Hello @philpa! Great video! Thank you for all the mentions
The one thing I wanted to clarify is that in the Cardano UTxO model (unspent transaction outputs) - a single address is not equal to a single wallet, but rather one wallet may (and usually does) contain multiple addresses, where each address has its own balance. It is impossible to know if some two addresses are from the same wallet for anyone except the owner of this wallet, and except some rare cases - this is a part of the privacy system (the same as with bitcoin).
You can think about it like this. In Ethereum -
address=wallet, but nothing is stopping a single person from owning multiple wallets, so when someone is analysing the Ethereum addresses - it is a stretch to say that each wallet represents a person. In a UTxO model (like Bitcoin, or Cardano) - the same is more true, because for each address, there’s a bigger chance that the same person also owns some other addresses.
Unfortunately, more accurate analysis becomes exponentially harder, because only this information is public. But the good thing is that overall analysis of the distribution stays somewhat the same, but with a slight shift to the right part of the chart, if you want to approximate the owners of addresses, to account for each person owning multiple of those. This is also the reason why there’s so much addresses with a low amount of coins (less than a single ADA) - this is because a single wallet may contain a “good” address (or few) with some average balance, and multiple “change” addresses, where only a dust from some previous transactions are stored, until the owner moves it.
So overall your conclusions are pretty good, with a slight correction that the number of addresses does not really correspond to number of persons, but it is very hard to estimate the actual number. Also as the balance of addresses goes up - the chance for multiple of them to have the same owner goes exponentially down - because almost every Cardano user have multiple addresses with <1 ADA, but a chance that the same user have multiple >100K addresses is already pretty low, so analysis for bigger addresses gets more accurate.
Also I want to point out that this clarification makes your point about millionaires even stronger, because there’s still a chance (even tho a lower one) that some of those >10M addresses are owned by same persons. So it shifts the distribution up a bit, meaning that theres a giant chance that a same person owns multiple >100 ADA addresses, and a big chance that a single person owns multiple >1K ADA addresses, and a decent chance that a single person owns multiple >10K addresses, etc
But I liked it very much, it is really nice to see the interest to this kind of stats, and I definitely gonna try to improve the technics and make stats even better and more representative!
P.S. Also added this comment on YouTube