Wall Observer BTC/USD - Bitcoin price movement tracking & discussion

[JorgeStolfi]

And lack of depth charts

The API used by chart sites to extract the data from the exchanges requires that the entire detailed order book be returned every few seconds or so to every API client.

That API was designed by chart sites assuming that order books would contain only a couple thousand entries at most.  But robots and lack of trading fees generate LOTS of small orders on the Chinese exchange books.  Understandably, the largest exchanges truncate the book after a few thousand entries, which may cover only a narrow range of prices around the spread.

Chart sites should indicate the truncation with "..." or something.  Since they don't, their unwary clients keep assuming that what they see on the chart is the entire order book.

The chart-reporting API should be redesigned to return the order book SUMMARY in log scale (namely, total orders up to to 5, 10, 50, 100, ... USD above and below the spread; or, conversely, what the final price would be if one sold/bought 5, 10 50, 100, ... BTC or USD in one go.  But even that may be too expensive for the Chinese sites to compute and deliver in real time.

[JorgeStolfi]

ccess.

EDIT: Oops, the inventor of Dinosaur Analysis was @spooderman it seems.

It was my glorious achievement, dont take it away from me

OOPS, sorry! Re-editing my post now,

[magicmexican]

Looks like its going down, 2h macD usually does not fuck around and its rarely a trap, so i expect huobi going to 2550 minimum, maybe 2450-2500 even

[ChartBuddy]

Explanation

[Carra23]

A brief history of dinosaur patterns:



that was a PAIN to do. funny to go back though and read posts from people saying "well I think we definately can't go below 700" etc.

Thats a cute dino.

Once I become a billionaire I will adopt one as a pet.

[spooderman]

I can confirm that it was NOT me. Magicmexican gets the credit.

I merely went back and found his post from December.

And yes TERA, everyone WAS saying "the cycles are faster now, this bear market will be over faster than all the previous ones."

If we hit the 300s again I am BUYING.

[magicmexican]

saw a prediction to 400$ in 2.5 hours~, lets see if it holds

[Davyd05]

saw a prediction to 400$ in 2.5 hours~, lets see if it holds

China is going to bed so I expect a lull, but 400 hmmm I would hope stamp could up the volume but I don't think anyone is in a rush to buy.

[niothor]

And lack of depth charts

The API used by chart sites to extract the data from the exchanges requires that the entire detailed order book be returned every few seconds or so to every API client.

That API was designed by chart sites assuming that order books would contain only a couple thousand entries at most.  But robots and lack of trading fees generate LOTS of small orders on the Chinese exchange books.  Understandably, the largest exchanges truncate the book after a few thousand entries, which may cover only a narrow range of prices around the spread.

Chart sites should indicate the truncation with "..." or something.  Since they don't, their unwary clients keep assuming that what they see on the chart is the entire order book.

The chart-reporting API should be redesigned to return the order book SUMMARY in log scale (namely, total orders up to to 5, 10, 50, 100, ... USD above and below the spread; or, conversely, what the final price would be if one sold/bought 5, 10 50, 100, ... BTC or USD in one go.  But even that may be too expensive for the Chinese sites to compute and deliver in real time.

we're not talking about the data from the api , we're talking about the data on their websites
And gox was showing that even with their huge orderbook , i can't understand why huobi and okcoin are hiding this except .... fake data?

[Davyd05]

And lack of depth charts

The API used by chart sites to extract the data from the exchanges requires that the entire detailed order book be returned every few seconds or so to every API client.

That API was designed by chart sites assuming that order books would contain only a couple thousand entries at most.  But robots and lack of trading fees generate LOTS of small orders on the Chinese exchange books.  Understandably, the largest exchanges truncate the book after a few thousand entries, which may cover only a narrow range of prices around the spread.

Chart sites should indicate the truncation with "..." or something.  Since they don't, their unwary clients keep assuming that what they see on the chart is the entire order book.

The chart-reporting API should be redesigned to return the order book SUMMARY in log scale (namely, total orders up to to 5, 10, 50, 100, ... USD above and below the spread; or, conversely, what the final price would be if one sold/bought 5, 10 50, 100, ... BTC or USD in one go.  But even that may be too expensive for the Chinese sites to compute and deliver in real time.

we're not talking about the data from the api , we're talking about the data on their websites
And gox was showing that even with their huge orderbook , i can't understand why huobi and okcoin are hiding this except .... fake data?

the best part if the Data is faked and they hold less coins or equal amounts to those outside of China then we shouldn't expect an over whelming rush of coins leaving their market for ours.  

some selling volume building*

[niothor]

And lack of depth charts

The API used by chart sites to extract the data from the exchanges requires that the entire detailed order book be returned every few seconds or so to every API client.

That API was designed by chart sites assuming that order books would contain only a couple thousand entries at most.  But robots and lack of trading fees generate LOTS of small orders on the Chinese exchange books.  Understandably, the largest exchanges truncate the book after a few thousand entries, which may cover only a narrow range of prices around the spread.

