Is the current volume 'real'?

NuBits
#1

I know there was some issues with bots colliding yesterday. Was that resolved? Is the volume we’re seeing for NuBits today reflecting real trading, possibly due to Bitcoin’s downward trend?

#2

I just noticed that too… 310k(24h) in volume? That’s insane

#3

The trading volume is unfortunately not accurate.

#4

What’s the latest as to why?

#5

I think it’s still the what you pointed out with the bot collisions. I don’t think we’ve moved away from that period to not be reflected in the chart. Though we may have more volume due to Bitcoin sinking. It’s hard to tell at this point.

#6

Here’s my analysis of what happened.

We are digging into the bot logic to see if we can identify why it occurred. Once we know that, we’ll be able to come up with a solution for how to mitigate it in the future.

All indications point to this being an occasional problem since the bot was turned on. I’ll let @KTm post her update later, but it is my understanding that there will be a full accounting review needed to identify these erroneous wall collisions and then figure out how to best resolve them.

It’s been a very busy last two days as we’ve been tracking down these logic conflicts – it doesn’t really help that BTC prices have been at their most volatile, which has caused the walls to be moved ~1500 times in the last 48 hours. A vast majority of those moves have been cleanly updated.

#7

So, we have two issues:

  1. Two bots on the same non-USD currency pair often trade with each other when walls move
  2. The NuBot sometimes gets confused and trades with itself

Is this correct?

#8

Yes, those are the two negative scenarios that we are aware of right now. I’m going to be working with @KTm for the next couple days to data mine the logs we have to see if there are more variations of those two (or others that we haven’t yet come across) that can be identified.

#9

So this is still happening right now then?

#10

Because the defect still exists. We have two options:

  1. Shut down the bots and stop supporting the markets;
  2. Deal with the outcome of leaving the bots up (namely, transaction fees on the exchanges)

No one has indicated that 1 is preferred over 2.

#11

Increased velocity of wall movement == increased risk of collisions

#12

The 24h volume appears to have exceeded $1M on coinmarketcap.com. It would be useful to find out how much the unnecessary transaction fees equate to if the volume isn’t purely organic growth.

#13

This should be fixable in software. Query all open orders and double-check them before placing any order. It’s slower, but it’s safer.

#14

It already does that. These are situations outside the programmed logic. What you describe is a fundamental behavior already imbued in the code, as it was something that is obvious.

Something is occasionally short-circuiting the standard behavior. That corner case is what we’ll be working on identifying and fixing.

I believe it has something to do with an incoming order immediately before or during the time when a wall shift is initiated.

#15

This may be urgent. See this screenshot from NBT/BTC on BTER: two huge trades, less than an hour apart. If this was the NuBot trading with itself, we’re leaking a lot of value in wasted trades!!

#16

Shutting down operations on BTER is the only option that would prevent this from happening. Until it can be resolved. Personally, I’m more comfortable with paying the fees than the negative publicity that is sure to appear for pulling the walls.

#17

I’m kind of surprised with this thread, for a moment I thought that the BTER volume is real taken into consideration what’s been going on in the markets over the weekend. I’ve noticed that 4 hours ago the NBT volume at BTER was 100k USD, I’ve checked it now and it’s over a million. I had some thoughts to share, but then I’ve stumbled across this thread.

#18

is any of the volume real…

#19

Yes, the buy wall is nearly 50’000 NBT on that exchange and currency pair. That’s real.

I count this happening 5 times over a one-hour period, according to the trade history. This is severe. Maybe we can run with 10% of the currently-funds active on the pair, to mitigate losses?

#20

Agree with Ben and I understand there are no trading fees on BTER anyway.

Until we know what the issue is I don’t think we know what part of the trading is real