Miners are getting jacked for 15-20% of earnings!!!!

[martychubbs]

Well, perhaps, but those hoppers didn't do the sometimes 4 hours of processing on the server that we did in order to get a shot at a fresh block.

While we're slaving away at calculating that hash...they're off trying to get more than their due many other places.

Anyway, I can solve the problem with a relative degree of ease.

You create another class of shares called 'loyalty shares'.  You create a number of shares equal to the estimated share count at 5 minutes with current has power.

You distribute loyalty shares proportional to the payouts on your last block.

So...

Two things...

1) Pool hoppers would be fucked. (their margins would be reduced to tatters.  If they happend to hop in on a 2 pool that closed in 2 minutes, their share would be super miniscule due to the loyalty shares.

2) It would ease attrition on long length blocks because folks wouldn't get the loyalty shares if they bailed.

This way... instead of your most loyal folks getting shafted in short rounds, we'd end up doing the best.

Doubt this will help you out either... but here are some numbers that prove that miners are getting jacked for 15-20% of earnings on blocks shorter than 12 minutes. on average.

https://docs.google.com/spreadsheet/pub?key=0AhhejNMzWwx8dGpYZE9FYXk3d05MQjVDNUxLcHNXZkE&output=html

It proves that either A) pool hoppers are doing better than your massive brain thought they could... or B) for some reason slush is stealing from us, but only on short blocks.

Please direct posts to Sargasm, I just helped him set up his own thread.

[slush]

thanks :-)

[martychubbs]

but here are some numbers that prove that miners are getting jacked for 15-20% of earnings on blocks shorter than 12 minutes. on average.

No, your math is incorrect, basically because you didn't considered hashrate change. Again, there was a lot discussion about it and, honestly, I trust Meni's calculations much more than yours. You obviously DIDN'T read his paper, so please do it and stop spamming here. If you want to continue in the discussion, please open separate thread. I'm sorry that I'm angry, but I'm little tired of such "smart guys" entering the discussion in the middle, without reading the papers, math and discussion history.

Look Sargasm, you have your first reply!  Please respond here.  Thanks

[RandyFolds]

His solution does not sound like it could be implemented with a 'relative degree of ease'. It sounds like it would be an absolute monster to code and manage.

[Inaba]

Yes, it would be a disaster to manage and adds complexity where none is needed.  The solution:  Don't mine on a proportional pool.  Problem solved.  No mathematical magic or acrobatics needed.

[jkminkov]

miners are jacked more than 15-20% when pool op keeps found blocks for himself

[organofcorti]

Well, perhaps, but those hoppers didn't do the sometimes 4 hours of processing on the server that we did in order to get a shot at a fresh block.

While we're slaving away at calculating that hash...they're off trying to get more than their due many other places.

Anyway, I can solve the problem with a relative degree of ease.

You create another class of shares called 'loyalty shares'.  You create a number of shares equal to the estimated share count at 5 minutes with current has power.

You distribute loyalty shares proportional to the payouts on your last block.

So...

Two things...

1) Pool hoppers would be fucked. (their margins would be reduced to tatters.  If they happend to hop in on a 2 pool that closed in 2 minutes, their share would be super miniscule due to the loyalty shares.

2) It would ease attrition on long length blocks because folks wouldn't get the loyalty shares if they bailed.

This way... instead of your most loyal folks getting shafted in short rounds, we'd end up doing the best.

I'll explain it to you like you're a five year old... we'll see if that helps.

Ok, jimmy... you see it's like there is a really tough puzzle that has to be figured out and when it is, everyone gets paid a portion of the total work for their efforts.

But there are some bad kids, jimmy.  These kids have figured out that if they only help right at the beginning, statistical averages mean that they will earn ~(this means approximately)20% more than by working on every puzzle till its done.

You see sometimes the puzzles take a long long time and are very difficult.  So by focusing only on brand new puzzles, they increase their chances of hitting a lucky one.

So instead, we're going to calculate a certain number of points that are a different class than the points you earn.  These points will be used to calculate your share and they will be based upon the amount you worked on the LAST puzzle.

This way, when bad kids want to play at your puzzle, since he wasn't there for the last puzzle, he won't make any money and he'll cry.  This is because when you add up the loyalty points and earned points, while the total scores amongst all the players that play every round will be proportionally equivalent , the bad kids score will be less.  If it is a very short round, it will be much less.

That way everyone that likes working hard will make more money!

Isn't that great jimmy?  Now run along and stop being such a fucking idiot.

I've looked at proposed (edit: non-provably fair) payout systems before, and none have done anything to prevent pool hopping at a proportional pool - some reduce hopper margins, some just make hopping more work and some make things actually better for hoppers while making the payout system seem fair. I haven't seen any that work as simply as pplns or as well as DGM. So when I see a new proposed payout system, I'm interested. To be honest, I'm expecting it to fail, but I can't sure until I simulate it and I'm always happy to be proven wrong.

