Visual comparison of pool payout methods based on real-world data

[Luke-Jr]

I finally have real-world example graphs for most payout methods (warning: heavy-load webpage), based on real-world data from Eligius for a random ~800 MHz miner over a ~4.5 day period. I also simulated the graphs as if he were pool hopping by reassigning his shares beyond 43% of a block to another user. This way, you can see visually how he would have been paid under the Slush Score, Proportional, Maximum PPS, PPS, and PPLNS systems. I would have added the Geometric method too, but I couldn't figure out how to make it work.

The pool hopper spends about 2/3 of his time mining for other pools.

Please use this thread to comment on the various methods (especially as they relate to Eligius's future), nominate new methods (which I might simulate a graph for, if it seems sensible), etc...

Edit: Nominations for other methods should be made in this thread

These graphs took many hours to code and generate. If you find them useful, please donate: 12GJ27Fr9Wme2JT3ZkQMhzvHXNE8MXrAux

P.S. congrats 1AjsBb..., you're famous!

Click here for latest version

[Artefact2]

That's very nice. PPLNS is my favorite by the way ! 

[gentakin]

I'd like to know how long the pool hopper is not actively mining because the pool is above 43% of the difficulty (that is, how much time can the hopper spend on a different pool?). In percent.

It helps to figure out by how much the pool hopper gets overpaid for each payout method.

[Luke-Jr]

I'd like to know how long the pool hopper is not actively mining because the pool is above 43% of the difficulty (that is, how much time can the hopper spend on a different pool?). In percent.

It helps to figure out by how much the pool hopper gets overpaid for each payout method.

About 2/3 of the pool hopper's shares were spent mining for other pools.

[gentakin]

It's a pretty interesting chart. Thanks for your work on it. I will send the results of a few hours of my mining work to you. Which is not that much, but still.


A fair reward for the pool hopper is probably about 4.2/3 = 1.4BTC. With prop, he gets 2.7BTC. That's pretty.. unfair.

(Now I notice I could have figured out the pool hopper's away-from-poolness ratio by looking at PPS..)

[shamathana]

is poolhopping still an issue? the graphs i saw on eligius, are quite constant 24/7.

if i wanted slush, i'd go to slushs server.

if i wanted pps, i'd go to one of the many pps-servers.

maximumpps stands against the no registration needed, because every time u change, u have to get ur maximum up again.

pplns looks like the über ripoff:#
Variables: N = difficulty * 2
# Easy math: (your shares / N shares) * reward
basically it's like prop but the size is always two times time the average ( N = difficulty * 2). meaning it's averaged out prop/2.
on averge u would be ripped off 50%, on faster rounds even more, on doulbe than average rounds it would be like prop, and on longer than 2*average round u would decrease the 50%profit of the 50btc + fees margain of the operator.

eligius is the best(2nd best when it comes to payouts only) of 5 servers i have tested. never change a running system, just fix the us-server.

[Luke-Jr]

is poolhopping still an issue? the graphs i saw on eligius, are quite constant 24/7.

Whether people are actually doing it or  not, I don't know. I don't think the hashrate graph gives enough info to tell. With proportional, pool hopping is a good idea and a miner's right. The problem is that it will kill the pools involved eventually.

if i wanted slush, i'd go to slushs server.

if i wanted pps, i'd go to one of the many pps-servers.

Slush and PPS are only on there for comparison/reference. I'm not considering either as viable options.

maximumpps stands against the no registration needed, because every time u change, u have to get ur maximum up again.

You still don't need to register. I agree, the lack of being able to change addresses on a whim is a concern. Any suggestions to deal with it?

pplns looks like the über ripoff:#
Variables: N = difficulty * 2
# Easy math: (your shares / N shares) * reward
basically it's like prop but the size is always two times time the average ( N = difficulty * 2). meaning it's averaged out prop/2.
on averge u would be ripped off 50%, on faster rounds even more, on doulbe than average rounds it would be like prop, and on longer than 2*average round u would decrease the 50%profit of the 50btc + fees margain of the operator.

