Lost bitcoins - a bigger flaw than originally thought?

[nwbitcoin]

Having used the search facility a bit, all the discussion on lost coins seems to be focused on its only a bit, and it will make the rest more valuable so its not an issue.  However, I think there is a problem with this thinking.

All physical currencies have a percentage that are lost.  Its generally around 2-3% per year, and mostly effects the smaller denominations.

If we have a situation where we know there will only be 21 million coins, eventually, all the coins will become lost.

As things were, when you had so many coins that thousands are needed to buy a pizza, then that's when even the biggest denominations are worthless enough to lose.  You also have the issue with losing wallets and that whole learning cycle that early adoptors suffered.

As the mainstream start using the coins, they too will lose a few coins as they get used to the security needed to not lose this new currency.

Assuming a 2% lose per annum, within 20 years, half of the total number of coins will have been lost.

Does anyone else see this as a problem, or is the idea that because the coins can be split down, there is still enough currency to be used by everyone on the internet - for ever?

[1714880326]

1714880326

[1714880326]

1714880326

[1714880326]

1714880326

[1714880326]

1714880326

[cho]

a 2% loss per year does not lead to a 40% loss in 20 years.

"all the coins will become lost" : no, not with a flat x% rate of loss per year. For all the coins to become lost, you need a 100% loss the last year.

[luke.watson]

His maths might be off but it's true

We start with 21 M (assuming no coins are lost until all coins have been mined)

Year 1 : 21M
Year 2: 20.5M
Year 3: 20.1M
Year 4: 19.8M

This could in theory continue until we are down to 1 Bitcoin and everyone is trading in 0.0000000001s for a new Ferrari, its true though the loss of coins with no coins being added is a issue, even if only 0.5 BTC were added a year it would still help to combat the loss of coins..

Although, coins are not actually lost they are just inaccessible ie the owner has forgotten his password, so I assume there will be a big business in tracking down lost BTC and recovering them in the future especially for the wallets of innovators who might have hundreds of coins stashed away and at $40 USD for 1 BTC I wouldn't mind finding one of these accounts even now

[Tacticat]

The more valuable the coins are the less loss there will be.

[herzmeister]

It is said that people can't think exponentially, but they also can't think asymptotically.

In 2013, John Doe I. may lose his 400 USD or 10 BTC wallet.
In 2014, John Doe II. may lose his 400 USD or 1 BTC wallet.
In 2015, John Doe III. may lose his 400 USD or 0.1 BTC wallet.

[beeblebrox]

A loss of 2% per year gives a half-life of about 36 years,  ie: that's how long it takes to loose 50%.
So, say we've already lost 1,000,000 due to early folly so far but from now on only 2% lost then:
In about 36 years you'll have about 10,000,000 left (this not really that accurate-- because I've based it on the max 21000000 being mined already now when really the next approx~10,000,000 will be mined over this period-- but most of them are mined in the next 8 years so it not so bad)
in about 72 years you'll have about 5,000,000 left
~108yr -> 2,500,000
~140yr -> 1,250,000
~170yr -> 500,000 or so.

500,000 is plenty enough to still be able to buy small cheap items.

(as-an-aside: Personally I believe that we should allow miners to recover lost coins,  but for reasons other than coin base skrinkage.)

[Technomage]

It is said that people can't think exponentially, but they also can't think asymptotically.

In 2013, John Doe I. may lose his 400 USD or 10 BTC wallet.
In 2014, John Doe II. may lose his 400 USD or 1 BTC wallet.
In 2015, John Doe III. may lose his 400 USD or 0.1 BTC wallet.

This pretty much covers the problem in its entirety. The average user stands to lose less and less every year. However, I think that the security of Bitcoin holdings will increase and people will figure out how to handle inheritance, etc etc. So I don't believe there will be a fixed percentage of lost coins, it will likely be a smaller percentage in the future. When and if Bitcoin hits mainstream, there could be a temporary spike in lost coins due to negligent use.

[Piper67]

It is said that people can't think exponentially, but they also can't think asymptotically.

In 2013, John Doe I. may lose his 400 USD or 10 BTC wallet.
In 2014, John Doe II. may lose his 400 USD or 1 BTC wallet.
In 2015, John Doe III. may lose his 400 USD or 0.1 BTC wallet.

