Page 4 of 7

Re: so.... here we go...

PostPosted: Wed Feb 15, 2017 1:17 am
by Shadow86
Lansett wrote:
Chrumps wrote:Can you point out a single instance when it did not extend the season ?

Yes some times ago an Everbloom followed a Bloodmoon.


We was in an everbloom tho, I remember that one. So again it just extended the season we was already in.

EDIT: Not saying I support this bugged season theory, I just wanted correct info out there.

Re: so.... here we go...

PostPosted: Wed Feb 15, 2017 8:09 am
by nadde991
Chrumps wrote:
nadde991 wrote:Looks like the Salem community needs to learn their statistics.
As TMeow said:
TotalyMeow wrote:go read about Gambler's Fallacy, Confirmation Bias, Law of Averages, and Independence of Events.

You clearly didn't read about these.

Yes, I did read about all of them: Gambler's Fallacy is the assumption that probability changes depending on past results, Confirmation Bias is when you try to confirm one's preexisting beliefs or hypotheses, whilst not considering alternative possibilities, Law of Averages states that the likelihood of certain events evens out if the frequency is increase (more attempts, higher probability) and finally Independence of Events is about two events whose probability does not affect one another.

Crumps wrote:Can you point out a single instance when it did not extend the season ?

By that logic I should also say that the probability of getting an arcane amethyst is bugged because me and my friends got 3 on one blood moon.
0.03 % is not that low, when you think about how many rare items there are and how many people are selling them.

Re: so.... here we go...

PostPosted: Wed Feb 15, 2017 8:20 am
by matan002
nadde991 wrote:Yes, I did read about all of them: Gambler's Fallacy is the assumption that probability changes depending on past results, Confirmation Bias is when you try to confirm one's preexisting beliefs or hypotheses, whilst not considering alternative possibilities, Law of Averages states that the likelihood of certain events evens out if the frequency is increase (more attempts, higher probability) and finally Independence of Events is about two events whose probability does not affect one another.


Image

Re: so.... here we go...

PostPosted: Wed Feb 15, 2017 8:28 am
by Qiresea
if the probability for something to happen is 50% at each event, it means that the probability for the event not happening is an inverse exponential function of time, so in principle, you can wait for an infinite amount of time for it to happen. i wouldnt be surprised by the season not changing after a few bloodmoons, the statistics is very low.

Re: so.... here we go...

PostPosted: Wed Feb 15, 2017 9:10 am
by Chrumps
nadde991 wrote:Yes, I did read about all of them: Gambler's Fallacy is the assumption that probability changes depending on past results,

Where did I assume the probability has changed depending on something else than a suspected bug ?
nadde991 wrote:Confirmation Bias is when you try to confirm one's preexisting beliefs or hypotheses, whilst not considering alternative possibilities,

No, I am considering two hypotheses: 1-It is dependent on the previous season, 2-It is random with 50/50 chances. So far it is like 5:2.5 for the first hypothese.
nadde991 wrote:Law of Averages states that the likelihood of certain events evens out if the frequency is increase (more attempts, higher probability) finally Independence of Events is about two events whose probability does not affect one another.

Again, where did I state that some frequency has changed?
nadde991 wrote:
Crumps wrote:Can you point out a single instance when it did not extend the season ?

By that logic I should also say that the probability of getting an arcane amethyst is bugged because me and my friends got 3 on one blood moon.
0.03 % is not that low, when you think about how many rare items there are and how many people are selling them.

I agree with you in one point: YOU should learn about statistics.

Qiresea wrote:if the probability for something to happen is 50% at each event, it means that the probability for the event not happening is an inverse exponential function of time, so in principle, you can wait for an infinite amount of time for it to happen. i wouldnt be surprised by the season not changing after a few bloodmoons, the statistics is very low.

Actually waiting infinitely and explaining to oneself it's just statistics would be a particularly sick case of Confirmation Bias.
The chance for the particular series I am referring to is 1/32 (=0.5^5).

