ZSR Forums

Grunz and AKA would be helpful here..

I'm messing with bushes to try to determine how the RNG works

from what I can tell, the only things that matter are

1. which frame the bush is destroyed
2. which bush it is (if you kill 2 different bushes on the same frame they will have different results)
3. some sort of system where it changes the drops based off of what you have (if you have full hp, you will never find a heart or fairy)
4. method of destruction ? (

what does not seem to matter:
1. any input that does not directly affect the bush. example: drop a bush on a certain frame. doesn't matter what buttons you spam while it's falling, the item inside will not change

if I am wrong about any of this I would be delighted to know.

yeah you also have to shove the controller in your ass and fart to manipulate stuff

I took the time today to set up an experiment with the 5 bushes around the tree in Kakariko. I had a savestate near the bushes and would just go up and break them all and observe how many bombs I got. I did different things to keep it as random as possible, but kept one thing constant: I shield dropped them all instead of throwing them (next time I'll throw them all and see if there is a large variation). I talked to Navi, talked to the dudes, picked them in different orders, held them over my head and ran around, I would press cojiro or the bomb button, etc. Just a bunch of stupid random shit. I did this 200 times to give me a total of 1000 bushes broken. This actually took less than an hour, so if anyone else wants to try it, just do the RBA route until you get to these bombs and have at it. Just be sure to keep good results so it can be added with my data.

Here were my results:

-I got 0/5 bombs 148 times.
-I got 1/5 bombs 44 times
-I got 2/5 bombs 8 times
-I got 3+/5 0 times.

This means I saw a total of 60 bombs in 1000 bushes, for a probability of hitting a bomb being 6%, or 1 in every 16.67 bushes. 1000 bushes isn't a huge sample, but I will continue to add to it and see how the percentage changes. My guess is it will stay around 6% and maybe change by +/- a percent or two.

Since I grab 2 bushes in my segments, I think that means I continue to try the superslide about once every 8.33 attempts at the bombs. And someone tell me if I am not doing this correctly:

The chances of getting 2 segments in a row with a bomb drop would be: (1/8.33) * (1/8.33) = 0.0144 = 1.44% chance.

EDIT:

I did a throwing bombs test today. I threw them with the A button. My results were quite similar:

-I got 0/5 bombs 147 times.
-I got 1/5 bombs 47 times
-I got 2/5 bombs 6 times
-I got 3+/5 0 times.

This means in this sample I saw 59 bombs out of 1000, or 5.9%. This is very good as I was doing things quite randomly in both cases and both cases show up just one bomb difference. This most likely means the variance is low over large samples and therefore if you were able to come up with something like 100 bombs per 1000, you'd be on to something.

I will continue to do some random testing and see what kind of results I get. Then maybe try some kind of test where I control as many things as possible.

You're talking about one specific kind of luck manipulation. Something which haven't fully grasped, all these methods mentioned definitley play a factor. I've also heard of what angle you slash the bush also counts as well. One thing I've learned about about enemy and cucco positions is that you can manipulated during a transition by the holding and moving the analougue into different positions. Button presses play no factor. Don't forget that spread manipulation also plays a factor when manipulating rupees.

Now that I've mentioned it, I wonder if moving the analouge about while exiting Link's house could affect manipulation. It's easy to test via use of hex editing (providing you know how to hex).

wouldnt the way to figure this out be to look at the code and to see where the random variable comes from so that you could see how the requirements change