Except that the last N shares are always paid. So except for blocks at least 2*diff shares long, your shares on the previous block(s) would get paid more than once.

eligius is the best(2nd best when it comes to payouts only) of 5 servers i have tested. never change a running system, just fix the us-server.

The proportional system is self-defeating due to pool hopping. As far as actual earnings, the graphs clearly show that for the ordinary miner, all the systems pay more or less the same.

[HanSolo]

Very nice charts!

Regarding how they show PPLNS (http://eligius.st/wiki/index.php/Pay-Per-Last-N-Shares):

If people care a lot about some shares earning zero after unlucky periods, then N could be much larger: say difficulty * 10. This means it takes longer before a contributed share is sure it's collected all its payments, maybe days, but also that hardly any shares get a total goose-egg. In contrast to MaxPPS, it is not required that a miner continue contributing in hopes things eventually 'even out' on their account. They eventually get credited their earnings-by-rule (or the shares expire by moving out of the last-N-window) as long as the pool continues operating.

In my mind one of the largest benefits, worthy of highlighting via a bottom-bullet, is that PPLNS minimizes the record-keeping and reserve-holding of the pool operator. Even with 100% trust in the operator, holding a reserve balance at the pool involves risks: real-life crises, security-break-ins, bitcoin economy crashes, etc. A system which allows most accounts to be 'zeroed' asap after generation fees earned means less risk.

(A way to indicate this graphically could be to show the actual surplus of generation fees over payouts as another data series, or in the case of pure PPS, the required deficit the operator sometimes has to cover. The determinism of PPLNS means neither surplus nor deficit need be carried at any point.)

[bitcoin0918]

---

[Luke-Jr]

In my mind one of the largest benefits, worthy of highlighting via a bottom-bullet, is that PPLNS minimizes the record-keeping and reserve-holding of the pool operator. Even with 100% trust in the operator, holding a reserve balance at the pool involves risks: real-life crises, security-break-ins, bitcoin economy crashes, etc. A system which allows most accounts to be 'zeroed' asap after generation fees earned means less risk.

I'm not sure this is a real benefit for practical reasons:

• Full record-keeping is still needed for auditing
• Reserve-holding is needed only until the minimum payout is met (and there is enough generation to make the payouts) under all systems

It does remove the requirement for maintaining non-reward credit, but that is strictly a negative of the MaxPPS system, not so much a benefit of this one IMO.

[HanSolo]

In my mind one of the largest benefits, worthy of highlighting via a bottom-bullet, is that PPLNS minimizes the record-keeping and reserve-holding of the pool operator. Even with 100% trust in the operator, holding a reserve balance at the pool involves risks: real-life crises, security-break-ins, bitcoin economy crashes, etc. A system which allows most accounts to be 'zeroed' asap after generation fees earned means less risk.

I'm not sure this is a real benefit for practical reasons:

• Full record-keeping is still needed for auditing
• Reserve-holding is needed only until the minimum payout is met (and there is enough generation to make the payouts) under all systems

It does remove the requirement for maintaining non-reward credit, but that is strictly a negative of the MaxPPS system, not so much a benefit of this one IMO.

Yes, all systems require balances to meet some minimum threshold (or age).

MaxPPS also holds at the pool amounts generated that aren't yet matched by 'value'. These could be larger than the minimum payout.

To illustrate with an extreme example, assume a MaxPPS system with 50 miners. each submits 1 share, the 50th makes a full block. They each get 1BTC in 'earnings' but only an infinitesimal amount in 'value'. So they are credited tiny amounts, but (50-epsilon) is held indefinitely by the pool operator, until some overlong-round arrives.

If the next 10 rounds all take exactly the expected amount of shares, then for those, they each get 1BTC value/earnings/payout. But until an overlong round arrives, that first-round surplus is still held at the pool.

If some miners dabble and then leave, and thus never catch up on 'value', that reserve is held.. forever? (Or does it eventually revert to the operator or other miners?)

[twmz]

Even though it is more complicated to wrap your head around, I prefer MaxPPS.  It is a smoother payout than proportional, doesn't reward pool hopper if they bail on long blocks, and doesn't risk the pool being shut down from losing BTC like PPS does.

