Great Scientist: Academy vs Discover Tech

[Bibor]

By the time you have both a university and an observatory you're probably not going to have that many renaissance techs left. There's also the fairly strong argument that a tech now is better than a tech in 100 turns time, and another 150-200 turns after that, especially if you're slingshotting deep.

That's because you make the two most common mistakes that people do:
a) that there's an infinite number of turns
b) that there's an infinite number of available scientists

Lets say you need to research an average of 60 techs per game (depending on your victory conditions, of course). Also lets say the game lasts, on average, for 350 turns.

How many great scientists can you produce during the course of that game? Lets say 6.

Lets say you get your 1st Academy in the Classical era and you build a national college there:
5 *1,5 = 7.5 ; average tech cost 110 takes 15 turns to pay off.

You get the 2nd Academy at time you get your University in the Medieval era:
10 * 2.0 = 20 ; average tech cost around 400 takes 20 turns to pay off.

You get the next 3rd Academy early renaissance and you build an Observatory there:
15 *2,5 = 37,5 ; average tech cost 800 takes 21 turn to pay off.

You get the 4th Academy at late renaissance and you build a Public school there:
20 *3,0 = 60 ; average tech cost 1450 takes 24 turns to pay off.

You add a 5th Academy in the industrial era:
25 *3,0 = 75 ; average tech cost 2100 takes 28 turns to pay off.

You add your final 6th Academy in the modern era as well as a research lab:
30 *4,0 = 120 ; average tech cost is 3000 so it finally takes 25 turns to get the return profit out of it.

***

Total number of beakers needed to research 60 techs: 53225
Average number of possible beaker gains from bulbing: 7860
Real effect of bulbing: 14% of total science needed (excluding the last 14 techs).

If all research would be equal during those 350 turns, the research pace would need to be 152 per turn (152*350=53200). With an average benefit of Academies being around 60 per turn (most benefit comes in the last 100 turns) total research generated by Academies would be 21000. This city alone with science buildings and Academies would make up 40% of your total needed research.

How about that.

The only thing you need to do is to build your science city adjacent to a mountain and with lots of production.

EDIT: High empire-wide research lowers the effects of both bulbing and Academies.

 

[The Great Apple]

That's because you make the two most common mistakes that people do:
a) that there's an infinite number of turns
b) that there's an infinite number of available scientists

These were not assumed by my argument. I really can't see the logic you used to reach this conclusion.


Two issues with your maths. Firstly: the whole point of bulbing is not to bulb an average tech. The tech that is being bulbed is often quite a bit ahead of the average tech for the era. The obvious example of this is slingshotting to rifles/artillery.

Secondly, and less importantly: you're forgetting the 2 GPT this tile would normally be kicking out from a trading post. This 2 GPT lets me get more city states (potentially more happiness/food/culture/military power/science/great people), hurry buildings/units and trade more efficiently. You could argue that 2 GPT will get me 1-2 techs per game purely based on research agreements.

My third issue is with timing. While academies are long-term investments, bulbing pays instantly. Bulbing a military tech will let me get an army into battle 10 turns earlier. Other techs will let me win wonder races, or build earlier universities/other science buildings. Other advantages include earlier adoption of latter-era social policies, and the city state bonuses from changing eras earlier. Basically, now can be a lot better than later.

 

[Bibor]

These were not assumed by my argument. I really can't see the logic you used to reach this conclusion.


Two issues with your maths. Firstly: the whole point of bulbing is not to bulb an average tech. The tech that is being bulbed is often quite a bit ahead of the average tech for the era. The obvious example of this is slingshotting to rifles/artillery.

Secondly, and less importantly: you're forgetting the 2 GPT this tile would normally be kicking out from a trading post. This 2 GPT lets me get more city states (potentially more happiness/food/culture/military power/science/great people), hurry buildings/units and trade more efficiently. You could argue that 2 GPT will get me 1-2 techs per game purely based on research agreements.

My third issue is with timing. While academies are long-term investments, bulbing pays instantly. Bulbing a military tech will let me get an army into battle 10 turns earlier. Other techs will let me win wonder races, or build earlier universities/other science buildings. Other advantages include earlier adoption of latter-era social policies, and the city state bonuses from changing eras earlier. Basically, now can be a lot better than later.

