That thread has already proven to be full of inconsistencies, and was coming directly from a Monero supporter (Flenst) with a history of failed attempts to discredit Dash PrivateSend
(which is why he titled his Reddit thread 'i traced a privatesend
this time no educated guess'). He also created a second Reddit thread about Dash PrivateSend named
'Around 13% of DASH's privateSends are traceable to their origin' in the Dash-unfriendly sub-Reddit 'CryptoCurrency" shortly after creating his DashPay Reddit thread, indicating to me a form
of Dash smear campagne at the time. See :
https://www.reddit.com/r/CryptoCurrency/comments/bkwat1/around_13_of_dashs_privatesends_are_traceable_to/* Flenst thought it was a 4 round mixed PrivateSend transaction that he could trace. Turns out it was a 2 round mixed PrivateSend transaction, which of course is the very lowest form of mixing
that Dash offers optionally. Users are sacrificing privacy for speed, when they set back the default number of rounds from 4 to 2 rounds.
UdjinM6 : Dash Lead Developer
* Flenst understanding of how Dash PrivateSend works had quite a few incorrect assumptions, making his thread even less of an educated guess
* Flenst claim to trace a 2 rounds mixed PrivateSend transaction was tied to a fixed known originating address at the start of the mixing process, something users should avoid during mixing
(fixed addresses like that of an Dash exchange address, that can then be traced back and forward by blockchain analyzers)
* Flenst never manage to produce a tool to verifiable de-anonymize Dash PrivateSend transactions, it was just a whole lot of assuming (spread over two Reddit threads) on his part so far i could tell.
To conclude :
Dash users that want to mix and use PrivateSend transactions are responsible for not including any fixed known originating addresses (exchange addresses for example) at the start of their mixing process, as that will negate any privacy build-up. Dash users are also responsible for choosing the level of privacy by leaving the mixing rounds on either 4 rounds (which is the default option and provides good privacy) or by raising the number of rounds to mix all the way to 16 rounds (which provides very strong privacy, but mixing takes longer) or by lowering the number of rounds to mix to 2 (which provides the weakest form of privacy, but mixing goes a lot faster). That is the part that Dash leaves optionally for users, so they can set their own level of privacy on their transactions. Also Dash users should never send the exact same amount before mixing and after mixing through PrivateSend.
If Dash users lower the number of rounds from 4 to 2 and use a fixed known originating address before the mixing, they may as well not do any mixing at all as it will be pretty pointless.
The correct Reddit title should have been 'Around 13% of DASH's privateSends are traceable to their origin, when using a fixed known originating address', but i'm not even sure he verifiable proved even that