Possible WOB interpretation: Totally balanced BTC orderbooks --> neither strong up nor strong down movements are imminent. No guarantees, though ;-)

Thanks a lot. Much appreciated.

Any questions I can answer?

Do you have hints for us? WOB is only a part of step 1 of 8 steps in Newbium, but I want to make sure to build it as good as possible.



I understand your doubts.

We are still far away from that scenario though. When the totals were converging (i.e. adding even more exchanges did little to influence the aggregated outcomes) ... I was surprised to see the total order volume ... to be that small. Tiny, literally. Worldwide BTC sell orders (the part that is publicly visible, in 27 exchanges) have a total of ~25 mio USD. We are seeing only 1/290 of all Bitcoin in existence in the sell orderbooks (or ... 1/1400 of your 75mio nightmare).

Obvious explanations: People do keep their coins outside the exchanges, good. Some exchanges are truncating the depth to e.g. 500 or 2000 order around the current price, bad. And yes, orders can be hidden / conditioned - and then no one is able to count them in. Missing data is sad, but ubiquitous. Still, many people do aggregate datasets with missing data, all over science.

For sure: Any secondary analysis depends on the quality of the primary data.  That applies to all data analysis sites, also coinmarketcap, bitcoinwisdom, etc. - I guess you then also never trust, nor use such sites yourself? How do you navigate Bitcoinworld, never using any data aggregators? I would not want to trade like that, but that's just me.

But yes: I see my highest goal to provide reliable analysis, so I am worried about such questions myself. Essentially you can always only trust data that you have generated yourself.  But with your approach you assume in Cryptoland near to 100% are scammers? I don't believe that. The incentive for an exchange to report a false orderbook ... I don't yet see how they would profit from that. But yes, it might happen.

There are three effects which mitigate the problem:
* Exchanges can not profit financially from reporting fake orderbooks. (Or perhaps they do - can you invent a scenario?)
* A law of large numbers - the more exchanges we query (now 30+), the less important the black sheep will be in our aggregation.
* The bear scammers and the bull scammers ... can cancel each other out.

And most importantly, and as we will never be able to sample the whole totality: Do not mainly look at the absolute numbers - but look at the ratio of green(buybook) and red(sellbook) column.

Even if our new metrics is not "truth" itself - but only a good proxy for the truth - in cross-comparison, it really helps to uncover reality. Examples:

balanced https://twitter.com/altsheets/status/735506529466646528
sell pressure https://twitter.com/newbiumcoin/status/733508871747051520
buy pressure https://twitter.com/altsheets/status/735817783666200576

But of course: If an exchange will be reported to be providing fake data, then we'll simply cut them out of our data collection. Forever.

Good point, thanks for the hint. And please alert us when you hear about such API data fraud.

Sunny greetings