Bitcoin Forum
June 24, 2018, 11:12:42 PM
 News: Latest stable version of Bitcoin Core: 0.16.1  [Torrent]. (New!)
 Home Help Search Donate Login Register
 Pages: 1 [2]  All
 Author Topic: I'm the unluckiest lucky person in the world  (Read 3416 times)
DamienBlack
Jr. Member

Offline

Activity: 56
Merit: 0

 June 18, 2011, 02:44:27 PM

I think it's you who doesn't understand probability - go Google the Gambler's Fallacy to see why you're wrong.

Well, glad to see I'm not the only one who understands this. I thought this was common knowledge. Kids stuff. I found it odd that people were arguing with me. Welcome to the internet.

You're both dumb-asses.

The Gambler's Fallacy pertains to the FALLACY of believing in an INCREASED probability of a particular outcome based on past outcomes. I.E. believing that since the coin flipped heads 80/100 tosses that tails is SURELY due to come out ahead over the next 100 tosses.

This doesn't change the FACT that a coin flipped X number of times still maintains a 50/50 probability of being heads or tails. In a sufficiently large sample size, a natural distribution nearing a perfect 50/50 ratio will emerge.

Bitcoin hashing is ALSO such a distributed probability. In the example of MtRed, their probability of uncovering blocks at the same rate they had been is not diminished OR increased, it's just PROBABLE based on the natural distribution, that they will not maintain such luck, and will in fact over time see a smaller number of block awards over the same time frame as the distribution equalizes over the rest of the network.

Fucking bloody hell you wikipedians...

Take a fucking course in logic.

Yes you are right, it is unlikely that they will maintain such luck. But it is just as unlikely that they will go on a dry spell, as originally suggested. Most likely, they will continue creating block at about the rate they should given their hashing power.

Someone said "They're likely in for a long walk of shame when few/no blocks show up for an entire difficulty tier...". They are not likely that just because they've had some good fortune.

Someone also said "Also, that probability will still throw huge gains at an agent, followed by an equally brutal dearth of gains, over time." That isn't how probability works. One has no reason to follow the other.

I trade bitcoin options at https://bitoption.org/ ... Join me.
I play poker at https://betco.in/ ... Join me.
Support the bitcoin economy, what do you do?
Tips: 1NfXhiTFEdKQTdLy49s6DYAP1K7MeFWyao
Even in the event that an attacker gains more than 50% of the network's computational power, only transactions sent by the attacker could be reversed or double-spent. The network would not be destroyed.
Advertised sites are not endorsed by the Bitcoin Forum. They may be unsafe, untrustworthy, or illegal in your jurisdiction. Advertise here.
1529881962
Hero Member

Offline

Posts: 1529881962

Ignore
 1529881962

1529881962
 Report to moderator
MacRohard
Full Member

Offline

Activity: 222
Merit: 100

 June 18, 2011, 02:48:34 PM

You submit all of your share back. The shares are blocks that work at the "minimum difficulty", but not necessarily the "current difficulty". You get paid based on how many valid shares you produce compared to others. The shares can and are checked, and they prove you've been doing work. They don't just trust you on it.

The minimum difficulty requires several million hashes. The way we usually denote the current difficulty is by dividing it by the minimum difficulty to indicate how many shares you'd need to find, on average, as opposed to how many hashes.

The only way they could check your share then would be to submit it to another miner (or reprocess it themselves I guess).. so almost the same thing but I guess they can do it on invalid blocks as well as valid ones.

No they check your share by doing that one hash, on the server. Just a single one. Even a crap card can do 5,000,000 hashes a second, so they have no problem checking all the hashes themselves. Basically, we search for sort of rare hashes, and submit them, they check them, proving that we are doing work (because they are indeed sort of rare) and sometimes it'll be a really rare hash that solver a block. Then everyone gets paid by how many sort of rare attempts they sent.

Because we only submit rare ones, they don't have to do very much work at all, we already weed out the vast majority, and they satisfy themselves that we really did work because otherwise we wouldn't have rare hashes to show them.