This pretty much covers the problem in its entirety. The average user stands to lose less and less every year. However, I think that the security of Bitcoin holdings will increase and people will figure out how to handle inheritance, etc etc. So I don't believe there will be a fixed percentage of lost coins, it will likely be a smaller percentage in the future. When and if Bitcoin hits mainstream, there could be a temporary spike in lost coins due to negligent use.

It's already increased. Multiple backups, online wallets (themselves backed up). I don't think we're going to see anywhere near a loss rate of 2% yearly. More like .2%.

[proudhon]

A loss of 2% per year gives a half-life of about 36 years,  ie: that's how long it takes to loose 50%.
So, say we've already lost 1,000,000 due to early folly so far but from now on only 2% lost then:
In about 36 years you'll have about 10,000,000 left (this not really that accurate-- because I've based it on the max 21000000 being mined already now when really the next approx~10,000,000 will be mined over this period-- but most of them are mined in the next 8 years so it not so bad)
in about 72 years you'll have about 5,000,000 left
~108yr -> 2,500,000
~140yr -> 1,250,000
~170yr -> 500,000 or so.

500,000 is plenty enough to still be able to buy small cheap items.

(as-an-aside: Personally I believe that we should allow miners to recover lost coins,  but for reasons other than coin base skrinkage.)

How do you distinguish a lost coin from a saved coin?

[beeblebrox]

.....
(as-an-aside: Personally I believe that we should allow miners to recover lost coins,  but for reasons other than coin base skrinkage.)

How do you distinguish a lost coin from a saved coin?

I knew someone would ask this.  

Firstly, let me state clearly that don't want to argue on this thread about this-- cause that's not what the thread is about. So, I'll give just this one reply.

Ok, If you allow the miners to sweep up the untouched coins then the ones that remain untouched are most likely lost (yes, before you mention it: I know that a very small percent of the time it maybe that they're not lost its just that the owner can't make any transaction-- say if the owner is imprisoned or in a coma, etc-- but this would be rare. To them I'd say it's just tough luck)
Personally I'd suggest a schedule where you can leave them untouched for a year or two, but then after that the miners can periodically take a percentage of the original amount so that it takes say 5-10years before they disappear completely.

The reason why I feel miners should be allowed to collect dead coin is two fold:
1) I believe the transaction fee model is fundamentally flawed due to the fact that you can trade bitcoin off-chain.  I've just been arguing this on a another thread-- no-one else seems to believe me though.  (This flaw will take a few decades from now to reveal itself and I'll be a very old man by then so in reality I don't really care about it.)

2) The second reason is an argument from a sense of equity.  It's to do with the question of who should pay to maintain network security.  Take me for example, almost 2yrs ago now I bought approx. 2500 coin as an investment and have left it sit on the network accumulating dust.  I've never made any transactions since so haven't paid any fees to the miners-- however it is the miners that are protecting my coin.  Isn't it fairer if I was forced to contribute a bit to the cost of protecting my investment?  If miner's were allowed to collect dead coin then I'd be forced either to pay the miners by the fee incurred in making a transaction from one of my addresses to to another in order to revive my coin, or pay them direct by dead-coin loss.

[proudhon]

.....
(as-an-aside: Personally I believe that we should allow miners to recover lost coins,  but for reasons other than coin base skrinkage.)

How do you distinguish a lost coin from a saved coin?

I knew someone would ask this.  

Firstly, let me state clearly that don't want to argue on this thread about this-- cause that's not what the thread is about. So, I'll give just this one reply.

Ok, If you allow the miners to sweep up the untouched coins then the ones that remain untouched are most likely lost (yes, before you mention it: I know that a very small percent of the time it maybe that they're not lost its just that the owner can't make any transaction-- say if the owner is imprisoned or in a coma, etc-- but this would be rare. To them I'd say it's just tough luck)
Personally I'd suggest a schedule where you can leave them untouched for a year or two, but then after that the miners can periodically take a percentage of the original amount so that it takes say 5-10years before they disappear completely.

The reason why I feel miners should be allowed to collect dead coin is two fold:
1) I believe the transaction fee model is fundamentally flawed due to the fact that you can trade bitcoin off-chain.  I've just been arguing this on a another thread-- no-one else seems to believe me though.  (This flaw will take a few decades from now to reveal itself and I'll be a very old man by then so in reality I don't really care about it.)

