Does anyone know the in-game formula for calculating the total experience required at each skill level?
Lew Zealand
Helper
10-term polynomial
aka: arbitrary
I will also accept the following as answers:
random from dartboard toss
total g earned per year from Salmonberry season
digitized Brownian motion
C-14 decay times in years for 10 actual atoms
number of minutes between original thoughts in my head
Cosmic ray energy values in 10MeV increments
aka: arbitrary
I will also accept the following as answers:
random from dartboard toss
total g earned per year from Salmonberry season
digitized Brownian motion
C-14 decay times in years for 10 actual atoms
number of minutes between original thoughts in my head
Cosmic ray energy values in 10MeV increments
Boo1972
Farmer
I suspect it arises from something like how many parsnips seem reasonable to plant and harvest to raise your farming skill to the next level. And then the rest of the skill levels fell in line with farming. Maybe my teenagers can figure out an equation with those fancy calculators of theirs.
ShneekeyTheLost
Farmer
In the actual code, it's an array, not a formula. I don't believe there is a f(x) for the progression.
Lew Zealand
Helper
All data can be fit to a curve! The entirety of the known Universe* is merely a slightly more complex than normal wave function being actively manipulated by miniature office cubicles full of hamsters and maybe gerbils! Could be nutria though, the office is located rather far away and I only have a cheap telescope to spy on them. But no matter! We shall not slump into despair at the feet of merely mostly intractable polynomials, instead we shall push on to find a talented mathematologist to untie this knot of some annoyance.
My kid is taking some sort of standardized test for math, maybe I'll ask him. What's the reward for solving a 10-term polynomial? Ice cream? Ice cream is always good stuff, maybe I'll give him a second scoop if he keeps the R^2 above .97
*they also work on those unknown universe bits. But only when the boss gets the OK for overtime pay.
SurprisedPikachuBob3
Farmer
Look at the wiki
ZythiQ
Greenhorn
So, I thought I posted here... guess I didn't—anyhow, ended up calculating a useless 10th-degree polynomial: -(5/18144)x^9 + (61/2016)x^8 - (3125/3024)x^7 + (279/16)x^6 - (145145/864)x^5 + (31083/32)x^4 - (1905110/576)x^3 + (3382439/504)x^2 - (840545/126)x + 2600
Thanks for the comments tho!
Thanks for the comments tho!
Lew Zealand
Helper
Ah, the new and improved emergency services number! Replaces the old and outmoded 0118 999 881 999 119 725...
3
3
SurprisedPikachuBob3
Farmer
I LOVE THAT SHOW
DraconianFleet
Greenhorn
Apologies for the necro, but I was also curious to the answer for this, and while I wasn't able to find an answer online. I was able to use desmos to get a function that would fit within +-15 exp for most levels, for the most part it is theoretically scalable for however many levels you need for mods too.
Explanation:
levels 3, 4, and 5 were a little mean, and I don't think there is a good solution for them, they probably have a lower exp threshold so that it is easier to get your first perk than it would otherwise be. To scale the function just add terms with the format (x/n)^(2.5n) where n = the level. The level 3 and 5 exponents had to be changed so that it generally fit, if someone can make a better curve, then I would love to see it. Another note is that aside from the first 5 terms since they are the base function, every 3rd term after that is a negative term, I could also modify the constant to get it to be a general better fit, but I am tired at this point. There has to be a more elegant way, it looks so clunky, and while I was able to get generally "close" with more elegant solutions, none of it was nearly as close as this.
TL:DR, the function:
-10 + 16x + 60x^1.5 + 26x^1.56 + (4x)^2.239 + (x/1.5)^4.575 + (x/2)^5 - (x/3)^7.65 + (x/4)^10 + (x/5)^10 - (x/6)^15 + (x/7)^17.5 + (x/8)^24 - (x/9)^22.5 + (x/10)^25
(is there a better way to format math here?)
edit: I have just realized my level 8 term is also not following the pattern, oops.
Explanation:
levels 3, 4, and 5 were a little mean, and I don't think there is a good solution for them, they probably have a lower exp threshold so that it is easier to get your first perk than it would otherwise be. To scale the function just add terms with the format (x/n)^(2.5n) where n = the level. The level 3 and 5 exponents had to be changed so that it generally fit, if someone can make a better curve, then I would love to see it. Another note is that aside from the first 5 terms since they are the base function, every 3rd term after that is a negative term, I could also modify the constant to get it to be a general better fit, but I am tired at this point. There has to be a more elegant way, it looks so clunky, and while I was able to get generally "close" with more elegant solutions, none of it was nearly as close as this.
TL:DR, the function:
-10 + 16x + 60x^1.5 + 26x^1.56 + (4x)^2.239 + (x/1.5)^4.575 + (x/2)^5 - (x/3)^7.65 + (x/4)^10 + (x/5)^10 - (x/6)^15 + (x/7)^17.5 + (x/8)^24 - (x/9)^22.5 + (x/10)^25
(is there a better way to format math here?)
edit: I have just realized my level 8 term is also not following the pattern, oops.
Lothengriol
Farmhand
Oh no, I should not have been allowed to look in here.
As Lew Zealand hinted above, there is a ninth-degree polynomial equation (with ten total terms) which can exactly fit this data. Below is the polynomial in question.
ZythiQ's equation has a typo, it should be a 567 in the denominator of the x^3 term instead of his 576 (correct form is below)
(-5/18144)x^9 + (61/2016)x^8 - (3125/3024)x^7 + (279/16)x^6 - (145145/864)x^5 + (31083/32)x^4 - (1905110/567)x^3 + (3382439/504)x^2 - (840545/126)x + 2600
This is found through a process known as Lagrange polynomial interpolation, and you can make an nth degree polynomial equation for any (n+1) points you want. For more on Lagrange polynomial interpolation, this is a good starting point: https://wmueller.com/precalculus/families/lagrange.html
If you go any further, you might end up taking a class on Numerical Analysis, which will be good fun if you like math and modeling. But as ShneekeyTheLost pointed out, the code just uses a table (array) instead of an equation! But hey, at least we had fun with math.
Alright, at least I had fun with math.
Lew Zealand
Helper
Math exists to have fun with. Different people have different ideas about what fun is but when they forget to properly describe their favorite math it's usually a simple oversight.