So, let me make sure I have this right:

1. loyaltyShares <- sum(totalSubmittedShares for a given 5 minute period)
2. minerShareProportion <- minerSubmittedShares/totalSubmittedShares
3. minerPayout <- (minerSubmittedShares(this round) + minerShareProportion(last round)*loyaltyShares)/totalSubmittedShares(this round) * 50btc

So my question is, what is the assumed pool hashrate? Why use '5 minutes' as your share amount rather than (say) 0.1*D? If you're ok with using a fixed share amount instead or a fixed proportion of difficulty, then let me know.

[jkminkov]

miners are jacked more than 15-20% when pool op keeps found blocks for himself

Any evidence for this?

ask Vladimir :p

[slush]

jkminkov, can you be please more specific? Do you have any evidence that any large pool is keeping his blocks for himself?

[organofcorti]

jkminkov, can you be please more specific? Do you have any evidence that any large pool is keeping his blocks for himself?

I think Vladimir thought he did, but I wasn't convinced. I think jminkov's response was a bit tongue in cheek.

[jkminkov]

jkminkov, can you be please more specific? Do you have any evidence that any large pool is keeping his blocks for himself?

why just large pools?
do you have evidence than pool X is 100% honest - if yes, then make a whitelist...

organofcorti, you wasn't convinced that pool was hiding blocks but you wasn't convinced that the pool was NOT hiding blocks either

[Clipse]

Anything vladimir is fairly a moment of lol so wouldnt be surprising

[slush]

why just large pools?
do you have evidence than pool X is 100% honest - if yes, then make a whitelist...

Yes, I have the evidence (well, it's stats, so nothing is 100%) that at least me and deepbit are honest (at least from the view of stealing blocks): https://bitcointalk.org/index.php?topic=48889.0

The point is that such statistics can be done only after many thousands of pool rounds, so it's impossible to tell if some new pool is honest or not.

[DeathAndTaxes]

While I agree that prop pool miners are losing money to hoppers (how much is subject to debate) his "solution" is one in search of a problem.

PPLNS, PPS (and variants), DGM (double geometric) are not hoppable.  Why try to "fix" prop pools which are horribly broken when proven unhoppable solutions already exist.

I mean it would be like someone getting lead poisoning from lead dishes and coming up w/ an ultra complex "solution" of sealing, verifying, and testing lead based dishes on a daily basis to avoid/reduce lead intake .... instead of just not using lead dishes.  

Prop pool = good if you are a hopper and want to exploit idiots
Prop pool = idiotic if you aren't hopping.
It is a zero sum game. 

Hell I see hopping as Bitcoin Darwinism.

[Inaba]

I think it's because prop is the most easily understood by the lowest common denominator.  I put in X work, which is divided by Y total work.  That's how much I get paid.  Done.

Everything else involves at least one more mental step, possibly with the exception of PPS.  Not saying it's right or good, just saying that's the likely reason.

[bitlane]

Yes, I have the evidence (well, it's stats, so nothing is 100%) that at least me and deepbit are honest (at least from the view of stealing blocks): https://bitcointalk.org/index.php?topic=48889.0

Do you honestly want to start this little game up again ? Why not be a man and just come out and call BTCGuild a cheat, rather than point to a controversial thread, then fill up your 'friends bandwagon' with Tycho ...? Can't you stand on your own 2 feet without needing to get someone else on your side ?

While we are re-hashing old crap, Care to talk about the beginning of Namecoin Merged Mining ? .....how about it, there, 'Honest Abe' ?

Slush, you are a coward.

[rjk]

Yes, I have the evidence (well, it's stats, so nothing is 100%) that at least me and deepbit are honest (at least from the view of stealing blocks): https://bitcointalk.org/index.php?topic=48889.0

Do you honestly want to start this little game up again ? Why not be a man and just come out and call BTCGuild a cheat, rather than point to a controversial thread, then fill up your 'friends bandwagon' with Tycho ...? Can't you stand on your own 2 feet without needing to get someone else on your side ?

While we are re-hashing old crap, Care to talk about the beginning of Namecoin Merged Mining ? .....how about it, there, 'Honest Abe' ?

Slush, you are a coward.

WTF are you moaning about? Namecoins earned at the beginning were distributed to those that had addresses posted, and those folks had (in my opinion) adequate time to prepare. Or are you crying about something that the voices in your head invented?

[slush]

re namecoins: lol, please don't start again :-).

Why not be a man and just come out and call BTCGuild a cheat

Because I'm not talking that btcguild cheated, but that the stats is telling that deepbit and me are honest. Do you see the difference?

Btw me and tycho are NOT friends, but we can talk each other when it's important.

[rjk]

Bullshit he never told anyone he was doing the merged mining just did it like thief in the night, no one had time to prepare for anything when it was done in secret. Then once done it took well over week to get the coins paid out as in his unseemly haste to scam the coins he had no payout method in place costing people well over third of the value of their coins when the flood of coins on the market had already taken place. And you got the nerve to talk about inventing shit...

A third of their value? That is definitely bullshit. I recall the value of namecoins going UP after payouts were enabled and moving smoothly. If you didn't cash out in time, it's your fault.

[slush]

You're seriously ill.

Nah, I cannot provide you any arguments because I provided them months ago and this is closed case for me. Actually I cannot believe that somebody is still solving this "issue".

And you got the nerve to talk about inventing shit...

facepalm