BITCOIN

zagfles

Cus said:

'Person A buys Bitcoin at $10k
Person B buys Bitcoin from A at $70k'

So person A now has more rubbish $, but A used to own a bitcoin but now doesn't have it anymore. And it would cost him $70k if he wanted it back. So is he really up?

Depends. If A is a true believer that bitcoin is the future then what's he doing converting it back to worthless $? If he was a true believer he'd want the price of bitcoin to fall so he can fill his boots with more bitcoin for his rubbish fiat. So he's a loser for selling out. B is the winner as he's the one owning the future.

But if A never really believed in bitcoin and just wanted to make a quick profit in boring old fiat, he's well up. Well done to him. He might have spent every day on the interweb desperately trying to convince gullible fools like B that bitcoin is the future, how it'll make them big profits, how it'll give them financial freedom, using carefully selected statistics in the same way politicians do, in order to ramp up the price so he could sell his bitcoin for more. So a stunning success for A, bad luck to B. Unless of course B manages the same and sells to C for $500k. It could go on for ever, right?

thegentleway

User232002 said:

MeteredOut said:

the ad-hominem attacks 

You made a really rather obvious error between A|B and B|A. This is Statistics 101. Thats not hyperbole; Bayes Theorem is literally taught in Stats 101 and this will be one of the first things that will be discussed. Its the type of error that nobody who has actually studied Statistics at a reasonable level would make, and so it immediately identifies you as being someone that lacks technical knowledge here.

As it said, it’s not a case of 'it’s not quite right, but it’s near enough.' It’s flat out wrong. And it’s very relevant to casual observers of this thread that the people dismissing Bitcoin are not actually technically capable. 

I love stats. Would you be so kind as to explain to me how you can do a Bayesian inference without any conditional probabilities? Because I’ve never seen that before! I thought that from Bayes’ theorem you needed p(A), p(B) *and* p(B|A) to calculate p(A|B).

For the casual observer: From those two statements you can’t infer the % of addresses in profit. You may dispute @MeteredOut’s assumption but given his assumption the % is correct.

User232002

thegentleway said:

I love stats. Would you be so kind as to explain to me how you can do a Bayesian inference without any conditional probabilities? Because I’ve never seen that before! I thought that from Bayes’ theorem you needed p(A), p(B) *and* p(B|A) to calculate p(A|B).

So, you completely missed the point I was making but instead of stopping to think that it might be you who is indeed incorrect, you thought I needed to know the variables involved in Bayes? The arrogance on this forum really is insufferable.

thegentleway said:

For the casual observer: From those two statements you can’t infer the % of addresses in profit. You may dispute @MeteredOut’s assumption but given his assumption the % is correct.

I would dispute his assumptions, but even if I allowed them to be correct his conclusion is absolutely bogus. And you stating this shows that you don't understand it either. 

Here were the original assumptions and conclusion;

MeteredOut said:

So, if 90% of bitcoins are in profit, and 1% of addresses hold 90% of bitcoins, it doesn't take too much stretching of the imagine to think that approaching 99% of people who own bitcoins are not in profit. 

So let me make this rather obvious;

Suppose we have 100 Bitcoins and 100 people with 1 wallet each. The probability of any individual Bitcoin, or fraction of a Bitcoin, being in profit is 0.9

1 Wallet holds 90 Bitcoins
99 wallets hold 0.1 Bitcoin each

The question is; What is the expected number of wallets that are in profit?

If you think the answer here is 1%, you are completely incorrect. Its also not 'nearly 1%.'  I'm not going to calculate it, but my gut feel says the expected value is probably around 80%.

The hint is that there is no fundamental difference between the permutation where the 1 big wallet has all the profitable Bitcoins, and the permutation of of all 99 wallets being in profit and about 80 of the Bitcoins in the big wallet being in profit. As well as all the other permutations in between. 

One aside is that it is possible for 90% of Bitcoins to be in profit, but 100% of wallets to be in profit too.

MeteredOut said:

(I know the maths doesn't strictly give us that, eg we can't assume the 1% holding 90% is hold the same 90% that is in profit, and a single person can own multiple addresses)

There's at least an awareness that this isn't strictly correct. But, the problem is that the effect is massively understated. This is exactly the same error as you get shown in Bayes 101 when they give you a sensitivity/specificity problem. Its always about a false positive drug test and doctors always make the exact same mistake illustrated above when given the same problem by stating that a 1% false positive means that if someone tests positive then there is a 99% chance they have the disease. In other words, the difference between A|B and B|A - which is why this gets taught in Bayes. So yes, its a common fallacy and doctors are smart people - but doctors aren't in online forums trying to convince people that their calculated statistics are correct either.

