Author Topic: Price Feed Review  (Read 43937 times)

0 Members and 1 Guest are viewing this topic.

Offline binggo

  • Hero Member
  • *****
  • Posts: 2374
  • 世间太多瘪犊子
    • View Profile
What's a ashamed feed price? can you feel that?

It's a shame for the BITasset and BTS holder.

If you can't find out what's wrong with your feed price, please don't feed it.

And this also said to some big voting Proxy, please don't act like a zombie.


Offline zhouxiaobao

     The price of DEX should be an important criterion to measure the accuracy of feed price.
     When Bitcny is not devalued, the feeding price should not be lower than the price of DEX. But in fact, this often happens, and that's why people complain.

Offline alt

  • Hero Member
  • *****
  • Posts: 2821
    • View Profile
  • BitShares: baozi
unvoted delegate-zhaomu and magicwallet.witness

didn't get any response.
unvoted for verbatech2 and witness.yao.
will wait others response for another 24 hours.

Offline binggo

  • Hero Member
  • *****
  • Posts: 2374
  • 世间太多瘪犊子
    • View Profile
@binggo, I didn't understand the calculation you made in which you claim my feed prices affects the median by more than 2%. I will study it more to try to make sense of it and will comment about that later.

Ok,correct the words,“you have more 2% influence price differential than fiat price.”


verbaltech2   0.4672   0.584   1.6   1.02   14 seconds ago
fox   0.4694   0.5633   1.6   1.01   8 minutes ago

BTC pirce
roelandp   0.4729   0.5675   1.6   1.02   3 minutes ago
xeldal   0.4734   0.5681   1.6   1.01   22 minutes ago
magicwallet.witness   0.4742   0.569   1.6   1.01   1 minute ago
xn-delegate   0.4756   0.5707   1.6   1.01   18 minutes ago
sahkan-bitshares   0.4758   0.571   1.6   1.02   19 minutes ago
blckchnd   0.4769   0.5723   1.6   1.02   44 minutes ago
bhuz   0.4776   0.5027   1.6   1.01   31 minutes ago
xman   0.4779   0.59   1.6   1.01   23 minutes ago
zapata42-witness   0.4781   0.5737   1.6   1.02   14 minutes ago
USDT pirce 6.90

witness.yao   0.4783   0.574   1.6   1.01   3 minutes ago
btspp-witness   0.4796   0.6149   1.6   1.01   9 minutes ago
delegate-zhaomu   0.4806   0.5768   1.6   1.01   8 minutes ago
in.abit   0.4807   0.6163   1.6   1.01   7 minutes ago
delegate.freedom   0.4809   0.6011   1.6   1.01   38 minutes ago
bangzi   0.4813   0.5776   1.6   1.01   13 minutes ago
zbbtsbp   0.4818   0.5353   1.6   1.01   52 minutes ago
clockwork   0.4829   0.5795   1.6   1.01   9 minutes ago
gdex-witness   0.483   0.637   1.6   1.01   1 minute ago
witness.hiblockchain   0.483   0.5796   1.6   1.01   13 minutes ago

so what?
« Last Edit: June 03, 2019, 06:36:54 am by binggo »

Offline Thom

median does not mean correct.

the median of a mess is also a mess.

it's not difficult for one user to find what a value is correct at a moment. only when every witness' feed price is good enough the median can be a dependable value.

Then one of us is definitely wrong, and I don't think it's me. Abit has commented as have others about the nature of the median price calculation. We use the median price, not an average for the very reason to exclude large influences by individual witnesses.

This leads me to think you're only trying to manipulate prices / trades through your pushing your perspective. It wouldn't be the first time you've done that.

@binggo, I didn't understand the calculation you made in which you claim my feed prices affects the median by more than 2%. I will study it more to try to make sense of it and will comment about that later.
Injustice anywhere is a threat to justice everywhere - MLK |  Verbaltech2 Witness Reports: https://bitsharestalk.org/index.php/topic,23902.0.html

Offline bitcrab

  • Committee member
  • Hero Member
  • *
  • Posts: 1928
    • View Profile
  • BitShares: bitcrab
  • GitHub: bitcrab
median does not mean correct.

the median of a mess is also a mess.

it's not difficult for one user to find what a value is correct at a moment. only when every witness' feed price is good enough the median can be a dependable value.
Email:bitcrab@qq.com

Offline binggo

  • Hero Member
  • *****
  • Posts: 2374
  • 世间太多瘪犊子
    • View Profile
As of this moment my feed for CNY is only off by 0.05 BTS from median price, that's less than $0.0034185, and that is just one witnesses' influence.

0.05bts?

your feed price: 0.4733cny*0.05bts=0.023665cny, i don't think it is a small influence, you have more 2% influence than fiat price.

Now the fiat bts price is: 0.485cny, can you explain how you got the 0.4733cny?