PPLNS has simpler math, but that doesn't mean the logic of it is easy to grasp. I guess the concept is that each time a block is found, we'll pay you for your relative contribution to the pool's total hashrate over the recent past (not specific to the block that was just found).  Once you get past the start, the payout pattern seems identical to proportional for ordinary miners, so it isn't any smoother of a curve.  You lose out right at the start because for the first couple blocks you are not getting credited for shares from previous blocks.  You recoup that initial deficit eventually when you stop mining and keep getting paid even after you quit (for the last few shares you submitted).

I prefer MaxPPS simply because it has more consistent payouts than PPLNS (which has payouts that vary just as much as proportional).  With MaxPPS, on any given 24 hour period, I will get paid closer to what I would have been paid with PPS.

That said, I could live with either.

[Luke-Jr]

MaxPPS also holds at the pool amounts generated that aren't yet matched by 'value'. These could be larger than the minimum payout.

This is not relevant, as it isn't funds due to anyone yet.

To illustrate with an extreme example, assume a MaxPPS system with 50 miners. each submits 1 share, the 50th makes a full block. They each get 1BTC in 'earnings' but only an infinitesimal amount in 'value'. So they are credited tiny amounts, but (50-epsilon) is held indefinitely by the pool operator, until some overlong-round arrives.

They get paid for the work they did. The remainder is not owed to them, and won't be until they do the work to earn it.

If some miners dabble and then leave, and thus never catch up on 'value', that reserve is held.. forever? (Or does it eventually revert to the operator or other miners?)

It's not a reserve. They have non-reward credit noted on their account, in case they decide to mine again, and at that time the pool might owe them more than it makes at that point in time.

Basically, MaxPPS is nothing more or less than PPS with safeguards for the pool operator, which allows him to charge a lower fee, or even no fee at all.

[HanSolo]

MaxPPS also holds at the pool amounts generated that aren't yet matched by 'value'. These could be larger than the minimum payout.

This is not relevant, as it isn't funds due to anyone yet.

I understand that's the logic of MaxPPS. But, they are real BTC funds delivered from the blockchain. They are under the control of the pool operator. And, if someone who has already built up a pending 'earnings' amount later becomes due that BTC by having their 'value' catch up, they need those funds to be safe in order to get paid.

What if in the meantime, you had a security break-in and those BTC were stolen? Or a catastrophic data-loss situation where you lost the keys to pay that out?

Yes, this is also a risk with minimum payouts but it's larger with MaxPPS. And given inactivity-payouts, it's a capped risk. Amounts due tend toward zero very quickly. With MaxPPS, credits-available-if-you-mine, which are matched by real pool-operator BTC held balances, can grow indefinitely.. especially if many miners wander away.

And, since those funds won't be used to pay anyone else, for example someone new who generates extra 'value' but not yet matching earnings, it's not the full story to say those funds aren't "due to anyone yet". They seem to be in an account-specific reserve. That makes them very susceptible to indefinite abandonment.

If that account never shows up with more mining, to get their 'earnings' and 'value' caught up, is that amount held in reserve in case they again show up forever? Or is it eventually redeployed to pool costs or other recently-active miners?

Practically, you might want to declare an expiration time for these unvested earnings, before it becomes an issue.

[Artefact2]

Even though it is more complicated to wrap your head around, I prefer MaxPPS.  It is a smoother payout than proportional, doesn't reward pool hopper if they bail on long blocks, and doesn't risk the pool being shut down from losing BTC like PPS does.

PPLNS has simpler math, but that doesn't mean the logic of it is easy to grasp. I guess the concept is that each time a block is found, we'll pay you for your relative contribution to the pool's total hashrate over the recent past (not specific to the block that was just found).  Once you get past the start, the payout pattern seems identical to proportional for ordinary miners, so it isn't any smoother of a curve.  You lose out right at the start because for the first couple blocks you are not getting credited for shares from previous blocks.  You recoup that initial deficit eventually when you stop mining and keep getting paid even after you quit (for the last few shares you submitted).

I prefer MaxPPS simply because it has more consistent payouts than PPLNS (which has payouts that vary just as much as proportional).  With MaxPPS, on any given 24 hour period, I will get paid closer to what I would have been paid with PPS.