Chart sites should indicate the truncation with "..." or something.  Since they don't, their unwary clients keep assuming that what they see on the chart is the entire order book.

The chart-reporting API should be redesigned to return the order book SUMMARY in log scale (namely, total orders up to to 5, 10, 50, 100, ... USD above and below the spread; or, conversely, what the final price would be if one sold/bought 5, 10 50, 100, ... BTC or USD in one go.  But even that may be too expensive for the Chinese sites to compute and deliver in real time.

we're not talking about the data from the api , we're talking about the data on their websites
And gox was showing that even with their huge orderbook , i can't understand why huobi and okcoin are hiding this except .... fake data?

the best part if the Data is faked and they hold less coins or equal amounts to those outside of China then we shouldn't expect an over whleming rush of coins leaving their market for ours. 

I am guessing that chinese exchangers hold between 1/10 and 1/5 of the coins that are on bitstamp and btc-e.

There is no way they could have acquired the amount they claim out of the blue.

Time will tell , but I doubt the numbers are that wrong.

[Davyd05]

I am guessing that chinese exchangers hold between 1/10 and 1/5 of the coins that are on bitstamp and btc-e.

There is no way they could have acquired the amount they claim out of the blue.

Time will tell , but I doubt the numbers are that wrong.

I am theorizing they have a lot of miners and or people who buy off of miners to flip on the markets

[JorgeStolfi]

PS. By the way, it is not just the Chinese exchanges that truncate the reported order books.  One could clearly see the truncation in some of the @Chartbuddy plots of MtGOX, when that exchange was alive.

The API specified by bitcoincharts is described here.  Presumably the other chart sites depend on this same API. 

I cannot see how the exchange could report their total bid/as sum through that API except by reporting the entire detailed order book.  In fact I see no reliable way to tell whether the book was truncated or not.  If the book is truncated, the bid/ask sums computed from it are wrong and mostly meaningless.

[knightcoin]

A brief history of dinosaur patterns:



that was a PAIN to do. funny to go back though and read posts from people saying "well I think we definately can't go below 700" etc.

Thats a cute dino.

Once I become a billionaire I will adopt one as a pet.

Cool pet
Talking about prehistoric monsters and old information machines producer  ( Edison Universal Stock Tickers -> http://www.stocktickercompany.com/stc/flash/homepageBanner.swf - Wall Street Memorabilia, ) , I wonder (daydreaming) how would be a "ticker" if Tesla had devoted some time to think about it ....ehhh, since Tesla just had improved every Edison inventions...  it would be kind of "bit-multiverse-ticker"  

[seleme]

Shit, my shitty bot bought at 430 last night.

[razibuzouzou]

PS. By the way, it is not just the Chinese exchanges that truncate the reported order books.  One could clearly see the truncation in some of the @Chartbuddy plots of MtGOX, when that exchange was alive.

The API specified by bitcoincharts is described here.  Presumably the other chart sites depend on this same API. 

I cannot see how the exchange could report their total bid/as sum through that API except by reporting the entire detailed order book.  In fact I see no reliable way to tell whether the book was truncated or not.  If the book is truncated, the bid/ask sums computed from it are wrong and mostly meaningless.

Indeed, some exchanges truncate their order book (BTC-E for example).
But the bid/ask sums may not be that meaningless, their scope is just limited to the provided range of orders.
Eventually, accuracy of this metric depends on the distribution of orders in the book.

[ChartBuddy]

Explanation

[JorgeStolfi]

we're not talking about the data from the api , we're talking about the data on their websites
And gox was showing that even with their huge orderbook , i can't understand why huobi and okcoin are hiding this except .... fake data?

Why would they bother to show the entire order book (which may well be larger than MtGOX's?)? For the vast majority of traders, the detailed book near the spread is all that matters.

Their website server has to query the database server, format the data, post it on the page, and serve the page to their clients  Depending on the size of the order book and on the volume of requests to both servers, showing the entire order book may be too expensive for them too.  (And they may even use the same API as the chart sites.)

On the other hand, the truncation may indeed be intentional. They probably would like to hide their bid/ask sums to avoid comparisons with other exchanges.  Huobi and OKCoin were clearly competing for the "largest exchange" title...

[Carra23]

snip.

Please stop spamming this thread

Please stop quoting the trolls.

Please stop quoting users who quote the trolls.

....

Sorry someone had to.

Had to do what.. quoting?  =)))

To keep increasing the chain.

[rpietila]

So, since you have popped in, how long til we get a head of steam up again and this all calms down (to enable a base for a decent rise)?

I note and agree with your points about exchange confidence keeping money off the exchanges in any depth, but greed will surely conquer wariness at some point.

I would welcome your thoughts on timescale back to a decent bull market - a month, three...  longer?

In my baseline scenario, we close April at above 500, and 1000 will be breached in July.

Then there is the slower scenario but I am quite hopeful that 2014 will be a year of a new ATH nevertheless.

As everyone (including me) expect a 2-3 month plateau, perhaps that is not what will happen