Later you give a feed price 0.4725cny?! please explain?

Quote
If they are off at certain times are they off enough to skew the median by a significant amount

This is what‘s your thoughts? please don't so childish.

Do you know what's the difference between the two USDT Price? why one is 0.48147? and why one is 0.47094?

Do you know what's the difference between the two BTC Price?

« Last Edit: June 02, 2019, 06:52:16 am by binggo »

Offline Thom

I've adjusted my feed pricing a couple days ago and have been monitoring the results.  I hope you find it more in line with expectations. 

I'd like to mention that depending on when you take a snapshot of feed prices it might not be an accurate representation of how well any witness it tracking price, with regard to how often a feed script is checking for price changes.  For example if taken just after a significant price movement it might appear that some prices are off from where they should be.  Where as, if given more time (at least 1 hour), their feed price script will make the appropriate adjustment.   We are looking at many different markets also, so picking a time where all of these have been relatively stable for a period of time would be important for having a snapshot be an accurate representation.

I'm not saying this applied to my feeds abnormal prices, just wanted to mention it, as it might for some look worse than it actually is.

I'm currently only checking once an hour for price changes due to certain free API limit restrictions.  As far as I understand it, this is still adequate.  Please inform me if this is no longer considered acceptable.

After looking at my feeds, I found it was in fact a bit off and needed attention.  I'll continue to keep a closer eye on this.  Thanks everyone for paying attention.

Indeed, but I don't believe after reviewing bitcrab's link to the "standard" he is using takes such things into account. His "standards" are simply not comprehensive enough, but that doesn't matter to him apparently, or do other factors that should be concerning the witness role, such as the number of feeds a witness provides, or how frequently they are updated, to say the least.

When I check my feeds most of the time they are < 0.1% from median. When they are checked makes a huge difference. If they are off at certain times are they off enough to skew the median by a significant amount (do my "bad" feeds skew the median by more than 1% ?) I seriously doubt it. The charts on the page in black, how are they calculated? Which exchanges / sources are used? How often are they sampled?

Just throwing up a chart of some numbers to rank witnesses' feed accuracy without that info is not helpful. If that info is available consider publishing it for the benefit of the entire BTS community.

The tool I rely on is RoelandP's witness log: https://roelandp.nl/bitshareswitnesslog/ As of this moment my feed for CNY is only off by 0.05 BTS from median price, that's less than $0.0034185, and that is just one witnesses' influence. Calling for witness unvote for such small deviations is a witch hunt and is ridiculous. Not surprising to see such behavior from you tho. I am surprised to see alt agreeing with bitcrab tho.
Injustice anywhere is a threat to justice everywhere - MLK |  Verbaltech2 Witness Reports: https://bitsharestalk.org/index.php/topic,23902.0.html

Offline binggo

  • Hero Member
  • *****
  • Posts: 2374
  • 世间太多瘪犊子
    • View Profile
其实多数情况下,人民币与美元喂价基本都会低1%左右,部分情况下会低到2个点,甚至超过2个点,很少一部分情况会略微低0.5%,汇率换算及转换损耗需要及时注意,而且外盘BTC及USDT交易对的深度明显都不足。
从另一个方面来看的话,bitCNY现在MSSR=1.01,如果喂价偏低,会使MSSR变相偏大,进而影响内盘价格,最后再影响手续费。

看了下这个平台就是FIAT USD,USDT和美元汇率就是1.0对吗?
可以认为人民币通过BTC交易换美元导致美元溢价了2.2个点,这样是可以解释内盘美元比人民币溢价2.2个点

但是我当时是按美元计价得到内盘美元喂价低了超过2个点,人民币喂价差不多。
你看看bingo在我后面几分钟的截图,人民币计价 0.4728,按1美元7.05算是0.067美元。
有没有可能你们的脚本汇率转换逻辑有问题,把0.067美元的BTS按 6.9 转成人民币喂价 0.4627人民币,再按 7.05的汇率算成 0.0656美元了?
所以当时的实际情况是人民币和美元喂价都低了超过2个点。
« Last Edit: June 02, 2019, 12:00:58 am by binggo »

Xeldal

  • Guest
I've adjusted my feed pricing a couple days ago and have been monitoring the results.  I hope you find it more in line with expectations. 

I'd like to mention that depending on when you take a snapshot of feed prices it might not be an accurate representation of how well any witness it tracking price, with regard to how often a feed script is checking for price changes.  For example if taken just after a significant price movement it might appear that some prices are off from where they should be.  Where as, if given more time (at least 1 hour), their feed price script will make the appropriate adjustment.   We are looking at many different markets also, so picking a time where all of these have been relatively stable for a period of time would be important for having a snapshot be an accurate representation.

I'm not saying this applied to my feeds abnormal prices, just wanted to mention it, as it might for some look worse than it actually is.