My 14% vs. 40% is just a napkin calculation that shows the theoretical maximum gain from an "Academy spam" is powerful. When all things are taken into account (of which you mentioned quite a few) Academies are still not worse than bulbing. They might be slightly more or less powerful than bulbing, depending on circumstances.

 

[Martin Alvito]

Bibor, the tech costs are backloaded so heavily that you're better off having any GS but the first stand around until it can bulb a 3000S tech. You can start opening those between turns 150-180 on a Standard, Deity map if you optimize Science and use RAs properly.

If the game ran longer than it does in practice, you'd be correct that Academies could be worthwhile under certain conditions. But it doesn't, and they aren't.

 

[Bibor]

If the game ran longer than it does in practice, you'd be correct that Academies could be worthwhile under certain conditions. But it doesn't, and they aren't.

You can't just say "this doesn't hold water because my gameplay style is different". Some games last well into 400 turns. Are these games a sign of an incompetent player or are they perhaps a sign of a different approach?

 

[Polish]

I think that like many things in this latest version of the game the developers got so caught up in the pretty graphics they could put on the screen they lost sight of the point. The science academy is pretty and the graphic changes over time, I only ever built one. They are seriously not worth it. As has already been said, with Rationalism you get beakers in your trading posts making the academy even less valuable. The other thing is that you must have suprlus population to work the academy or you get nothinng for it. I think the academy really belongs in the city as a building--perhaps with a free specialist and it should serve as a science multiplier in that city. Personally I would have rather had the designers spend less time and effort on graphics and more on gameplay.

 

[Martin Alvito]

You can't just say "this doesn't hold water because my gameplay style is different". Some games last well into 400 turns. Are these games a sign of an incompetent player or are they perhaps a sign of a different approach?

Not all games are best served by pure optimization of Science. But with Research Agreements being what they are, there is no reason (other than Bollywood) for you to require 400 turns to get through the tech tree on a decent-sized map.

 

[alpaca]

Ah you meant the beakers per turn for the academy, I thought you meant global beakers per turn science output. That makes more sense, of course.

Yes, it's obviously a balancing question. If an academy yielded, say, 10 beakers instead of 5 I might seriously reconsider.

 

[Zechnophobe]

Hmm, Bibor you have made several errors in your logic.

First of all, you are comparing beakers per turn vs the average techs of an era, as opposed to the most expensive. The first tech I bulb in a game is Civil service.

Secondly:

10 * 2.0 = 20 ; average tech cost around 400 takes 20 turns to pay off.

This is incorrect. You have added the first academy towards paying off the second. You aren't proving that the two combined can pay off a bulb on the second Great Scientist, but rather that the second one is ALSO more valuable to turn into an academy. Therefore it must 'pay off' its own cost. And once again, you should be comparing against the greatest costing techs, not the average.

You have also not factored in the loss of an improvement on the tile in question.

You have also added in the requirement of building science buildings which cost maintenance and hammers. These are adding against the short term costs of not already having the tech from bulbing.

Lastly, while not a mathematical fault, you have made some assumptions about frequency of Great Scientists which is not arbitrary as well. You can get your first 17 turns after building your first library, and the second 34 turns later. So 3 in 51 turns. Yet you have the third scientist popping out in the renaissance. This could be true of a tech bulb player, but only because they bulbed TO the renaissance!

 

[Bibor]

Hmm, Bibor you have made several errors in your logic.

First of all, you are comparing beakers per turn vs the average techs of an era, as opposed to the most expensive. The first tech I bulb in a game is Civil service.

Secondly:

10 * 2.0 = 20 ; average tech cost around 400 takes 20 turns to pay off.

This is incorrect. You have added the first academy towards paying off the second. You aren't proving that the two combined can pay off a bulb on the second Great Scientist, but rather that the second one is ALSO more valuable to turn into an academy. Therefore it must 'pay off' its own cost. And once again, you should be comparing against the greatest costing techs, not the average.

You have also not factored in the loss of an improvement on the tile in question.

You have also added in the requirement of building science buildings which cost maintenance and hammers. These are adding against the short term costs of not already having the tech from bulbing.

Lastly, while not a mathematical fault, you have made some assumptions about frequency of Great Scientists which is not arbitrary as well. You can get your first 17 turns after building your first library, and the second 34 turns later. So 3 in 51 turns. Yet you have the third scientist popping out in the renaissance. This could be true of a tech bulb player, but only because they bulbed TO the renaissance!