User232002

User232002 said:

As usual with these threads across the internet, as soon as [actual quantitative data is presented], the users [conveniently find a reason to disengage from the discussion]. 

As I mentioned earlier, its depressing that the wallet distribution / Gini correlation data that I addressed earlier has been ignored, just as it was when it was first posted in this thread. As has the mathematical discussion as to why Bitcoin can't be a ponzi because the accepted mathematical models for ponzis do not fit its history.

Instead we get some asinine replies; Can't disprove a statistic? Just say that politicians make stats up all the time, so this one must be made up too. An especially ironic statement given the talk of A|B & B|A on this very page. Like, what kind of argument is this? From someone who has 20k posts on an internet forum. Less quantity, more quality please. Or how about we just throw the claim out there again that people in this thread are desperately trying to get people to buy a few hundred quid of Bitcoin? It trades tens of billions every day, but yes that few hundred quid is vital to support the price.

This is always the way in this thread. And now the actual data posted and the actual arguments presented will be buried and forgotten about, only for another poster to repeat the same nonsense in 10 pages time. 

Honestly, I'm a bit bored. I think I will, very conservatively, 5x my portfolio over the next 18 months. Continuing to post here will cost me money as it just takes too much time, as it did in 2021. Goodbye.


1. 9b7b7b170fa725e2b0c36c750f9cb055a8d81e0b4f3bae84f53c60c60209263a

2. 6034a8267b1308f8fe9bd862d614a839ecdd38285307a1eeeb0c3b21826dd7e9

3. d9ff2bae721942f76bd6171bc0041e3d039d2cb28fd2f1402148fee7ce5fbf9f

4. c4bd35b30452073c988ebc6acf6f4f9d62fa3fe78a33dfa2daf63d6b5016bd10

mooneysaver

thegentleway said:

User232002 said:

MeteredOut said:

the ad-hominem attacks 

You made a really rather obvious error between A|B and B|A. This is Statistics 101. Thats not hyperbole; Bayes Theorem is literally taught in Stats 101 and this will be one of the first things that will be discussed. Its the type of error that nobody who has actually studied Statistics at a reasonable level would make, and so it immediately identifies you as being someone that lacks technical knowledge here.

As it said, it’s not a case of 'it’s not quite right, but it’s near enough.' It’s flat out wrong. And it’s very relevant to casual observers of this thread that the people dismissing Bitcoin are not actually technically capable. 

 I thought that from Bayes’ theorem you needed p(A), p(B) *and* p(B|A) to calculate p(A|B).

This is a very common error, it's more commonly referred to as the 'No true Scotsman' fallacy (https://en.m.wikipedia.org/wiki/No_true_Scotsman).

zagfles

User232002 said:

User232002 said:

As usual with these threads across the internet, as soon as [actual quantitative data is presented], the users [conveniently find a reason to disengage from the discussion]. 

As I mentioned earlier, its depressing that the wallet distribution / Gini correlation data that I addressed earlier has been ignored, just as it was when it was first posted in this thread. As has the mathematical discussion as to why Bitcoin can't be a ponzi because the accepted mathematical models for ponzis do not fit its history.

Instead we get some asinine replies; Can't disprove a statistic? Just say that politicians make stats up all the time, so this one must be made up too. An especially ironic statement given the talk of A|B & B|A on this very page. Like, what kind of argument is this? From someone who has 20k posts on an internet forum. Less quantity, more quality please.

Now you're making up strawmen to argue against. Did anyone say politicians "make stats up"? Maybe you can quote it? No, because it's not there. So you quote your own post rather than the one you're really replying to hoping nobody notices. 

What I said was "using carefully selected statistics in the same way politicians do". Politicians rarely make up stats. Instead they quote usually correct stats but carefully selected ones which make the point they want to make.

On this particular stats argument, I don't give a rats backside what percentage of bitcoin owners are in profit. It's a pointless irrelavent stat from the point of view of whether it's a good investment now. The vast majority of VLS100 investors are probably in profit. As they were this time last year. 

So I don't care who's right in this tedious arguments about stats, the stats don't tell you anything useful.

