Can someone explain exactly how bulbing mechanics work ?

[Lord Yanaek]

If, by exploiting overflow, you would obtain free research of an amount equal to or greater than the amount by which you "miss" completion of Acoustics on the next-to-last turn (in ACTUAL BEAKERS, ignoring overflow on that turn), in the case of teching it by some other route, then switching to an overflow path will reduce your turns to Acoustics by 1 or more turns. And not otherwise.

Say what?

As long as you don't bulb something that has only a few turns left (1,2 or 3) you're fine.

Well, at least i can do this without adding a mathematical coprocessor to my brain

 

[Acken]

Not done further testing but on reflection perhaps my overflow in this case was capped at the last tech researched. In my example it was PP which has a base cost of 8800 beakers. This will need to be verified in code but perhaps the mechanics when you click 'next turn' is this:
Empty all overflow into selected tech, and if the tech is gotten that turn determine the new overflow(old_overflow + bpt -tech_cost). Decide based on base tech costs what the capping amount will be(either 5*bpt or last tech base cost, whatever is greater).

My understanding is that if overflow is bigger than the tech then yes it can get capped.

 

[HorseshoeHermit]

@Bob Morane:

Bob. There will be a turn when the game will say "1 more turn" until Acoustics, for your plan of teching there directly. On that turn, there is a certain difference from the actual cost of Acoustics, and your research in Acoustics.
If THAT AMOUNT is less than the free beakers you would get by exploiting overflow, then exploiting overflow will get you Acoustics faster. And if THAT AMOUNT is greater than the free beakers you would get, then exploiting overflow will not get you Acoustics faster. If the amounts are equal, then it won't get you Acoustics faster, but it does get you Acoustics + other stuff by the same point.


I am appreciative of the desire to test , as experiment is the crucible of thought.

And I misspoke some times.

Guh, I got the cap inverted. The cap is the greater of 5*bpt and the cost of the tech, not the lesser.
This still doesn't explain what phillipwyles observed. However, before I blame IGE, I have to ask more detail about what the setup was. Why were you sure your overflow was 25K? Why was Mobile Tactics relevant? I though you just gave yourself every tech before PArticle Physics.

----

This process has six parts, which depend on six values. I've scrapped the diagram and am just going to program a simple interactive animation to show the computations involved. I just keep getting sidetracked with insights into Design I'm having while designing it.

• The cost of the tech
• The research modifier for the tech.
• The modifier to the cost owing to the science penalty
• The cap applied in the April patch
• Your beakers per turn
• Overflow

The parts are:

1. The actual cost determination of the tech
2. Accumulation of beakers into the tech
3. The transfer of overshot beakers through the cap
4. The transfer of capped beakers into overflow
5. The movement of overflow into accumulated beakers each turn
6. The modification of those overflow beakers by the research modifier

0. Research modifier is a modified modifier. It's not a modifier group, I got this wrong before.

[Research modifier] = [Civ tech diffusion modifier] * ( (100 + [World Congress tech modifier]) / 100)​

[Civ tech diffusion modifier] = (Number of met civs know the tech) * [TECH_COST_TOTAL_KNOWN_TEAM_MODIFIER, from the Defines table] / (Number of alive civs)
[World Congress tech modifier] = 20 [U]if [I]Scholars in Residence[/I][/U] 0 [U]otherwise[/U]

1. The actual cost is given by:

[The cost defined in the database] * (100 / (100 + [The Research Modifier]) ) * [The science penalty]

This cost defines a moving finish line for beakers accumulated into this tech. Except for passing and then repealing Scholar's in Residence, or someone reviving a civilization who doesn't know the tech, this finish line only gets -closer- to you and your beaker count.​

2. When you get beakers, you put them into this tech, all at once. This is e.g. when you bulb, when you spread the Interfaith Dialogue religion, and when your turn starts. By adding this amount to your current research, you may well exceed the finish line. That excess is recorded and fed into a process.

3. The excess is processed as follows. The amount is capped by the following value:

MAXIMUM([the cost defined in the database] , (5 * [your bpt]) )

That is, if the excess is greater than both, instead it becomes the greater one of these two.

By "your bpt", I mean the amount sometimes called "Your science". The measure of change in beakers that applies to your empire, which is conveniently made visible for you at the top of your screen after you settle your original capital.​

4. The capped excess is added into your overflow.

5. When your turn starts, you get beakers automatically pushed into your current research. The amount of beakers added is given as:

[Bpt]+ [B]Modified overflow[/B]

6. Overflow is modified by taking:

[B]Modified Overflow[/B] = [B]Overflow[/B] * ((100 + [Research modifier]) / 100)

Your overflow is set to zero when it is pushed through this modification into your research.

I'm gonna make a picture, but it should only help with visibility and digestion of the information. I have no resource to offer those who do not know what the preceding sequence means.


