Formula for calculating GB rewards?

[DeletedUser33220]

You think you get it, but your sophomoric understanding of the math involved is limiting you. There are multiple variables that they added in to balance buildings, some of which were done retroactively, some kick in to varying degrees at varying levels. I'm not going to sit here and pretend that I can spout the math off the top of my head, but I can tell you that it's more difficult then something you can plug into your calculator.

It's because there is not a set formula, it's a more complex algorithm that changes as the building progresses in level.

Ahh, now we are getting somewhere. Splendid, let's both continue as adults (aside from your continued attempts to belittle my intelligence, regardless the fact that I'm still not entirely convinced you understand what you're talking about, let alone what I'm talking about).

No. It is a formula, not an algorithm. Believe me - if you want a computer to do it, you write code to do it. Sometimes the formula is short, sometimes it is long, and sometimes people improperly refer to it as an algorithm - but there's still a single, if not simple, formula. This formula is not arbitrary, and is the same for all buildings of a given age. Don't believe me? Pick any two buildings from the same age, and compare their level costs and their contribution rewards. They're the same, promise.

As previously mentioned, it's very likely the exact same formula for all great buildings across the game, with a different base value depending on the age of the building.

 

[DeletedUser32389]

As previously mentioned, it's very likely the exact same formula for all great buildings across the game, with a different base value depending on the age of the building.

It changes based on level. Say, one formula for levels 1-10, then another for 10-30 and so on. If you go back through your buildings you'll see what I mean. Different buildings have different cut-offs but they all have one at 10 for sure.

aside from your continued attempts to belittle my intelligence

I get testy when a person with 10 posts comes around and acts like I'm an idiot for trying to help. Had I TRUELY understood what you were asking I would have never replied. "Stop being a bitch" is number one on my to-do list

 

[DeletedUser29726]

It changes based on level. Say, one formula for levels 1-10, then another for 10-30 and so on. If you go back through your buildings you'll see what I mean. Different buildings have different cut-offs but they all have one at 10 for sure.

They're not so much cutoffs (though that's an easy way to look at it), but the artifacts of rounding. Typical progression for 1st place reward on a building goes like:

up by 5 each level
up by 5 some levels and 10 some levels, more frequently becoming 10 as you go through levels
up by 10 each level
up by 10 some levels and 15 some levels, more frequently becoming 15 as you go through levels
up by 15 each level
etc.

what this is really representing is it going up by nonmultiples of 5 that are overall increasing but getting rounded off. The prerounding formula would intrinsically cover all the "cutoffs".

Edit: I would add, it is quite possible pre-10 is a different formula or just a table but post-10 is almost certainly following a single set formula for each of FP cost, and 1st place reward. For total FP cost that formula is CEILING(FP10*1.025^(level-10),1) i believe while pre-10 is a very different formula or possibly just a table even.

Edit 2: This one is easier to figure out because it's rounding off to the nearest 1 on numbers in the thousands. The rounding doesn't obscure much of anything. The FP rewards though are starting around 100 rounded off to the nearest 5 around level 10. And don't change that fast as to the rate they're increasing. It'd take plotting of best fit curves for multiple buildings to get a guess as to the underlying type of formula it is, followed by regression fits, and then much futzing about to turn your regression factors into numbers that catch each rounding right and determining what rounding method is used (floor/round/ceiling).

Or datamining the code itself (if the client in fact knows how to calculate the numbers and isn't just fed them by the server).

 

[DeletedUser33220]

I get testy when a person with 10 posts comes around and acts like I'm an idiot for trying to help. Had I TRUELY understood what you were asking I would have never replied. "Stop being a bitch" is number one on my to-do list

Fair enough, but perhaps that should serve as a reminder to not judge a book by its cover. Just because I don't post much doesn't mean I don't know what I'm talking about.

 

[DeletedUser29726]

simple regression fits for arc's 1st place payout for levels 11-147 (x= level)


Quadratic is really close. I'll see what the numbers look like for another age's building with quadratic fit (will be over a lot less levels)

 

[DeletedUser29726]

CdM over levels 11-86

Quadratic is still very close. The acceleration term (x^2) is pretty close to the same and the difference could be rounding artifacts and the other two terms are pretty much just defining the situation at level 10 which would be what is defined by age.

 

[DeletedUser33220]

They're not so much cutoffs (though that's an easy way to look at it), but the artifacts of rounding. Typical progression for 1st place reward on a building goes like:

up by 5 each level
up by 5 some levels and 10 some levels, more frequently becoming 10 as you go through levels
up by 10 each level
up by 10 some levels and 15 some levels, more frequently becoming 15 as you go through levels
up by 15 each level
etc.

what this is really representing is it going up by nonmultiples of 5 that are overall increasing but getting rounded off. The prerounding formula would intrinsically cover all the "cutoffs".

