right about saving wealth (research 00% / wealth 100%) until a new research multiplier building is constructed and then go to 100% research and profit. This should be the #1 reason for using binary research. It is also best to maximize research slider and minimize the wealth slider while building a wealth multiplying building and maximize wealth slider after it has been completed to profit.
Sun Tzu Wu
Seraiel
If you want anything from I please ask in German
- Joined
- Sep 6, 2011
- Messages
- 8,165
Hmm, I just found out that my feeling for the Math on this was actually wrong.
100 / 0 is the way to go, as long as Research-multipliers and Gold-multipliers are at least balanced.
The only case, in which it's better to run with a lower slider, is once Banks are reached, then building Research and producing Gold (via the Slider) is actually more efficient, because Banks in general are the most efficient multiplier-building available.
Learned something.
100 / 0 is the way to go, as long as Research-multipliers and Gold-multipliers are at least balanced.
The only case, in which it's better to run with a lower slider, is once Banks are reached, then building Research and producing Gold (via the Slider) is actually more efficient, because Banks in general are the most efficient multiplier-building available.
Learned something.
This should explain how research works...
War Academy: Technology Research Explained
And if I can comment on the binary research debate: Yes, binary research is a standard strategy with a good reputation to its name. However, if you have met most of the other civilizations early (and are researching a tech "all" the other civilizations already have), you may end up losing some beakers per turn by setting your slider to 0% research (as any beaker you put towards research will get a multiplier (from other civilizations (that you've met) having that tech already).
If you're leading the tech-race, however, there is no such loss of beakers (as none of those civilizations you know have the tech...). [Maybe binary research is better on lower difficulty levels?
]
Yours Sincerely
Kjotleik of Norway
War Academy: Technology Research Explained
And if I can comment on the binary research debate: Yes, binary research is a standard strategy with a good reputation to its name. However, if you have met most of the other civilizations early (and are researching a tech "all" the other civilizations already have), you may end up losing some beakers per turn by setting your slider to 0% research (as any beaker you put towards research will get a multiplier (from other civilizations (that you've met) having that tech already).
If you're leading the tech-race, however, there is no such loss of beakers (as none of those civilizations you know have the tech...). [Maybe binary research is better on lower difficulty levels?
]Yours Sincerely
Kjotleik of Norway
pomthom
Drive & Reverb
So this doesn't apply if you don't select a tech to research right?
This doesn't make any sense to me.
fyermind
Chieftain
- Joined
- Apr 26, 2008
- Messages
- 96
My understanding here might be off, but iirc if you don't select a tech to research you don't get the world multipliers for how many AIs have researched it or something. I know when forget to pick my starting tech on T0 and select it on T1 I always loose beakers (usually ~8% of my research or 1 beaker)
In the later game, I think you get the tech rate of your previous tech on overflow, but a code diver should probably confirm that.
drewisfat
King
- Joined
- Feb 8, 2011
- Messages
- 995
I think some of the main reasons for binary research aren't even rounding errors / multiplier considerations -- but from trades. Trading a tech for gold is much better than the converse (Since you don't lose anything). It's pretty easy to trade techs for gold to keep maxing out research -- if the AI has a lot of gold you can even tell if he's researching that tech sometimes. Similarly it's usually worth it to sell of excess (or even non-excess) resources for GPT.
Also it helps you know what tech to get. You could research tech X for 10 turns at 50% slider. Or you could wait 5 turns banking gold and THEN decide if you want to get that tech, based on what techs the AI have researched / other changes in the game.
Also it helps you know what tech to get. You could research tech X for 10 turns at 50% slider. Or you could wait 5 turns banking gold and THEN decide if you want to get that tech, based on what techs the AI have researched / other changes in the game.
MyOtherName
Emperor
- Joined
- Dec 7, 2004
- Messages
- 1,526
Originally, the most important reason was rounding errors, because 
and 
used to be rounded to integers in individual cities rather than over your whole empire, which was a very big deal when your cities aren't producing much commerce. I think libraries, etc, kicked in after rounding too, so you effectively got penalized twice with the rounding errors. i.e. if you produced 4 
at 90% science in a city with a library, you were actually 0.40 and 1.50 per turn due to rounding errors!
These days the rounding happens at the empire level, but it can still be fairly important in the very early game: e.g. when you're producing only 14, only getting 13 + due to rounding is a nontrivial loss.
_____________
The idea that if you have better multipliers, then acquiring extra (e.g. by trades or building wealth) is more valuable than acquiring an equivalent amount of extra (e.g. by trades or building research) -- and the opposite should you have better multipliers -- is something different. Or at least, it used to be -- is this idea also going by the name "binary research" these days?
and used to be rounded to integers in individual cities rather than over your whole empire, which was a very big deal when your cities aren't producing much commerce. I think libraries, etc, kicked in after rounding too, so you effectively got penalized twice with the rounding errors. i.e. if you produced 4 at 90% science in a city with a library, you were actually 0.40 and 1.50 per turn due to rounding errors!These days the rounding happens at the empire level, but it can still be fairly important in the very early game: e.g. when you're producing only 14
, only getting 13 + due to rounding is a nontrivial loss._____________
The idea that if you have better
multipliers, then acquiring extra (e.g. by trades or building wealth) is more valuable than acquiring an equivalent amount of extra (e.g. by trades or building research) -- and the opposite should you have better multipliers -- is something different. Or at least, it used to be -- is this idea also going by the name "binary research" these days?
pandamancer
Prince
- Joined
- Aug 18, 2013
- Messages
- 335
In terms of the research slider, I do a 0-50-100% in research
The idea is almost the same as binary researching (0 or 100) but sometimes I also go for 50% research if most of the conditions are true:
a) you do have research+gold or research+espionage generating buildings
b) late game
c) you want to stall the ai so he won't get the research bonus, yet you also want to keep ahead to be first in getting that technology
d) an acceptable research speed without a huge strain in economy
e) you want to give some to espionage without sabotaging your research a lot
To do a 50% without loss in rounding, you must first ensure that the raw commerce in each city is a even number. By applying 50% to each raw commerce, you will always get a whole number. Applying the research/gold/espionage percentage buildings and civics (assuming they are of 25%-50%-75% percentages), you will always get .25/0.5/0.75/0.0 in the final computation of research/gold/espionage. With some painstaking effort, you must balance them so that the whole empire must gain a bpt, gpt and ept of a whole number.
As a consequence, buildings like monastery/seowon/stock exchange/mint or civics like free religion screws up the 50% research speed.
The idea is almost the same as binary researching (0 or 100) but sometimes I also go for 50% research if most of the conditions are true:
a) you do have research+gold or research+espionage generating buildings
b) late game
c) you want to stall the ai so he won't get the research bonus, yet you also want to keep ahead to be first in getting that technology
d) an acceptable research speed without a huge strain in economy
e) you want to give some to espionage without sabotaging your research a lot
To do a 50% without loss in rounding, you must first ensure that the raw commerce in each city is a even number. By applying 50% to each raw commerce, you will always get a whole number. Applying the research/gold/espionage percentage buildings and civics (assuming they are of 25%-50%-75% percentages), you will always get .25/0.5/0.75/0.0 in the final computation of research/gold/espionage. With some painstaking effort, you must balance them so that the whole empire must gain a bpt, gpt and ept of a whole number.
As a consequence, buildings like monastery/seowon/stock exchange/mint or civics like free religion screws up the 50% research speed.
Similar threads