The "paying off" is a mistake, I acknowledge that. But that part is completely irrelevant actually. I don't even know why I left it there.

It's true, I didn't calculate in the tile loss either and building costs either. However It's questionable if the city would actually work those same tiles if they were mere trade post tiles (2F3.5G with marketplace+bank). I also didnt' calculate in the hammer cost for buildings because this is supposed to be a "science city" which focuses primarily and mostly on building science buildings along with happiness/culture when possible.
You're also right that I didnt' calculate in the building maintenance, but it can be easily paid off with trade route and those few commerce tiles the city works.

And lastly, something I forgot to mention, this kind of a city would have a lot of scientist slots, meaning it would also be its own Great scientist factory.

As for the Great scientists early. Okay, yes. You can get them that early. But costs for new ones goes up quickly, new slots for scientists are hard to come by, and burning them that early also reduces their beaker value. Keeping them around also costs you unit maintenance that can easily reach building maintenance costs you mentioned earlier).

Even with all the flaws you found in my calculations, my point still stands: a city that I described will generate a lot of research.

 

[Zechnophobe]

It will, for sure. But your empire will have made more gains by popping them for techs.

It is fun to have a city producing a lot of science. It is cool to look at I guess. But it just isn't efficient in the majority of cases.

 

[fed1943]

And how much pop the science city would need?

6 to work the academies+n to work hammer tiles (as the city must have high prodution) +n to be
scientists (to be a GScientist farm) = ?

The grow time must be factored in and that works against the academy approach IMHO.

If the academies were split among cities than the grow cost would be lower but the hammers cost
would be higher.

 

[avandol]

By the time you have both a university and an observatory you're probably not going to have that many renaissance techs left. There's also the fairly strong argument that a tech now is better than a tech in 100 turns time, and another 150-200 turns after that, especially if you're slingshotting deep.

1. He was saying that this accumulates to a renaissance tech. I f you dont understand a comparison, please dont use it as a counter argument.

2. Your second point is no argument at all, since you completely forgo strategy in favor of tactics. Thats neither good nor very successful.

Basically you are saying here its better to have one free tech now rather than 3 free techs in a hundred turns.
If you'd have won in that time, then yes. In any other case, no.

This really is not that hard to understand. Its number crunching. And either you can do the math, or you dont.

-------

The only real point to discuss is the exact moment where the favor switches to bulb. That is really hard to define, since it depends on a wide set of variables and differs from game to game.

 

[pi-r8]

Yeah Academies on grasslands.

With a University and an Observatory it's a 10 tile that feeds itself.
Or a 20 tile if you add a Research lab.

What a crappy improvement.

Lets see, 100 turns, 10 per tile... 1000. That's a renaissance tech!

Moral of the story: if you have half a brain and place your academy properly (near a city that can build an Observatory), you'll get at least 2-3 techs out of it before the game ends. And please don't counter with "GAME'S ALREADY OVER BY THAT TIME LOLZ" arguments because you need 4 techs to do a horseman rush.

what if you save him for 100 turns, and then bulb a modern era tech? Or use him now on a renaissance tech that can help you a lot immediately?

from what I've seen, each "column" of the tech tree costs about 300-400 beakers more than the column before it. If it takes 10 turns to research a tech, and there's 4 techs in a column, that's 40 turns to advance to the next column. So saving the scientist to bulb later is effectively almost 10 beakers/turn. But actually it get's much better than that, because you can advance on just one part of the tech tree and save multiple scientists to bulb together, which makes it worth more like 20-30 beakers/turn. Settling it just can't compare.

 

[avandol]

And about keeping the scientist in a city ...:
If I use him right away as academy I will have my free tech at the time that "it pays off" (in your example , and how the hell do you actually know when it pays off??) to bulb with him.

I play on epic/marathon speed. It takes ages to research and I regularly play games with several hundred rounds. In that game setting, academy is a must. But again, the interesting point is still when to switch from academy to bulbing.
The argument to have some military unit or a wonder ten rounds earlier is not valid either.
If you lose in the early game due to you foe having musketeers ten rounds before you, you'd probably have lost anyway, since you made some bad choices.
There is no single unit or wonder that is so killer that you have to have it before everyone else does.

