Can coinjoins also be not so successful? Or even fail?
Two huge successful coinjoins within 15 minutes 🔥🔥🔥
360 inputs
346 outputs
42 bitcoin
https://mempool.space/tx/620370cfa56101d5eb5df3eb36cc25dc1b80ded0182a56d50a500db67b8086cd
369 inputs
357 outputs
39 bitcoin
https://mempool.space/tx/5dd0bf6b406ea40c6f19f0729be7564e47d189b265583987e162071b2991b474
Discussion
Yes!
If only one single input out of those 300+ does not sign the transaction, it is not valid according to Bicoin consensus rules, and other nodes will reject the transaction.
This can happen when a user shuts down his computer, or the wifi goes off, or a Tor node is overloaded, or the user is malicious.
The first attempt of signing a coinjoin transaction has only ~25% success rate.
Then the coordinator kicks out the non-signing input, and attempts a second round, which has ~50% likely hood of success.
So it usually doesn't take more than 5 attempts to succeed.
I see. But losing funds in the process is not possible right?
Correct!
A coinjoin is atomic, so either nothing happens, or everyone gets all his money on a new address.
Coinjoin is also non-custodial, meaning you never give up control over your private keys and utxos.
A coinjoin client assumes that the coinjoin coordinator and other participants are malicious, and it has strong defenses against shenanigans.