Ahh Ok. I think I get it. Interesting.. so if you wante to be a real bastard you could just not submit the 'good' hashes and still get paid almost as much.. but there's no financial incentive to do that since it would be worthless to you..hmmm..

elggawf
Sr. Member

Offline

Activity: 308
Merit: 250

 June 18, 2011, 02:49:53 PM

You're both dumb-asses.

The Gambler's Fallacy pertains to the FALLACY of believing in an INCREASED probability of a particular outcome based on past outcomes. I.E. believing that since the coin flipped heads 80/100 tosses that tails is SURELY due to come out ahead over the next 100 tosses.

Okay, so what's the difference between the fallacy of thinking that a string of heads means an increased probability of tails, and thinking that a string of short-found blocks on MtRed means an increased probability of a dry spell?

^_^
DamienBlack
Jr. Member

Offline

Activity: 56
Merit: 0

 June 18, 2011, 02:51:39 PM

Ahh Ok. I think I get it. Interesting.. so if you wante to be a real bastard you could just not submit the 'good' hashes and still get paid almost as much.. but there's no financial incentive to do that since it would be worthless to you..hmmm..

Yes, the address that the block reward will be paid to is already baked into the work that the pool sends to you. There is no way of changing it without changing the hash, which would ruin the "good" hash. The system is clean. No way to mess with it, no trust required. Besides hoping that the pool pays you your fair share that is.

You could hold it back, but why?

I trade bitcoin options at https://bitoption.org/ ... Join me.
I play poker at https://betco.in/ ... Join me.
Support the bitcoin economy, what do you do?
Tips: 1NfXhiTFEdKQTdLy49s6DYAP1K7MeFWyao
Synaptic
Member

Offline

Activity: 84
Merit: 10

 June 18, 2011, 02:55:54 PM

I never said dry spell, I said EQUALLY.

It may manifest as a dry spell, it may manifest as never, ever having the same luck again. But it will reach equilibrium somehow.

Either way, the gains of their pool so far, relative to their hashing power, means they're likely in for a "disappointing" run.

Not disappointing probablistically, but disappointing psychologically, hence the "brutality" of it.

Numbers mean nothing without human interpretations.

EDIT:  I did suggest a dry spell with the "no blocks for an entire difficulty."  I never meant this as a fact, just an illustration of the probable disappointments upcoming for them.
MacRohard
Full Member

Offline

Activity: 222
Merit: 100

 June 18, 2011, 02:58:22 PM

Ahh Ok. I think I get it. Interesting.. so if you wante to be a real bastard you could just not submit the 'good' hashes and still get paid almost as much.. but there's no financial incentive to do that since it would be worthless to you..hmmm..

Yes, the address that the block reward will be paid into is already baked into the work that the pool sends to you. There is no way of changing it without changing the hash, which would ruin the "good" hash. The system is clean. No way to mess with it, no trust required. Besides hoping that the pool pays you your fair share that is.

You could hold it back, but why?

Well here's one scenario that just came to me.. A rival pool operator seeking to drive another pool out of business could setup a sort of proxying setup where their own users are actually processing work from the other pool. They the withhold all the payout hashes but submit all the others to the pool they're trying to kill. That way the other pool is mostly paying for it's own destruction.

DamienBlack
Jr. Member

Offline

Activity: 56
Merit: 0

 June 18, 2011, 02:59:11 PM

I never said dry spell, I said EQUALLY.

It may manifest as a dry spell, it may manifest as never, ever having the same luck again. But it will reach equilibrium somehow.

Either way, the gains of their pool so far, relative to their hashing power, means they're likely in for a "disappointing" run.

Not disappointing probablistically, but disappointing psychologically, hence the "brutality" of it.

Numbers mean nothing without human interpretations.

Fair enough, sorry I misunderstood.