On a more practical side:
I am looking for sequences like:
    Everbloom - Blood Moon - Coldsnap
    Coldsnap - Blood Moon - Everbloom
So far it was (IIRC):
    EB-BM-EB
    CS-BM-CS
    EB-BM-EB
    CS-BM-CS
    CS-BM-CS

The last 3 were in 2017, the other 2 in mid 2016 so I might be missing something.
It is written now so I can wait a few bloodmoons more to see a larger sample.

Re: so.... here we go...

PostPosted: Wed Feb 15, 2017 9:21 am
by Qiresea
no, it is not a case of confirmation bias. it may need a very long time to happen because the probability function is asymptotic. what i am trying to say is that i dont have any reason to believe that there is a bug in the code at this level of statistics.

Re: so.... here we go...

PostPosted: Wed Feb 15, 2017 9:41 am
by Chrumps
Qiresea wrote:no, it is not a case of confirmation bias. it may need a very long time to happen because the probability function is asymptotic. what i am trying to say is that i dont have any reason to believe that there is a bug in the code at this level of statistics.

Well, provide a better statistics then.

Re: so.... here we go...

PostPosted: Wed Feb 15, 2017 9:58 am
by punker
Chrumps wrote:
nadde991 wrote:Yes, I did read about all of them: Gambler's Fallacy is the assumption that probability changes depending on past results,

Where did I assume the probability has changed depending on something else than a suspected bug ?
nadde991 wrote:Confirmation Bias is when you try to confirm one's preexisting beliefs or hypotheses, whilst not considering alternative possibilities,

No, I am considering two hypotheses: 1-It is dependent on the previous season, 2-It is random with 50/50 chances. So far it is like 5:2.5 for the first hypothese.
nadde991 wrote:Law of Averages states that the likelihood of certain events evens out if the frequency is increase (more attempts, higher probability) finally Independence of Events is about two events whose probability does not affect one another.

Again, where did I state that some frequency has changed?
nadde991 wrote:
Crumps wrote:Can you point out a single instance when it did not extend the season ?

By that logic I should also say that the probability of getting an arcane amethyst is bugged because me and my friends got 3 on one blood moon.
0.03 % is not that low, when you think about how many rare items there are and how many people are selling them.

I agree with you in one point: YOU should learn about statistics.

Qiresea wrote:if the probability for something to happen is 50% at each event, it means that the probability for the event not happening is an inverse exponential function of time, so in principle, you can wait for an infinite amount of time for it to happen. i wouldnt be surprised by the season not changing after a few bloodmoons, the statistics is very low.

Actually waiting infinitely and explaining to oneself it's just statistics would be a particularly sick case of Confirmation Bias.
The chance for the particular series I am referring to is 1/32 (=0.5^5).

On a more practical side:
I am looking for sequences like:
    Everbloom - Blood Moon - Coldsnap
    Coldsnap - Blood Moon - Everbloom
So far it was (IIRC):
    EB-BM-EB
    CS-BM-CS
    EB-BM-EB
    CS-BM-CS
    CS-BM-CS

The last 3 were in 2017, the other 2 in mid 2016 so I might be missing something.
It is written now so I can wait a few bloodmoons more to see a larger sample.

tilt

Re: so.... here we go...

PostPosted: Wed Feb 15, 2017 10:28 am
by MaxPlanck
Actually the 2 week summer was followed by a coldsnap after BM but nice try.

Re: so.... here we go...

PostPosted: Wed Feb 15, 2017 4:23 pm
by nadde991
Listen dude, your statement was it has a 1/32 chance of occurring therefore it's bugged. But guess what? The probability of getting Winter, Summer, Winter, Summer, Winter is also 1/32.
If Winter had a probability of lets say 1% and summer had a probability of 99% I would think it was strange to get 5 in a row but 50 % is really high.
Also I never said that Gambler's Fallacy, Comfirmation Bias... was directly related to this topic, I just meant that if you think getting 5 in a row on a 50 % probability you should probably learn some statistics.