Considering that bitcoin has not yet reached mass adoption, its 21 million cap is as unknown a bitcoin itself to majority of people.
Yeah, to claim that it is the most famous number is completely wrong. However, I think that it will enter history eventually as one of the most significant things to happen in this era. There is nothing like Bitcoin and pretty much everything has unlimited supply these days. Even many altcoins that had fixed supply switched to infinite supply after some time.
Usually those who "guess" the number of lost coins are simply counting the UTXOs that have not moved for a certain number of years. I say this to everyone who makes these claims: I started with bitcoin back in 2014 and I have small amount of coins from those years that I have not moved for 11 years. While I still own the keys, these "guessers" count my coins as lost.
Those that do it that way are dumb, it is similar to those that think that they understand quantum computers and the potential risks that they carry after reading a few articles or posts on X. The problem does not lie in the observation of coins based on again, the problem lies in inference. The internet is full of dumb people unfortunately, most of them don't even know what they don't know. From a statistics perspective, what those people are doing is basically like this:
- Observed variable: time since last movement or spend.
- Inference: probability that keys are lost after N years is equal to 1, where N is a random number decided by random guess work by some internet idiot.
This is an error that is as basic as it gets when it comes to statistics. If you think in terms of latent variables,
"coin is lost" is not an observable state except in a few very specific cases where it is in a provably unspendable state (OP_RETURN being one)[1]. A proper analysis by a researcher would probably model coin movement as a stochastic process (and not a binary state of lost vs. not lost) and then use historical data to gather probabilities. For example, what is the probability that a coin that has not moved for N years will move again? This is done by analyzing all coins and movements. It could also make some distinction based on address types or other metrics.
So yes this can be done right but it requires a lot of statistical knowledge, Bitcoin technical expertise and research experience. However, the conclusion will never be dumb as the one we usually get. It will only gives us probabilities.
Their silence for over ten years isn't just a strategy, it’s a testament to a level of conviction that is rare in today’s attention driven market.
Another problem with these guesses is that they guess that an early day miner only owns the coins that are not moved and didn't own anything else! In other words if they haven't moved dozens of 50
BTC rewards from early blocks, they may have already moved a single block reward that is worth almost $4.5 million. Do they need to move all their coins? Not really. Therefore the rest are untouched and that doesn't show "conviction" of any kind...
As of time of writing, only people with P2PK addresses should definitely move all their coins. Not doing so is a mistake and risky to both the user and the network. For everyone else (assuming normal and secure address derivation of course) you are correct.
[1] @stwenhao I summon thee.