I trade bitcoin options at https://bitoption.org/ ... Join me.
I play poker at https://betco.in/ ... Join me.
Support the bitcoin economy, what do you do?
Tips: 1NfXhiTFEdKQTdLy49s6DYAP1K7MeFWyao
DamienBlack
Jr. Member

Offline

Activity: 56
Merit: 0

 June 18, 2011, 03:01:18 PM

Ahh Ok. I think I get it. Interesting.. so if you wante to be a real bastard you could just not submit the 'good' hashes and still get paid almost as much.. but there's no financial incentive to do that since it would be worthless to you..hmmm..

Yes, the address that the block reward will be paid into is already baked into the work that the pool sends to you. There is no way of changing it without changing the hash, which would ruin the "good" hash. The system is clean. No way to mess with it, no trust required. Besides hoping that the pool pays you your fair share that is.

You could hold it back, but why?

Well here's one scenario that just came to me.. A rival pool operator seeking to drive another pool out of business could setup a sort of proxying setup where their own users are actually processing work from the other pool. They the withhold all the payout hashes but submit all the others to the pool they're trying to kill. That way the other pool is mostly paying for it's own destruction.

The payout address is baked into the work. If the evil pool changed it, they couldn't submit the hashed to the good pool. If they don't change it, then submitting that hash to the network would pay the original address, which belongs to the good pool. It doesn't work.

I trade bitcoin options at https://bitoption.org/ ... Join me.
I play poker at https://betco.in/ ... Join me.
Support the bitcoin economy, what do you do?
Tips: 1NfXhiTFEdKQTdLy49s6DYAP1K7MeFWyao
MacRohard
Full Member

Offline

Activity: 222
Merit: 100

 June 18, 2011, 03:05:25 PM

Ahh Ok. I think I get it. Interesting.. so if you wante to be a real bastard you could just not submit the 'good' hashes and still get paid almost as much.. but there's no financial incentive to do that since it would be worthless to you..hmmm..

Yes, the address that the block reward will be paid into is already baked into the work that the pool sends to you. There is no way of changing it without changing the hash, which would ruin the "good" hash. The system is clean. No way to mess with it, no trust required. Besides hoping that the pool pays you your fair share that is.

You could hold it back, but why?

Well here's one scenario that just came to me.. A rival pool operator seeking to drive another pool out of business could setup a sort of proxying setup where their own users are actually processing work from the other pool. They the withhold all the payout hashes but submit all the others to the pool they're trying to kill. That way the other pool is mostly paying for it's own destruction.

The payout address is baked into the work. If the evil pool changed it, they couldn't submit the hashed to the good pool. If they don't change it, then submitting that hash to the network would pay the original address, which belongs to the good pool. It doesn't work.

I understand that. It doesn't matter. The victim pool would pay the evil pool because of the non-payout hashes they submitted. They would turn around and repay that money to their users who were doing the work. In the middle, they strip out all the payout hashes that their users found. It doesn't help them but it hurts the victim.

DamienBlack
Jr. Member

Offline

Activity: 56
Merit: 0

 June 18, 2011, 03:11:21 PM

I understand that. It doesn't matter. The victim pool would pay the evil pool because of the non-payout hashes they submitted. They would turn around and repay that money to their users who were doing the work. In the middle, they strip out all the payout hashes that their users found. It doesn't help them but it hurts the victim.

Oh I understand now. They are just trying to destroy the pool, not make money. Well, honestly, since most pools pay proportionally the pool would just have a very unusually low rate of block finds. I guess if it lasted long enough some people would leave the pool. If they were doing pay-per-share the brutally bad odds may drive them out of business. I suppose it would be a sort of evil attack.

I think you get the evil mastermind award. We'll name it the MacRohard attack.

I trade bitcoin options at https://bitoption.org/ ... Join me.
I play poker at https://betco.in/ ... Join me.
Support the bitcoin economy, what do you do?
Tips: 1NfXhiTFEdKQTdLy49s6DYAP1K7MeFWyao
 Pages: 1 [2]  All
 « previous topic next topic »