That said, I could live with either.

You are right. Yet, maxPPS never pays more than proportional, and, you said it yourself, PPLNS payouts are very close to the proportional payouts. I'd rather have my payouts depend on the pool's luck rather (and get them immediately) than having constantly an (even tiny) amount of my reward "withhold" by the pool in case of "unlucky" rounds.

[Luke-Jr]

MaxPPS also holds at the pool amounts generated that aren't yet matched by 'value'. These could be larger than the minimum payout.

This is not relevant, as it isn't funds due to anyone yet.

I understand that's the logic of MaxPPS. But, they are real BTC funds delivered from the blockchain. They are under the control of the pool operator. And, if someone who has already built up a pending 'earnings' amount later becomes due that BTC by having their 'value' catch up, they need those funds to be safe in order to get paid.

What if in the meantime, you had a security break-in and those BTC were stolen? Or a catastrophic data-loss situation where you lost the keys to pay that out?

Then it can pay from future generation, at the expense of the excess getting into the pool wallet. It all evens out, that's why it's even workable.

If that account never shows up with more mining, to get their 'earnings' and 'value' caught up, is that amount held in reserve in case they again show up forever? Or is it eventually redeployed to pool costs or other recently-active miners?

Practically, you might want to declare an expiration time for these unvested earnings, before it becomes an issue.

Since it isn't theirs, it isn't necessarily in reserve. Of course, not keeping it as a reserve would create a liability to the pool operator should they come back later. Since pool hopping will be pointless and PPS generally a losing proposition, I expect there won't be all that much size to these abandoned NRC (non-reward credits). Specifying an expiration for such credits might be a good idea just in case-- perhaps it could help defer/reduce fees.

[shamathana]

is poolhopping still an issue? the graphs i saw on eligius, are quite constant 24/7.

Whether people are actually doing it or  not, I don't know. I don't think the hashrate graph gives enough info to tell. With proportional, pool hopping is a good idea and a miner's right. The problem is that it will kill the pools involved eventually.

if thats true, then why did u implement it in the first place?! why don't u know if they r hopping, aren't u the admin with access to lots of stats?! because if they r not, we could cut the whole thread quite short! and ur pools hashpower rises and rises.

maximumpps stands against the no registration needed, because every time u change, u have to get ur maximum up again.

You still don't need to register. I agree, the lack of being able to change addresses on a whim is a concern. Any suggestions to deal with it?

not with keeping maxpps. but it's sounds shitty anyway maxpps = pps or less. shouldnt pps stand for a guaranteed value, which maxpps does not? also i chose eligius, because in the end everything gets paid out after a week, but with maxpps that would not be the case. so i would change the server, because it looks like the worst system to me.

pplns looks like the über ripoff:#
Variables: N = difficulty * 2
# Easy math: (your shares / N shares) * reward
basically it's like prop but the size is always two times time the average ( N = difficulty * 2). meaning it's averaged out prop/2.
on averge u would be ripped off 50%, on faster rounds even more, on doulbe than average rounds it would be like prop, and on longer than 2*average round u would decrease the 50%profit of the 50btc + fees margain of the operator.

Except that the last N shares are always paid. So except for blocks at least 2*diff shares long, your shares on the previous block(s) would get paid more than once.

ahh u should stress the retroactively part, where u get paid on average just half but it dates back twice as much, so it should average out(except really long round which own miners). now it makes sense.

interessting system. but it would just change hopping styles:
as i get it, u now join fresh rounds and leave when they take to long, and come back if a new one starts.
with pplns it's just the opposite. u join the long rounds, because they should be over soon and hopefully followed by shorter rounds, so u get paid triple and more. then leaving at the beginning of a new round in fear of a long round, where the first shares aren't worth anything. and being worth nothing should be a even stronger driving factor than being worth less.

all those systems have their ups and downs. but if u wanna combat hopping, pplns doesn't look to good. pps should be king, or if u wanna burden the miners with the risk it's maxpps.

[Luke-Jr]

is poolhopping still an issue? the graphs i saw on eligius, are quite constant 24/7.