Honestly, I'm a bit bored. I think I will, very conservatively, 5x my portfolio over the next 18 months. Continuing to post here will cost me money as it just takes too much time, as it did in 2021. Goodbye.

Bye then. At least we've got yet another prediction we can probably laugh at in 18 months time. Although not very original, we already had the 5x prediction in March 2021. Funny that poster never came back to gloat like he said he would.

thegentleway

User232002 said:

Here were the original assumptions and conclusion;

MeteredOut said:

So, if 90% of bitcoins are in profit, and 1% of addresses hold 90% of bitcoins, it doesn't take too much stretching of the imagine to think that approaching 99% of people who own bitcoins are not in profit. 

So let me make this rather obvious;

Suppose we have 100 Bitcoins and 100 people with 1 wallet each. The probability of any individual Bitcoin, or fraction of a Bitcoin, being in profit is 0.9

1 Wallet holds 90 Bitcoins
99 wallets hold 0.1 Bitcoin each

The question is; What is the expected number of wallets that are in profit?

Where’s the missing 0.001 bitcoin - lol

1 wallet holds 90 bitcoins in profit
99 wallets hold 0.101 bitcoin at a loss

This permutation yields only 1% of addresses are in profit.

In the absence on any other information, it’s just as valid as any other permutation.

thegentleway

mooneysaver said:

thegentleway said:

User232002 said:

MeteredOut said:

the ad-hominem attacks 

You made a really rather obvious error between A|B and B|A. This is Statistics 101. Thats not hyperbole; Bayes Theorem is literally taught in Stats 101 and this will be one of the first things that will be discussed. Its the type of error that nobody who has actually studied Statistics at a reasonable level would make, and so it immediately identifies you as being someone that lacks technical knowledge here.

As it said, it’s not a case of 'it’s not quite right, but it’s near enough.' It’s flat out wrong. And it’s very relevant to casual observers of this thread that the people dismissing Bitcoin are not actually technically capable. 

 I thought that from Bayes’ theorem you needed p(A), p(B) *and* p(B|A) to calculate p(A|B).

This is a very common error, it's more commonly referred to as the 'No true Scotsman' fallacy (https://en.m.wikipedia.org/wiki/No_true_Scotsman).

Oh dear, it’s contagious - lol. Your friend brought up Bayes and I highlighted why it’s not relevant here. 

Scottex99

Merry Xmas, even to the No-coiners 😃

I hope 2024 is prosperous for us all 

Cus

User232002 said:

User232002 said:

As usual with these threads across the internet, as soon as [actual quantitative data is presented], the users [conveniently find a reason to disengage from the discussion]. 

As I mentioned earlier, its depressing that the wallet distribution / Gini correlation data that I addressed earlier has been ignored, just as it was when it was first posted in this thread. As has the mathematical discussion as to why Bitcoin can't be a ponzi because the accepted mathematical models for ponzis do not fit its history.

Instead we get some asinine replies; Can't disprove a statistic? Just say that politicians make stats up all the time, so this one must be made up too. An especially ironic statement given the talk of A|B & B|A on this very page. Like, what kind of argument is this? From someone who has 20k posts on an internet forum. Less quantity, more quality please. Or how about we just throw the claim out there again that people in this thread are desperately trying to get people to buy a few hundred quid of Bitcoin? It trades tens of billions every day, but yes that few hundred quid is vital to support the price.

This is always the way in this thread. And now the actual data posted and the actual arguments presented will be buried and forgotten about, only for another poster to repeat the same nonsense in 10 pages time. 

Honestly, I'm a bit bored. I think I will, very conservatively, 5x my portfolio over the next 18 months. Continuing to post here will cost me money as it just takes too much time, as it did in 2021. Goodbye.


1. 9b7b7b170fa725e2b0c36c750f9cb055a8d81e0b4f3bae84f53c60c60209263a

2. 6034a8267b1308f8fe9bd862d614a839ecdd38285307a1eeeb0c3b21826dd7e9

3. d9ff2bae721942f76bd6171bc0041e3d039d2cb28fd2f1402148fee7ce5fbf9f

4. c4bd35b30452073c988ebc6acf6f4f9d62fa3fe78a33dfa2daf63d6b5016bd10

Shame, I enjoy your perspective. Considering that this thread went quiet for the many months that bitcoin slowly dropped in price through 2022 to 2023, I guess you were furiously working hard to not lose too much money and too busy to post here..