How often is arbitration used?

Just curious how often arbitration is necessary, i.e. how many times per week or per month do arbitrators get involved in trades on Bisq (if this data is being tracked)? Also is the number of arbitrator-assisted trades known as a percentage of total trades?

You can have a look here :
it gives an idea about the numbers.

Looking at :
gives also an idea about the percentage.

1 Like

(The support GitHub repo primarily tracks reimbursements due to failed trades. It won’t provide an answer to your question, @freesov.)

We don’t track the trade-to-dispute ratio precisely, but it floats somewhere around 10%. Most disputes are due to simple mistakes, bugs or networking issues, and get resolved quickly and amicably.

1 Like

Thanks for the helpful replies! This satisfies my curiosity. So it sounds like 1 in 10 trades require help of some sort from an arbitrator. It would be great to see that figure be reduced to 1 in 100 (if not too unrealistic). From the level of professionalism I have witnessed here so far, such a target is achievable. :wink:

Agreed. 1 in 10 sounds like a lot, but in practice it’s pretty manageable. Nevertheless, we’d love to get it below 5% as well. The main limiting factor is dev bandwidth. Just the ability to take the feedback and implement UI / UX improvements that help avoid mistakes, and to implement fixes to what can very often be quite subtle and difficult networking problems. Thanks for inquiring about this in any case. Cheers.

1 Like

Is it really 10% atm? In August it was < 5%… Atm there seems to be Tor issues which might lead to network issues, so that might be a reason for increased problems. 1% is our target…

1 Like

I measure by taking the number of disputes I arbitrate in a given month, multiplied by 2 (arbitrators) and then I divide that number by the total number of trades that month. So for example, if I arbitrate 50 disputes in a month (typical) and multiply that by 2, I get 100 disputes. I then divide that by 1000 (typical number of trades per month), and that comes out to be 0.1 == 10%. Perhaps August was an anomaly. I don’t run these numbers every month, but every month I have run them, it comes out around 10%.

With the latest releases lots of bugs should have be decreased (tx broadcast issue, msg not arriving). But lets see if that is true :wink: In August at least it was much lower…