In the late game I bulb as well. But that was never the question anyway.

 

[pi-r8]

And about keeping the scientist in a city ...:
If I use him right away as academy I will have my free tech at the time that "it pays off" (in your example , and how the hell do you actually know when it pays off??) to bulb with him.

I would argue that it's always better to either bulb, or save him for a bulb later. There's no magic timing for this of course, you have to figure it out based on the game you're in and what you want. One tech that's extremely good to bulb is dynamite (for artillery)- you can bulb both it and the 2 techs before it very quickly, when normally it would take a long time to research them. I don't know how long it takes to research on marathon, but that shouldn't matter. Even if you settle an academy right after getting writing, I doubt it would pay off in time to match this bulbing strategy in most games.

 

[avandol]

what if you save him for 100 turns, and then bulb a modern era tech? Or use him now on a renaissance tech that can help you a lot immediately?

I fail to see what there is so hard to understand here.
You are able to multiply 5x100, are you?

Stop mixing people, it does not lead to a solution of this problem.

If I can use him on a modern age tech in the foreseeable future, the game isnt in early times any more.
And if I can use him now on renaissance, its not in early time anymore either.

At least try to follow the thread, its not that long.

 

[avandol]

I would argue that it's always better to either bulb, or save him for a bulb later. There's no magic timing for this of course, you have to figure it out based on the game you're in and what you want. One tech that's extremely good to bulb is dynamite (for artillery)- you can bulb both it and the 2 techs before it very quickly, when normally it would take a long time to research them. I don't know how long it takes to research on marathon, but that shouldn't matter. Even if you settle an academy right after getting writing, I doubt it would pay off in time to match this bulbing strategy in most games.

At least try to use numbers when presenting arguments.

Saying "its always better this way" is nothing but wasted breath unless you present some facts.
Same counts for your assumption at the end.

Its very easy to count the beakers you accumulate by building an academy early on. Over the course of the game it nets you in more than one tech, just by itself.

It is not hard to understand. Do the math in your next game.

 

[pi-r8]

At least try to use numbers when presenting arguments.

Saying "its always better this way" is nothing but wasted breath unless you present some facts.
Same counts for your assumption at the end.

Its very easy to count the beakers you accumulate by building an academy early on. Over the course of the game it nets you in more than one tech, just by itself.

It is not hard to understand. Do the math in your next game.

did you see the part in my previous post where I did the math to justify my argument? I avoided using numbers because you said you play marathon, and I don't know what the tech costs are in marathon.

In standard speed, a modern era tech costs about 3000 beakers. The game will probably end in less than 300 turns after the first great scientists comes out (more like 200 turns after). So even if you can get 10 beakers/turn as soon as you build the academy, that's still not enough to justify the acadamy over simply saving him for a while and then bulbing later. Not to mention the massive benefits you can get from having a tech like dynamite 20-30 turns early.

edit: forgot about overflow beakers. Most of the time you won't get a tech exactly, so a lot of normal beakers get wasted. Bulbs, however, always get you the full value of the tech, with no waste.

 

[RealHuhn]

[...] The other thing is that you must have suprlus population to work the academy or you get nothinng for it. [...]

Surpisingly, this isn't true. I was pretty sure that an academy on a grass tile reduces the food from 2 to 1 but it does not. It remains at 2 + 5. Maybe a hidden change ?!?!

I've played a few games until the renaissance and two academies are probably worth it.
But the science city needs a few requirements:
* Grassland for growth and academies
* Hills for production so you can switch from science to production while building the college, university etc.
* A mountain tile for observatory.

The best policy tree is probably freedom because your is limited and 1 per 1 is awesome.

With two academies, teching speed to the renaissance is very very fast, backfilling is crazy .
But it's very difficult to balance early scientists against early expansion:
* The science city cannot produce anything but or the required buildings, maybe an odd worker here and there but that's it.
* 2nd city needs a library as well for the national college (which should be built after your first academy, imo)
=> Lack of military units and cities ...

Purchasing the university with gold is another possibility but you often have to neglect city states etc.

You guys should give it a try before saying it can't be done

Which is the best civ/leader?
I'd say Ghandi for happiness, France for early culture (monuments less important), Siam for buffed maritime CS, maybe China for gold ?!?!