Author Topic: would you support a cheater as the wintess?  (Read 1710 times)

0 Members and 1 Guest are viewing this topic.

Offline alt

  • Hero Member
  • *****
  • Posts: 2736
    • View Profile
  • BitShares: baozi
Re: would you support a cheater as the wintess?
« Reply #30 on: December 29, 2018, 08:46:21 am »
no it isnt... MEDIAN != MEAN
sure I know it's median instead of average.

Offline alt

  • Hero Member
  • *****
  • Posts: 2736
    • View Profile
  • BitShares: baozi
Re: would you support a cheater as the wintess?
« Reply #31 on: December 31, 2018, 03:20:34 am »
@alt, Do not wast time of the community, let's make it simple, how about i share my active private key, feeding program(open sourced) and the config file to the community member who we both trust, let the community see who is lying, if you were wrong, you own me a apologize coz of u keep defaming me, and if i am cheating the community like what you said, i quit the witness by myself, apologize to the community, and leave the community forever. deal?
it's fair.
you don't need share your key, you just need give your code and config file to a trusty memeber, log the price every 5 minutes to a file, check if it's accuracy as usually.
what's going on?
I am still wating

Offline alt

  • Hero Member
  • *****
  • Posts: 2736
    • View Profile
  • BitShares: baozi
Re: would you support a cheater as the wintess?
« Reply #32 on: December 31, 2018, 03:28:32 am »
it happend again and again, just now crazybit's feed price update from 3.70370 BTS/CNY to  3.70702 BTS/CNY followed the median price.
http://bts.ai/block/33576189
http://bts.ai/block/33576388

Offline alt

  • Hero Member
  • *****
  • Posts: 2736
    • View Profile
  • BitShares: baozi
Re: would you support a cheater as the wintess?
« Reply #33 on: December 31, 2018, 11:15:20 am »
it's  funny, suddenly, crazybit lost the accuracy forever.
it's so clear he got the accuracy with cheating.

Offline crazybit

Re: would you support a cheater as the wintess?
« Reply #34 on: January 01, 2019, 01:43:01 pm »
@alt, Do not wast time of the community, let's make it simple, how about i share my active private key, feeding program(open sourced) and the config file to the community member who we both trust, let the community see who is lying, if you were wrong, you own me a apologize coz of u keep defaming me, and if i am cheating the community like what you said, i quit the witness by myself, apologize to the community, and leave the community forever. deal?
it's fair.
you don't need share your key, you just need give your code and config file to a trusty memeber, log the price every 5 minutes to a file, check if it's accuracy as usually.
what's going on?
I am still wating


just back from holiday, @clockwork already very kind to elaborate the feeding activity, but seems you still choose not to believe.

so i uploaded whole log https://drive.google.com/open?id=1ClIR80hPvOUAxyT7bZ_jjgH9__WdK5u9 of pass months for your reference, my feeding program is [email protected], you also can download and review the source code from(https://github.com/crazybits/bitshares-pricefeed), the original config file is also pmed you, you could simulate the feeding activity by yourself.

if u still do not believe and insist on your previous conclusion , i think i have no ability to convince you.

Offline alt

  • Hero Member
  • *****
  • Posts: 2736
    • View Profile
  • BitShares: baozi
Re: would you support a cheater as the wintess?
« Reply #35 on: January 01, 2019, 09:57:17 pm »
@alt, Do not wast time of the community, let's make it simple, how about i share my active private key, feeding program(open sourced) and the config file to the community member who we both trust, let the community see who is lying, if you were wrong, you own me a apologize coz of u keep defaming me, and if i am cheating the community like what you said, i quit the witness by myself, apologize to the community, and leave the community forever. deal?
it's fair.
you don't need share your key, you just need give your code and config file to a trusty memeber, log the price every 5 minutes to a file, check if it's accuracy as usually.
what's going on?
I am still wating


just back from holiday, @clockwork already very kind to elaborate the feeding activity, but seems you still choose not to believe.

so i uploaded whole log https://drive.google.com/open?id=1ClIR80hPvOUAxyT7bZ_jjgH9__WdK5u9 of pass months for your reference, my feeding program is [email protected], you also can download and review the source code from(https://github.com/crazybits/bitshares-pricefeed), the original config file is also pmed you, you could simulate the feeding activity by yourself.