2) The second reason is an argument from a sense of equity.  It's to do with the question of who should pay to maintain network security.  Take me for example, almost 2yrs ago now I bought approx. 2500 coin as an investment and have left it sit on the network accumulating dust.  I've never made any transactions since so haven't paid any fees to the miners-- however it is the miners that are protecting my coin.  Isn't it fairer if I was forced to contribute a bit to the cost of protecting my investment?  If miner's were allowed to collect dead coin then I'd be forced either to pay the miners by the fee incurred in making a transaction from one of my addresses to to another in order to revive my coin, or pay them direct by dead-coin loss.

I think it's fine to discuss this here because this is an attempt to address the OP's claim that lost coins are a flaw that needs addressing.  But, again, how do you distinguished between lost coins and coins that I've saved, say, in offline, split-key, wallets?  You never answered that.

You said:

Ok, If you allow the miners to sweep up the untouched coins then the ones that remain untouched are most likely lost (yes, before you mention it: I know that a very small percent of the time it maybe that they're not lost its just that the owner can't make any transaction-- say if the owner is imprisoned or in a coma, etc-- but this would be rare. To them I'd say it's just tough luck)

Wait, what?  Again, the issue is that you want to take coins from addresses that haven't had activity at all or for a very long time.  The question can be rephrased as, "How do you know that untouched coins are lost, and can thus be taken without depriving anyone of their digital property?".  Your answer is, "Take (sweep) the untouched coins...".  But if the question is about how we can know when we're justified in taking them, the answer can't just be, "well, take them!".  Do you see how you've failed to answer the question?

Having issued that criticism, I think that there is some sense to your reasons for wanting to collect fees on dormant coins, but you can argue that case without saying anything about lost coins; and I think you're better off not trying to say anything about lost coins because there really is no way to tell from the blockchain alone what is lost and what is saved.

[greyhawk]

Personally I'd suggest a schedule where you can leave them untouched for a year or two, but then after that the miners can periodically take a percentage of the original amount

I don't know who you are. I don't know what you want. If you are looking for a get-rich-quick-scheme, I can tell you I don't have one. But what I do have is a very particular set of coins; coins I have acquired over a very long trolling career. Coins that have been untouched by people like you. If you let my coins go now, that'll be the end of it. I will not look for you, I will not pursue you. But if you don't,

[nwbitcoin]

I have a feeling nobody is seeing the scale involved here.

How many people are currently using bitcoin? lets be generous, and say 1 million.  If we are looking to have just the rich countries of the world using bitcoin, we are talking 1 billion people.  If each one was to lose, 0.01 btc at some time over their 1st year of usage, that would amount to 10 million bitcoins.  I've lost real money in the past - everyone does - but with coins, I don't really care.  It has very little value to me, and the government make sure there is enough in circulation by producing more!

The point is that its not the bit coins people will be loosing, but the decimals as they lose wallets, delete files and generally act as proper ignorant users - anyone who has done any IT support will be aware of how stupid some user errors can be.

With a billion people using bitcoins, we are also going to be seeing a lot of transactions - possibly 100k per second or more, meaning that using the figures from today to extrapolate on future behaviour isn't the best way of preparing for this possibility.

Is there some scope to deal with this issue built into the plan?

[DannyHamilton]

Ok, If you allow the miners to sweep up the untouched coins then the ones that remain untouched are most likely lost (yes, before you mention it: I know that a very small percent of the time it maybe that they're not lost its just that the owner can't make any transaction-- say if the owner is imprisoned or in a coma, etc-- but this would be rare. To them I'd say it's just tough luck)

Wait, what?  Again, the issue is that you want to take coins from addresses that haven't had activity at all or for a very long time.  The question can be rephrased as, "How do you know that untouched coins are lost, and can thus be taken without depriving anyone of their digital property?".  Your answer is, "Take (sweep) the untouched coins...".  But if the question is about how we can know when we're justified in taking them, the answer can't just be, "well, take them!".  Do you see how you've failed to answer the question?

It seems that the point he was making is that if it is known by all users that any coins that aren't spent/moved within some official timeframe (say 2 years) will automatically be swept (or have a significant fee), then any rational actor will devise a coin storage plan that includes moving the coins on a regular basis (within 2 years).  Since any rational actor will move their coins regularly to avoid the penalty, any that are not moved can be considered lost.

