Comparing both charts, the values are almost the same. Since I don't know the exact methodology used to get their numbers, this chart is close enough to the original chart to be safe to say that I'm using an almost identical methodology.

The cumulative amount deposited into the addresses only has a 1% difference to the number reported by Elliptic which is $93M, once again confirming the methodology used.

With that out of the way, let's dig into how the numbers are obtained.

3.1 Similar to Sam, I extracted 26 addresses from the documents, and aggregated all of the funds going into the addresses.

3.2 Unlike Sam, I used Anton's dataset instead of relying on BitQuery.

There is a major flaw if we use this method to calculate the actual money the was deposited into the addresses:

4.1 It doesn't account for money sent back to the senders, i.e. exchanges.

4.2 The reason we need to account for this is that otherwise, the numbers will be extremely bloated.

4.3 For example, if someone deposits $100 into an address, and the address sends the $100 back to the sender, the actual net flow is 0.

4.4 Another example, if the sender sends $100 back into an address after their intended recipient sends it back, we will be double counting the $100. So, instead of $100 being counted as amount deposited, we're counting it as $200. This back and forth action bloats the actual number by n times the transactions occur.

4.5 This problem worsens when more than 2 parties are involved, imagine A -> B -> C -> A -> B, where B and C are operated by the same people. Transactions such as this are much harder to process, but it still means that only one deposit occured.

To simplify things, I will be assuming transactions like 4.5 never occurred or are insignificant.

Using the same methodologies as the ones used above, we find out that we get the same reported amounts by Elliptic and BitOK, which totals to around $130M.

This in turn means that the problem stated earlier exists in this report too.

While the methodologies are the same, the bloated number is much more severe in this case, which is 131% more than the actual amount.

The difference in the reported and actual amounts is jarring, where $73M has been returned to the sender.

Since this dataset is inclusive of PIJ addresses, we are able to subtract the amounts obtained above to get the amounts deposited into non-PIJ addresses.

This means that out of $129M of the total funds, only $35M is deposited into non-PIJ addresses, $6M lower than the number reported.

Deducting $53M out of the total $56M we found by adjusting the numbers, we find out that only $3M is deposited into those addresses.

The deposited amount is actually at least 90% less than the calculated $35M or the reported $41M.