if u still do not believe and insist on your previous conclusion , i think i have no ability to convince you.
there is only one way to support you are not cheating, give your code and config file to a trusty one, and run it with the same accuracy.
but since even yourself have lost the accuracy forever from the date 2018.12.31, I think nobody can implement the accuracy again.
let's just remember the wonderful accuracy from 2018.12.26 to 2018.12.31, it's a legend from crazybit.
@clockwork, what's your thought about he can never recreate the accuracy again?
@xeroc, you are a doctor, do you believe the accuracy in the 5 days is not cheating?
« Last Edit: January 01, 2019, 10:30:54 pm by alt »

Offline alt

  • Hero Member
  • *****
  • Posts: 2736
    • View Profile
  • BitShares: baozi
Re: would you support a cheater as the wintess?
« Reply #36 on: January 01, 2019, 10:40:20 pm »
let me give a math model.
le'ts suppose there is a reasonable price in every moments, it's x
the finally median price y is about random in a range of x*0.99 ~ x*1.01, or more.
suppose you are a honest witness, which give your feed price z in reasoanble, without any information about other's feed price and median price,
the posibility of z close to  y less than 0.01% is about 0.01%/0.02, which is far more less than 1%
if you can do it twice, the possibility is 1%^2, less than 0.01%
if you can do it 3 times, it's 0.01^3.
but crazybit have do it every 5 minutes in almost 5 days, in both CNY/USD market.

of course if'ts very easy to get the information about others price, and median price
so with a cheating, you can get the accuracy very easy.
« Last Edit: January 01, 2019, 10:44:22 pm by alt »

Offline alt

  • Hero Member
  • *****
  • Posts: 2736
    • View Profile
  • BitShares: baozi
Re: would you support a cheater as the wintess?
« Reply #37 on: January 01, 2019, 11:29:13 pm »
in fact, there are some pattern changed obviously since 2018.12.31
why do you think crazybit changed to another totally diffirent feed price strategy?

1. the accuracy totally lost after this day.
2. the feed price changed every 5 minutes in a very stable freequency after this day.
    In the before, he only updated price feed if the median price changed,
    in my previous snapshots you can see there are several times he didn't update feed price for more than 30 minutes.

Offline Thul3

  • Full Member
  • ***
  • Posts: 163
    • View Profile
Re: would you support a cheater as the wintess?
« Reply #38 on: January 01, 2019, 11:58:02 pm »
I didn't wanted to post here before since i haven't seen the code however i'm pretty good in math as i had also to study it.
When seeing first time this thread i instantly thought about probability calculation.
The probability calculation makes it hard to belive that this is just a coincidence or such a well defined script.

Alt explained it very well how big the mathamatical possibility is to hit exectly median price so often.
« Last Edit: January 02, 2019, 12:00:57 am by Thul3 »

Offline crazybit

Re: would you support a cheater as the wintess?
« Reply #39 on: January 02, 2019, 01:39:11 am »
let me give a math model.
le'ts suppose there is a reasonable price in every moments, it's x
the finally median price y is about random in a range of x*0.99 ~ x*1.01, or more.
suppose you are a honest witness, which give your feed price z in reasoanble, without any information about other's feed price and median price,
the posibility of z close to  y less than 0.01% is about 0.01%/0.02, which is far more less than 1%
if you can do it twice, the possibility is 1%^2, less than 0.01%
if you can do it 3 times, it's 0.01^3.
but crazybit have do it every 5 minutes in almost 5 days, in both CNY/USD market.

of course if'ts very easy to get the information about others price, and median price
so with a cheating, you can get the accuracy very easy.

let me tell you where you wrong. the wrong part is you make a wrong assumption, and you keep to prove the mathematic impossibility of the wrong assumption, i gave you the source code, origin config file. and you make a conclusion again even do not simulate it.

Offline alt

  • Hero Member
  • *****
  • Posts: 2736
    • View Profile
  • BitShares: baozi
Re: would you support a cheater as the wintess?
« Reply #40 on: January 02, 2019, 03:32:46 am »
I didn't wanted to post here before since i haven't seen the code however i'm pretty good in math as i had also to study it.
When seeing first time this thread i instantly thought about probability calculation.
The probability calculation makes it hard to belive that this is just a coincidence or such a well defined script.