[greyhawk]

It seems that the point he was making is that if it is known by all users that any coins that aren't spent/moved within some official timeframe (say 2 years) will automatically be swept (or have a significant fee), then any rational actor will devise a coin storage plan that includes moving the coins on a regular basis (within 2 years).  Since any rational actor will move their coins regularly to avoid the penalty, any that are not moved can be considered lost.

So everyone will have coin storage plans that will automatically (we know people's memories can't be trusted) move the coins.

Here's the keyword: Automatically. The coins get moved automatically. So there's nothing there to sweep because -again- all the coins get moved automatically. There's no reason to sweep because there's nothing to sweep.

So the logical course of action is: Keep systems simple. Dont' sweep.

[Blazr]

We only need 1 Satoshi to keep Bitcoin going, we can always add more decimal places and use new units, such as pricing things in μBTC.

Lost coins are not an issue at all.

[beeblebrox]

I have a feeling nobody is seeing the scale involved here.

How many people are currently using bitcoin? lets be generous, and say 1 million.  If we are looking to have just the rich countries of the world using bitcoin, we are talking 1 billion people.  If each one was to lose, 0.01 btc at some time over their 1st year of usage, that would amount to 10 million bitcoins.  I've lost real money in the past - everyone does - but with coins, I don't really care.  It has very little value to me, and the government make sure there is enough in circulation by producing more!

The point is that its not the bit coins people will be loosing, but the decimals as they lose wallets, delete files and generally act as proper ignorant users - anyone who has done any IT support will be aware of how stupid some user errors can be.

With a billion people using bitcoins, we are also going to be seeing a lot of transactions - possibly 100k per second or more, meaning that using the figures from today to extrapolate on future behaviour isn't the best way of preparing for this possibility.

Is there some scope to deal with this issue built into the plan?

It really isn't a problem for atleast a few of centuries take a look at the figures I gave above.

[Soros Shorts]

At some point you could always change the client/protocol to allow 1 Satoshi to be broken up, e.g. into a trillion pieces. This would be a relatively minor change I don't think anybody would have any objections.

[drawingthesun]

If a satoshi is split into a trillion pieces, then the lost coins won't matter at all.

I also think that a method for accepting infinite divisibility should be incorporated into the Bitcoin protocol now.

Any user should be allowed to take their bitcoin and turn it into a trillion trillion smaller parts, this type of flexibility will help give bitcoin the ability to be everlasting.

[beeblebrox]

Ok, If you allow the miners to sweep up the untouched coins then the ones that remain untouched are most likely lost (yes, before you mention it: I know that a very small percent of the time it maybe that they're not lost its just that the owner can't make any transaction-- say if the owner is imprisoned or in a coma, etc-- but this would be rare. To them I'd say it's just tough luck)

Wait, what?  Again, the issue is that you want to take coins from addresses that haven't had activity at all or for a very long time.  The question can be rephrased as, "How do you know that untouched coins are lost, and can thus be taken without depriving anyone of their digital property?".  Your answer is, "Take (sweep) the untouched coins...".  But if the question is about how we can know when we're justified in taking them, the answer can't just be, "well, take them!".  Do you see how you've failed to answer the question?

It seems that the point he was making is that if it is known by all users that any coins that aren't spent/moved within some official timeframe (say 2 years) will automatically be swept (or have a significant fee), then any rational actor will devise a coin storage plan that includes moving the coins on a regular basis (within 2 years).  Since any rational actor will move their coins regularly to avoid the penalty, any that are not moved can be considered lost.

Ok, so I will make a comment.  

Yes, this is exactly what I meant.  I'm glad there's atleast there is someone here who has 1/2 a brain, stick around Heaven knows that bitcoin needs you.

Ok, I made the mistake that I'm taking with normal people, not bitcoiners.  I'll try to give a little analogy just incase people here don't understand what Danny has just said.  If you see a mouse in your kitchen and you pick up your cat and rub its nose in the mouse's face and the mouse doesn't run away, then chances are that the mouse is dead.  Do you get it now?   In, general you can assume that people take actions try to avoid an obvious loss-- unless of course that person is a bitcoiner, because bitcoiners will flap there hands about and say that there is no loss coming it a government trick or or that neither sashotshi nor anyof the other bitcoin gods have told me there is a loss coming so I don't believe there is or .....