I'm currently only checking once an hour for price changes due to certain free API limit restrictions.  As far as I understand it, this is still adequate.  Please inform me if this is no longer considered acceptable.

After looking at my feeds, I found it was in fact a bit off and needed attention.  I'll continue to keep a closer eye on this.  Thanks everyone for paying attention.

Offline alt

  • Hero Member
  • *****
  • Posts: 2821
    • View Profile
  • BitShares: baozi
看了下这个平台就是FIAT USD,USDT和美元汇率就是1.0对吗?
可以认为人民币通过BTC交易换美元导致美元溢价了2.2个点,这样是可以解释内盘美元比人民币溢价2.2个点

但是我当时是按美元计价得到内盘美元喂价低了超过2个点,人民币喂价差不多。
你看看bingo在我后面几分钟的截图,人民币计价 0.4728,按1美元7.05算是0.067美元。
有没有可能你们的脚本汇率转换逻辑有问题,把0.067美元的BTS按 6.9 转成人民币喂价 0.4627人民币,再按 7.05的汇率算成 0.0656美元了?
所以当时的实际情况是人民币和美元喂价都低了超过2个点。

@alt

为什么CNY:USD交易价格在7以上,而不是“官方”规定的6.9?因为“官方”汇率是被操纵的。而加密货币交易市场是自由的。

我的上次喂价:
Kraken BTC 报价 8550 美金
Bittrex BTC 报价 8549 美金
ZB BTC  报价 60241 RMB
AEX BTC 报价 60284 RMB

综合结果:
人民币:美金 约等于 7.05

最新报价截图


60300, 60202 / 8582, 8570 ~= 7.015, 7.036
« Last Edit: June 01, 2019, 01:14:14 pm by alt »

Offline zhouxiaobao

@alt

为什么CNY:USD交易价格在7以上,而不是“官方”规定的6.9?因为“官方”汇率是被操纵的。而加密货币交易市场是自由的。

我的上次喂价:
Kraken BTC 报价 8550 美金
Bittrex BTC 报价 8549 美金
ZB BTC  报价 60241 RMB
AEX BTC 报价 60284 RMB

综合结果:
人民币:美金 约等于 7.05

最新报价截图


60300, 60202 / 8582, 8570 ~= 7.015, 7.036

       那既然是自由汇率那为什么脱离不了官方汇率很多呢?所以这是中外比特币的价格差,是由搬砖成本决定的,不是汇率差。
      喂价应当反应的是bts与法币之间转换的价格,如果中间加上btc这一层,要把转换成本算上。
     为啥btc/bts交易对的价格总是比法币交易对的价格低?环节越多搬砖成本越高啊。所以在主要交易所和内盘当中,最高价格最接近实际法币转换价格。bitcny和bitusd都应当用最高价,为防止价格操纵,可以加一个偏离顶,比如若最高价大于平均价超过5%,则不再上调喂价。

Offline abit

  • Committee member
  • Hero Member
  • *
  • Posts: 4664
    • View Profile
    • Abit's Hive Blog
  • BitShares: abit
  • GitHub: abitmore
@alt

为什么CNY:USD交易价格在7以上,而不是“官方”规定的6.9?因为“官方”汇率是被操纵的。而加密货币交易市场是自由的。

我的上次喂价:
Kraken BTC 报价 8550 美金
Bittrex BTC 报价 8549 美金
ZB BTC  报价 60241 RMB
AEX BTC 报价 60284 RMB

综合结果:
人民币:美金 约等于 7.05

最新报价截图


60300, 60202 / 8582, 8570 ~= 7.015, 7.036
« Last Edit: June 01, 2019, 11:39:26 am by abit »
BitShares committee member: abit
BitShares witness: in.abit

Offline alt

  • Hero Member
  • *****
  • Posts: 2821
    • View Profile
  • BitShares: baozi
didn't get any response.
unvoted for verbatech2 and witness.yao.
will wait others response for another 24 hours.

USD/BTS price is 0.0658, while CNY/BTS price is 0.464,
means bitCNY/bitUSD rate is 7.05, while the fiat rate of CNY/USD is 6.9

I will unvote those witness who give feed price under 0.0658 if can't get a reasonable explain.


Offline BTSMoon

  • Full Member
  • ***
  • Posts: 91
    • View Profile
西方程序员的月薪大约是$7000美金,比特股满足不了他们啊,不如把喂价的事情交回给社区,我们只需要猴哥,Binggo 和 GDEX 三个就行了。
晕死,你还没搞明白什么是见证人吧,还三个人包场,我真是醉了

我可能还没弄明白什么是见证人,但是Abit大神应该很清楚吧:
https://bitsharestalk.org/index.php?topic=28418.msg330453#msg330453

我坚决的认为与其继续让无心的见证人搞喂价,不如让社区的有心人搞好喂价。让100个无心人喂价,不如让3个有心人喂价。