Dummy Scrolling

I'm pretty sure dummy scrolling doesn't work, but I guess people will do it when they feel a scroll will fail after 'X' amount of times. A guildie was dummy scrolling on a few equips with 30%s and they all boomed on the first try. Likewise, another guildie dummy scrolled a 30% after having 2 of them work on the intended equip and the dummy equip boomed. It's really a mixed bag and anything can happen since the only variable with scrolling is luck.

 

tried dummy scrolling on one of these high rates private servers:
failed 10 10% scrolls inarow and then scrolled each slot:
1/19
2/20
3/26
4/32
5/33
6/41
7/43
8/46
9/51
10/52
11/61
12/66
13/88
14/114

dummy scrolls used - around 560
---------------------------------------------------------------------------------------------------------
this year in high school I've studied in math probability, this forumla applies to scrolling but I'm not sure whether it works since you do use your scrolls on another item.

A - probability to land a 10% scroll ( P(A) = 0.1 )
B - failed atleast 9 out of 10 10% scrolls ( P(B) = 0.736 )


P() = C 1/10 * 0.1 * (0.9) times 9 + (0.9) times 10

the result is 0.3874/0.736 = 0.526 => the next scroll has 52.6% success chance.

we can see that I haven't landed the scrolls in the dummy test on the server at around 1/2 chance so even tho I want to believe dummy scrolling works it doesn't seem so, it's not as effective as it should be. perhaps you also need to take into account the dummy scrolls you land and then it becomes a way more complicated math equation.
maybe you should even take into account all the 10% scrolls you've ever used you can't know :|.

I haven't done probability in two months so I might be wrong on the calculation please write if you see something wrong there..



update all these calculations are wrong, previous events cannot change the next event's rate:
chance of landing a scroll considering you failed 2 before: (doing with a small number because it's easier, if it increased the chance it'll be slightly above 10%)
0.9*0.9*0.1 =0.081
P(A/B) = P(AxB)/P(B)
0.081/0.81 = 0.1
hence statical advantages don't matter when it comes to dummy scrolls if it does work (which I no longer believe) it's based on flaws in the rng code and on exploiting the flaws, I dunno really.

 

This is wrong.
All the calculation is correct. However, B is not something that will definitely happen.
You are assuming when u fail 9 10%'s, and u scroll the next slot, B will happen. (actually you scroll after u fail 10 10%'s), which is wrong.
^^ This assumption is correct.

A trivial point: Also, B contains the case that 1st (or 2nd to 9th) is landed.
^^^^This becomes nontrivial now.

Besides, if the success rate is 52.6%, you will only need 14/0.526=26.6 scrolls only. Whereas, you use 114 scroll to finish it.

 

The next scroll DOES NOT have a 52.6% success chance. The above formula is assuming that the probability of A is dependent on the probability of B, when in fact EACH individual event is completely independent from any other event. As stated here: http://www.stat.yale.edu/Courses/1997-98/101/condprob.htm

 

If dummy scrolling works, it shouldn't be attributed to Statistics. Rather, you should find clues from the code. For example, if the code says the success rate of 100 consecutive same kind of scrolls (like claw att 10%) is 10%, then dummy scrolling would work. If the code says, when you scroll, it does not depend on anything happening whether or not it will be landed, then dummy scrolling would definitely no work, which is written on textbook.

 

I wrote that it's very unclear whether it works because you use scrolls on different equips and you don't know if the time on which you used them matters (as in if scrolls from a year ago or hours ago should be taken into statics), and I mentioned that the scrolls I used didn't support the result.
ye I calculated if I fail 9 before using a scroll but I failed 10 before using, it shouldve been around 2/3.5 scrolls worked but anw it serves it's purpose: showing that the statics don't really fully apply/ or apply at all here.

correct me if I'm wrong but they are dependent:

they are independent if P(A)*P(B)=P(AxB)
P(A)=0.1
P(B)=0.736
P(AxB)=0.3874
0.0736 != 0.3874.

 

Ya, A and B are dependent.
However, you are calculating the wrong thing.

 

We can go back and forth but the definitive answer for anyone who isn't convinced: go into the source code of sources available online and see how it handles scrolling. You will see how 10% really is 10% and has no dependence on what happened before or after it (even if you want to argue pseudo randomness of RNGs, you still have to consider all the other people on the server scrolling in between your dummy scrolling and actual scrolling )

 

hm, B - failed atleast 9 out of 10 scrolls which can be any of the 10, it could be 1,2,3,4,6,7,8,9,10
and then the 5th scroll will get the P(A/B) effect, however the same chance applies if the 1-9 fail to the tenth slot tho right?

perhaps dummy scrolling should be done right after a sc then ?
yeah statics applies when there's a complete random chance but the source code is really near random isn't it?

 

It is indeed as random as RNGs allow. Dummy scrolling is a myth.

 

Wont stop me from dummy scrolling 2 infinity & beyond

 

When you go 0/23 not dummying, and 1/2 1/5 1/2 dummying, it's just too hard to believe dummying doesn't work... sigh

 

It's called luck... Your individual cases are just a few in many failures. You are free to believe what you want, but it doesnt make it right. Go check the source and you will see that past or future scrolling has absolutely no impact on your current scroll.

Tfw people dont understand basic probability.

https://en.m.wikipedia.org/wiki/Gambler's_fallacy

 

what if i say this has happened 5x already

 

Read the wiki link and read the source. Doesnt matter how many times it has happened to you, it is still just luck. There is absolutely nothing in the source code which would say that the way you claim it is, is the way scrolling works. Since it is not in the source, it isnt valid/the way scrolling works. Also read the wiki link.

Each scroll is an independent event not dependent on other events, past or future.

 

I've seen it, I'll believe what I want. You can believe what you want.

Tfw ppl dont understand advanced statistics

 

I dont need to believe. I know what is correct here. If you are going to make that claim please do expand on why you think i dont understand statistics. What tou are believing in is a common fallacy called the Gambler's Fallacy. Please read the wiki article on it and the one below since you clearly dont understand independence of events.

https://en.m.wikipedia.org/wiki/Independence_(probability_theory)

 

I like how even after the coders and people who have messed with the source of the game TELL YOU that this DOES NOT work. People are still like....

well it worked for me before, therefore your wrong.

 

I'm not saying you're wrong. I'm saying you're free to believe what you want to. He's the one saying I'm wrong.