1 Guise in 24 hours? no manastone in 32 Hours

[Dolalin]

Imagine being outraged at the randomness of the random number generator. Jesus wept.

[Tecmos Deception]

Originally Posted by Dolalin [You must be logged in to view images. Log in or Register.]

Imagine being outraged at the randomness of the random number generator. Jesus wept.

To be fair, the randomness of P99's RNG IS pretty fucking wonky.

[fadetree]

Originally Posted by Tecmos Deception [You must be logged in to view images. Log in or Register.]

To be fair, the randomness of P99's RNG IS pretty fucking wonky.

Yep. Specifically, it is globally random but not very random locally, meaning that it's randomness has a very coarse ganularity - for a small sample size you will generally get a much lower randomicity than you will for a large one. What this means in our terms is that there can be much more streakiness than you would expect from a smoother stochastic function. Meaning 'suck it, expect long stretches of absurdly bad randomness'. I don't know if this is because of the algorithm they use, or if the entropy gathering is bad....but some of it probably comes from compressing a random range into a 'fail/succeed' binary. We never see the underlying range or the set of fail rolls vs. the set of success rolls, so it looks super streaky to us.
I am also suspicious about the pattern I usually see when TS-ing - set of 10-15 fails followed by a set of 3-5 successes. This pattern goes on a lot and it implies that there is somethign else going on underneath that has some kind of internal state.
And yeah, I don't get why people want these items so badly. Seems nuts to me, I don't think they are particularly useful enough to justify.

[Isomorphic]

Originally Posted by Tecmos Deception [You must be logged in to view images. Log in or Register.]

To be fair, the randomness of P99's RNG IS pretty fucking wonky.

Evidence?

[Tecmos Deception]

Originally Posted by Isomorphic [You must be logged in to view images. Log in or Register.]

Evidence?

It's everywhere. There are simply too many streaks in all sorts of p99 activities given overall rates. I posted on resist streakiness a month or so ago with data if you care to dig a little, but I'm too lazy to provide the link.

Just a rough example for illustration: if p99 rng controlled coin flips, over the course of 10000 flips you'd see about 5000 heads and about 5000 tails. But you'd probably have 50 streaks of 10 instead of the 10 you should have seen based on 50/50 chances of heads vs tails.

[Isomorphic]

Originally Posted by Tecmos Deception [You must be logged in to view images. Log in or Register.]

It's everywhere. There are simply too many streaks in all sorts of p99 activities given overall rates. I posted on resist streakiness a month or so ago with data if you care to dig a little, but I'm too lazy to provide the link.

Just a rough example for illustration: if p99 rng controlled coin flips, over the course of 10000 flips you'd see about 5000 heads and about 5000 tails. But you'd probably have 50 streaks of 10 instead of the 10 you should have seen based on 50/50 chances of heads vs tails.

Are you saying that in 10000 coin fips the expected value of the number of times you get heads 10 times in a row is 10 but in EQ you will see 50? If so I have a few questions, are you only considering heads or does 10 tails count as one of these streaks. If we get 11 heads in a row does that count as 2 streaks of 10, counting 1-10 and 2-11. What is the variance of this expected value? How are you determining the expected value of this event in the first place?

[Tecmos Deception]

Originally Posted by Isomorphic [You must be logged in to view images. Log in or Register.]

Are you saying that in 10000 coin fips the expected value of the number of times you get heads 10 times in a row is 10 but in EQ you will see 50? If so I have a few questions, are you only considering heads or does 10 tails count as one of these streaks. If we get 11 heads in a row does that count as 2 streaks of 10, counting 1-10 and 2-11. What is the variance of this expected value? How are you determining the expected value of this event in the first place?

I called it a rough example for a reason. You did goad me into using Google for 8 seconds to find the thread i mentioned earlier, though:

https://www.project1999.com/forums/s...d.php?t=341531