Alt explained it very well how big the mathamatical possibility is to hit exectly median price so often.
good to know you known how ridiculous the accuracy are.
In fact many people don't understand math,
they don't regarded it as an efficient evidence until I take a photo about crazybit's witness console.
« Last Edit: January 02, 2019, 03:57:58 am by alt »

Offline alt

  • Hero Member
  • *****
  • Posts: 2736
    • View Profile
  • BitShares: baozi
Re: would you support a cheater as the wintess?
« Reply #41 on: January 02, 2019, 03:44:20 am »
let me tell you where you wrong. the wrong part is you make a wrong assumption, and you keep to prove the mathematic impossibility of the wrong assumption, i gave you the source code, origin config file. and you make a conclusion again even do not simulate it.
you still didn't tell me where I am wrong.
is it wrong that you have an amazing accuracy feed price between the date 2018.12.26 and 2018.12.30?
is it wrong that the possibility is very small to keep such an accuracy?
is it wrong that your feed price patten have changed obviously since 2018.12.31?

what did you expect me to do with the source code you provide?
can I got the same amazing accuracy like you did with this code?
if your answer is no, what's the point to run this?
if your answer is yes, I will try.
And it only make yourself more dubious everytime your feed price pattern change after I point out your flaw.

Offline pc

  • Hero Member
  • *****
  • Posts: 1426
    • View Profile
    • Bitcoin - Perspektive oder Risiko?
  • BitShares: cyrano
Re: would you support a cheater as the wintess?
« Reply #42 on: January 02, 2019, 10:17:46 am »
the finally median price y is about random in a range of x*0.99 ~ x*1.01, or more.
suppose you are a honest witness, which give your feed price z in reasoanble, without any information about other's feed price and median price,
the posibility of z close to  y less than 0.01% is about 0.01%/0.02, which is far more less than 1%

This assumption is wrong. The median corresponds exactly to at least one witness-provided feed value. So for 25 witnesses the raw probability that the median is yours is 4%.

This raw probability is in reality increased significantly by at least two more aspects:
* When a witness publishes a new feed value, his feed is the most recent and therefore more likely to be close to x than anyone else's.
* Some witnesses publish a new feed only if that would change the median. (This is a simple optimization which is perfectly valid and acceptable if the feed is checked often.)

IMO if you publish a price only if it changes the median, and your price is likely to be close to the median, the conclusion is that your feed value will be chosen as the median quite often.

Please provide proof if you accuse someone of cheating. The mere fact that someone provides an accurate feed can hardly be taken as such.
Bitcoin - Perspektive oder Risiko? ISBN 978-3-8442-6568-2 http://bitcoin.quisquis.de

Offline alt

  • Hero Member
  • *****
  • Posts: 2736
    • View Profile
  • BitShares: baozi
Re: would you support a cheater as the wintess?
« Reply #43 on: January 02, 2019, 11:17:52 am »
This assumption is wrong. The median corresponds exactly to at least one witness-provided feed value. So for 25 witnesses the raw probability that the median is yours is 4%.
even the probility is 4%, when it become 4%^10, it's still a task impossible.
and you have give your answer for a wrong question.
the pattern what crazybit did is that he just follow the median price exactely instead of given another number which became a new median price.
let me show you what's the different, when there are 3 numbers, (0.271,0.272,0.273), 0.272 is the median number from others, 0.273 is your current number
you can change your number to any between 0.271 and 0.272 which could become a median number.
but for crazybit, he just change to 0.272 instead of any others between 0.271 and 0.272, so the psossibility is much less than your answer.

This raw probability is in reality increased significantly by at least two more aspects:
* When a witness publishes a new feed value, his feed is the most recent and therefore more likely to be close to x than anyone else's.
you should see other's line chat from my previous post. https://bitsharestalk.org/index.php?topic=27687.msg326833#msg326833
it's unfortunately that the data seems not available after 2018.12.26 13:00, or you will see it more clearly.

* Some witnesses publish a new feed only if that would change the median. (This is a simple optimization which is perfectly valid and acceptable if the feed is checked often.)
IMO if you publish a price only if it changes the median, and your price is likely to be close to the median, the conclusion is that your feed value will be chosen as the median quite often.
exactely the opposite, crazybit never change the median, he always publish a new feed price which already in the median price, this means all his data is totally useless.
« Last Edit: January 02, 2019, 11:24:44 am by alt »

Offline alt

  • Hero Member
  • *****
  • Posts: 2736
    • View Profile
  • BitShares: baozi
Re: would you support a cheater as the wintess?
« Reply #44 on: January 02, 2019, 12:03:13 pm »
* Some witnesses publish a new feed only if that would change the median. (This is a simple optimization which is perfectly valid and acceptable if the feed is checked often.)

IMO if you publish a price only if it changes the median, and your price is likely to be close to the median, the conclusion is that your feed value will be chosen as the median quite often.
this really could be a good optimization, I will adopt this if I am a witness.