-----
I didn't mention the moving finish line trait before. Even though the finish line moves, the cost of the tech is actually only calculated at the times you gain research. So if the cost moves under your research level, you won't get the tech until the next time you get a beaker.* Actually, being awarded 0 beakers, depending on whether the programmers added a check like "If this is 0 do nothing" to the Interfaith Dialogue procedure, would be enough to get the tech.

This can -actually- happen if you meet a civ who knows the tech, or a met civ learns the tech, as doing so increases the research modifier by , apparently, 5% in a 6 player game, which will reduce the cost by (1 - 1/1.05) around 4.8% (of base). Differences of 2 beakers are common in my games for late ancient learns; for techs like Mobile tactics that's 440 beakers

 

[Acken]

Could you confirm that if Bpt + Modified Overflow > Tech_cost then the excess can get capped ?

Otherwise good and precise post.

 

[KoKid90]

@Bob Morane:

3. The excess is processed as follows. The amount is capped by the following value:

MAXIMUM([the cost defined in the database] , (5 * [your bpt]) )

That is, if the excess is greater than both, instead it becomes the greater one of these two.

MAXIMUM([the cost defined in the database] -> Is this the cost of the last tech researched ?

 

[KoKid90]

Could you confirm that if Bpt + Modified Overflow > Tech_cost then the excess can get capped ?

Otherwise good and precise post.

Isn't bpt + modified overflow after the cap amount possibility ?

 

[HorseshoeHermit]

MAXIMUM([the cost defined in the database] -> Is this the cost of the last tech researched ?

That is the cost of the current technology, at the time of evaluating that function.

Could you confirm that if Bpt + Modified Overflow > Tech_cost then the excess can get capped ?

Otherwise good and precise post.

That is a very open question because of the word 'can'. If you meant 'does', then this conditional is not valid. If you research the whole tech when applying bpt and modified overflow, but do not overshoot by more than the cost of the tech or 5* bpt, you do not get capped. But if it does overshoot by that much - perhaps because the overflow is quite excessive (because the exploit still exists!), perhaps because you had only 1 beaker remaining and overshot with a balanced amount - then it does get capped.


Civ didn't tell me this thread updated when Acken replied 3 months ago.
And I'm not drawing a picture. The geometry I had in mind is not accurate.

Isn't bpt + modified overflow after the cap amount possibility ?

The sequence of events you actually experience starts at (5), cycles back, and ends at (4).

 

[KoKid90]

That is the cost of the current technology, at the time of evaluating that function.

And it is the initial cost no matter how much is left in it right ?

That is a very open question because of the word 'can'. If you meant 'does', then this conditional is not valid. If you research the whole tech when applying bpt and modified overflow, but do not overshoot by more than the cost of the tech or 5* bpt, you do not get capped. But if it does overshoot by that much - perhaps because the overflow is quite excessive (because the exploit still exists!), perhaps because you had only 1 beaker remaining and overshot with a balanced amount - then it does get capped.

Ok I think I understand it now

 

[HorseshoeHermit]

And it is the initial cost no matter how much is left in it right ?

Yes, the database-defined value for the cost of the technology which is determined at the time the game is initialized from its assets.

 

[Nonevah]

... I really wish that they hadn't stopped the overflow exploit in such a stupid way.
I, personally, think it would have been best to change the already researched by others bonus to a percent decrease in the cost of the tech, not a percentage bonus to research against it. Or would that have just been way too complicated?

 

[HorseshoeHermit]

... I really wish that they hadn't stopped the overflow exploit in such a stupid way.
I, personally, think it would have been best to change the already researched by others bonus to a percent decrease in the cost of the tech, not a percentage bonus to research against it. Or would that have just been way too complicated?

That is exactly what happens actually.
The problem is that the bonus gets applied a second time (for no reason), and in the accelerating direction (for no reason).

Also, the problem you see developers give about overflow amounts... it's strange. They are right that overflow needs limits, but the reason I've ever heard gamedevs give is wrong. The problem is that overflow is a floating amount that lets you choose where to put the research after the fact instead of at the time - which isn't how hammers or beakers normally accumulate to your use. It's not that it gives you multiple things in one turn, that is just clearly not.... I mean if it did that would be closer to solving the problem. If you have enough hammers to build things then you should get them. The problem is not being forced to set those things in advance.

Gal Civ actually removes overflow, which takes us back to microing on build finishes, for this actual stated reason - they don't want multiple things you really built in one turn - this is Word of God.
/offtopic

But the third reason the fix is bad is because bulbs almost always end up 50% or more in overflow, naturally tend toward which is not regulated in Civ the way it ought to be. nah that's too inaccurate, ignore that.

MAIN REASON this doesn't work is because some programmer or his design document had a straight-up math derp. Twice.
edit: see, each of those three things would be a functional way of handling research modifiers and overflow. Except using ALL THREE of them is just.... derp. See, if the research was accelerated rather than the cost reduced, then if the overflow was decelerated into the next tech, it would cancel out precisely the extra beakers just from the bonus. It's a great design. But..... communication derp.

Derps. Derps everywhere.