Edit: I would add, it is quite possible pre-10 is a different formula or just a table but post-10 is almost certainly following a single set formula for each of FP cost, and 1st place reward. For total FP cost that formula is CEILING(FP10*1.025^(level-10),1) i believe while pre-10 is a very different formula or possibly just a table even.

Edit 2: This one is easier to figure out because it's rounding off to the nearest 1 on numbers in the thousands. The rounding doesn't obscure much of anything. The FP rewards though are starting around 100 rounded off to the nearest 5 around level 10. And don't change that fast as to the rate they're increasing. It'd take plotting of best fit curves for multiple buildings to get a guess as to the underlying type of formula it is, followed by regression fits, and then much futzing about to turn your regression factors into numbers that catch each rounding right and determining what rounding method is used (floor/round/ceiling).

Or datamining the code itself (if the client in fact knows how to calculate the numbers and isn't just fed them by the server).

I would agree that the first 10 follow a table. So near as I've been able to figure, and validated by a post on EN, is that for the first 10 levels, the number of FP to level appears to be "set", and then it increases my 2.5% per level after that, such that:

FP to level = #FP from level 10 * 1.025^(#level - 10)

So, as an example, for level 27 of the arc,

x = math.ceil(970 * 1.025 ^ 17)

 

[DeletedUser29726]

Left out one of the simple options in the earlier plot - power fit also works very nicely

 

[DeletedUser33220]

CdM over levels 11-86
View attachment 11304
Quadratic is still very close. The acceleration term (x^2) is pretty close to the same and the difference could be rounding artifacts and the other two terms are pretty much just defining the situation at level 10 which would be what is defined by age.

I appreciate that effort, but unfortunately, for my purposes, very close will not take the cake. Unless progress is made, or someone else comes along with more precise info, I may just unfortunately be stuck with manually entering the payouts for the 1st spot for each level and calculating the remaining spots.

 

[DreadfulCadillac]

This is what people are wasting their time on? You guys need to find something productive to do with yourselves, no offense

 

[DeletedUser29726]

Power fit for CdM:


The exponent appears to be a constant and the k in front would vary by age.

 

[DeletedUser29726]

This is what people are wasting their time on? You guys need to find something productive to do with yourselves, no offense

Playing with math is no less productive than playing FoE And more interesting than the direction they're taking it with settlements!

 

[DeletedUser33220]

This is what people are wasting their time on? You guys need to find something productive to do with yourselves, no offense

Trust me, once we have the answer, it'll be plenty productive.

 

[DreadfulCadillac]

Trust me, once we have the answer, it'll be plenty productive.

How does it help, to many pages to read through, can u just give me a quick overview of what this is about?

 

[DreadfulCadillac]

Playing with math is no less productive than playing FoE And more interesting than the direction they're taking it with settlements!

I Dissagree i like the settlements

 

[DeletedUser29726]

Yea it'll take a fair bit of work to find out the extra little terms. I'll play with it some more off and on and let you know if i come up with anything

 

[DreadfulCadillac]

How does it help, to many pages to read through, can u just give me a quick overview of what this is about?

can u plz explain for me

 

[DeletedUser32824]

I appreciate that effort, but unfortunately, for my purposes, very close will not take the cake. Unless progress is made, or someone else comes along with more precise info, I may just unfortunately be stuck with manually entering the payouts for the 1st spot for each level and calculating the remaining spots.

This is probably going to save you time in the long run. I've typed in tables for 80 levels for like 5 different GBs to calculate leveling them. While it is annoying, it's 100% accurate and much easier than reverse engineering a code.

The only time this wont work is when you want to calculate an unpopular GB to level (like I did with Voyager and Dynamic Tower). In that case I found a friend on my world with that GB leveled and manually looked through their GB and copied down the numbers.

In regards to people's attitudes in this thread: I would caution against calling people names, as it tends to get things off topic really quickly. If someone answers the question they think you asked because they misunderstood, just be nice and better explain yourself if you want that person's answer. If you don't want that person's answer just ignore them or thank them for their time.

 

[DeletedUser26120]

I've used this, seems to work well

https://foe-tools.github.io/gb-investment/

 

[DeletedUser29726]

can u plz explain for me

He likely wants to work on optimising a strategy for GB levelling and was hoping not to copy out the 1st place rewards for hundreds of levels of multiple ages and track them down himself. Having already copied out the 1st place reward for levels i'm interested in and using interpolation between neighboring ages for unpopular buildings has been 'good enough' for me. But I also like a fun math puzzle, so i'll tinker with it some