Help me solve this

I have already tried to contact the creator, and no one in the community knows about it

Sure 0.999 isn't bad, but everytime I try to calculate a value, it fails by too much.

When x=1, y=3000, when x=2 y=3394, and so on, whatever the formula may be, that's where I need help because all the formulas I came up with are not 100% precise and the creator couldn't have picked random values

Also, if you had read the other comments that I already posted, you would know what these values would represent, so read before asking.

If I gave you the values 1 2 4 8 16 32 64, I'm pretty sure you could figure out a formula, so this isn't any different.

programmer probably just use some random gen formula or something
pic related

if the numbers have to be integers, then the formula might involve rounding, or flooring

Do you not know what Excel is?

here is a less cluttered image of the plot

I mean, you can easily fit this curve with a polynomial, but it's not very interesting.

These are stock/crypto prices aren't they

It could be related to the distribution of prime numbers or something like that

OP here.
To
and also the rest:
Guys, I obviously had already tried all of these methods before, they are like the easiest and most obvious methods anyone would try when analysing a sequence, so of course I tried them immediately before trying other methods and eventually coming here to ask for help..

I had also figured out formulas and such, but as I said before, I'm not looking for a function with a correlation coeficient of 0.999, I already did that ages ago. Instead, I need a formula with 100% accuracy or I can't predict values accurately.

As expected, no one was able to help me, I think I might give up on this problem :/

I doubt they are, since they are from a game as I mentioned earlier.

Again, you are extrapolating and except you have all the numbers and the inputs It's literally impossible to get what you are looking for. Give up you manchild.

Shit trick that helped me once:
>dump them in excel
>graph them
>on graph properties, select show equation
>try n-degree polynomial

Will be not perfect, but helped me to aproximate the temperature function for a thermistor in a voltage divider
> GND---10K---Vx---Thermistor---VCC
I logged the voltage and temperature measured with a lab thermometer while heating the medium...

I think I gave you a decent list though, with like 40 different values or something..
As I mentioned before, if I gave you something like 1 2 4 8 16 32, you could guess it in seconds, so this is only a harder formula than 2^n but it should still be an existing formula, and I'm not sure if the developer generated random numbers to make a giant equation like the ones that have been suggested or that I found myself.
It should be something smaller and more elegant, I'm guessing.

But maybe I should really give up, I guess this problem is too hard for me and for anyone else in this board..

Thanks for the suggestion, but I also tried that already, along with other similar stuff, and only got formulas with 99.99% accuracy, and I was looking for a 100% one which I guess is a bit hard..

Oh, I see you've tried...
The curve is too complex to just establish the equation by looking at it, try looking something up under analytic geometry...

There is literally an infinite list of functions that fit the numbers you give. Look polinomial interpolation. Without any more information, it's impossible to give you more than a bunch of conjectures. It's not that hard to understand. It's like I give you the list 1 4 54 90 pr whatever and ask for something that fits exactly. The problem is that you will not be able to predict larger values. Basically you are asking that we infer from empirical values a clear rational mathematical formula. It's basic shit fag.

Anything that you could come up with that looks "elegant" enough is a solution that I would like to see, that's basically what I'm looking for.
I know it's hard because of so many functions that can fit them, but if 99% of the functions look super weird and only 1 or 2 look really "elegant", those 2 are most likely the correct ones, if you see what I mean.
If I wanted 99.99% accuracy, I wouldn't be asking for help here, because I already did that ages ago. It's just that there HAS to be something with 100%, whatever it may be.
If you think it's too hard for you, then don't bother I guess, but I still accept suggestions.

I'm just so intrigued because if you look at my previous comments, you will understand that this is supposed to be an equation used by a person in his code, and I can't believe that person used such a weird and complex equation, I expected it to be something simple but after trying all the possible methods that I know of, I still fail..
It's a big mystery to me..

Try this
math.stackexchange.com/questions/121212/how-to-find-curve-equation-from-data
Looks promising

Why does it matter if it's elegant? The programmer couldn't care less about that. What you need is something that gives you the exp at any level so you can predict then? Then elegant means jack shit. If you are just looking as to how this shit is done, then dive in the code and look other examples. It's coding, not pattern recognition. It's an impossible problem.

it would be safe to assume these numbers are chosen by the developers so as to deliver a satisfying gameplay experience.

you are probably retarded. that's ok though.

Sure, but if you were programming a game, would you use an equation like 25x^2+10x or would you use 5322.458x^3-453.2999x^2-5523.7 in your code?
I mean, I don't see the point of using such a weird and random equation, I personally would use something more consise.

So you are saying ALL of them were manually generated instead of belonging to some formula or something?
This game has levels up to 500 or more, I doubt this guy would have picked values manually for every single level, and I'm not sure why anyone would use such a weird equation in their code.

>99.9% correlation isn't accurate
you would actually fail any STEM lab course

i think this might just be a really indirect 0.999... = 1 troll.

Mathematician here. i work in Mathematical computer modeling, like what you are trying to do. First thing you should note:
if you have a graph which fits with near 100% accuracy, that is the right graph.
Next, the exponential function gives irrational values when you input an integer. it looks like the game Dev literally just cut off the number after the decimal and used that was the exp growth curve.

your tenacity for trying to get to the bottom of the problem is admirable. try using that on actual STEM instead of a video game.

Ok fair enough xd

Thank you very much for this comment

Our theories to describe the universe shouldn't be "just" 99.9% accurate though, or else they would just be an approximation :/
That's how we went from newton's gravity equations to einstein's gravity equations, because newton only did the 99%

You are such a fucking faggot.

Do you not agree? ._.

things can only be accurate within experimental observation. Newton didn't have access to the same level of measurement equipment Einstein did. Newton also had to invent calculus. Einstein didn't need to invent any of the math used in his theories. also, only recently have we been able to detect some of the things Einstein predicted.
this guy is right.

That's not my point.
I'm not saying Newton was bad or shouldn't be respected, because I think he was very important and his work was essential for the development of modern theories.
My point was, we should ALWAYS try to improve our models and try to reach that "100% accuracy", instead of saying "meh, 99.9% is fine, no one will notice"
I don't see why this is wrong..

Could be something using [math] \log\Gamma [/math]
It's a really convenient function for exp values

t. used to dev some shitty games in high school

you are welcome. carefully thought out and respectful comments raise the quality of the board.

I'm not arguing that your ideal is wrong. your heart is in the right place. This is my point:
it is impossible to achieve 100% experimental accuracy. measurement uncertainty is built into literally every experiment. moreover, it is not possible to prove that our Mathematical models are fact or just extremely good approximation. i agree with you that this is sad. as i said here i am a Mathematician. it is unsettling to know if what you do is grounded in reality, or just unsettlingly good approximation. you may find this interesting en.wikipedia.org/wiki/The_Unreasonable_Effectiveness_of_Mathematics_in_the_Natural_Sciences

please read the thread before posting

youve demonstrated a profound lack of understanding of a variety of topics, im actually impressed how quickly my opinion of you sullied. kill yourself.

Pic related; It's from a shitty rpg I made years ago. The growth rates are extremely similar.

comp.dit.ie/smulligan/interpolation.pdf

You can really only guess the precise formula. Don't forget that the values you are regressing are probably rounded...

Thanks for the suggestion, I'll try to work with that

Alright, thanks again

Based on?