Whether people are actually doing it or  not, I don't know. I don't think the hashrate graph gives enough info to tell. With proportional, pool hopping is a good idea and a miner's right. The problem is that it will kill the pools involved eventually.

if thats true, then why did u implement it in the first place?! why don't u know if they r hopping, aren't u the admin with access to lots of stats?! because if they r not, we could cut the whole thread quite short! and ur pools hashpower rises and rises.

Whether it's happening or not, it's basically stupid to not hop a prop. pool. And doing so will kill them.

maximumpps stands against the no registration needed, because every time u change, u have to get ur maximum up again.

You still don't need to register. I agree, the lack of being able to change addresses on a whim is a concern. Any suggestions to deal with it?

not with keeping maxpps. but it's sounds shitty anyway maxpps = pps or less. shouldnt pps stand for a guaranteed value, which maxpps does not?

MaxPPS isn't PPS. PPS carries an infinite risk for the pool operator. MaxPPS is PPS without that risk. ie, you get as much as PPS would pay, so long as it doesn't kill the pool.

also i chose eligius, because in the end everything gets paid out after a week, but with maxpps that would not be the case. so i would change the server, because it looks like the worst system to me.

No matter which system is used, including MaxPPS, everything you earn will still be paid out after a week.

pplns looks like the über ripoff:#
Variables: N = difficulty * 2
# Easy math: (your shares / N shares) * reward
basically it's like prop but the size is always two times time the average ( N = difficulty * 2). meaning it's averaged out prop/2.
on averge u would be ripped off 50%, on faster rounds even more, on doulbe than average rounds it would be like prop, and on longer than 2*average round u would decrease the 50%profit of the 50btc + fees margain of the operator.

Except that the last N shares are always paid. So except for blocks at least 2*diff shares long, your shares on the previous block(s) would get paid more than once.

ahh u should stress the retroactively part, where u get paid on average just half but it dates back twice as much, so it should average out(except really long round which own miners). now it makes sense.

Every system proposed averages out for ordinary miners.

interessting system. but it would just change hopping styles:
as i get it, u now join fresh rounds and leave when they take to long, and come back if a new one starts.
with pplns it's just the opposite. u join the long rounds, because they should be over soon and hopefully followed by shorter rounds, so u get paid triple and more. then leaving at the beginning of a new round in fear of a long round, where the first shares aren't worth anything. and being worth nothing should be a even stronger driving factor than being worth less.

Maybe, but this might be flawed with the gambler's fallacy: there's no reason to expect a short round after a long one. And really, there are no rounds with PPLNS at all.

all those systems have their ups and downs. but if u wanna combat hopping, pplns doesn't look to good. pps should be king, or if u wanna burden the miners with the risk it's maxpps.

PPS is king for miners obviously, but that means I'd be left paying them out of pocket at a loss. I could only afford that at probably 15% fee, which is IMO ridiculous. MaxPPS is the only viable way to do anything even close to feeless PPS.

[shamathana]

interessting system. but it would just change hopping styles:
as i get it, u now join fresh rounds and leave when they take to long, and come back if a new one starts.
with pplns it's just the opposite. u join the long rounds, because they should be over soon and hopefully followed by shorter rounds, so u get paid triple and more. then leaving at the beginning of a new round in fear of a long round, where the first shares aren't worth anything. and being worth nothing should be a even stronger driving factor than being worth less.

Maybe, but this might be flawed with the gambler's fallacy: there's no reason to expect a short round after a long one. And really, there are no rounds with PPLNS at all.
[/quote]
yes yes, a long last round does increase the chance that the next will be a fast one, but i think many believe that, there is so much scam about this on youtube.

anyway,you gamble in both cases in hope of a short round. with prop its the round u just started, if ur lucky all share will have more value than on average. with pplns it's the following round, all shares will get paid triple or even more.
in prop shares at the beginning may have a higher value because of a short round and the longer the round the less the value.
opposite with pplns: shares at the beginning have a chance of not being worth anything at all, but shares at the end may be worth triple, quadruple or even more...
in both hopping can be nice if ur lucky, u just have to hop at a different points in time.

PPS is king for miners obviously, but that means I'd be left paying them out of pocket at a loss. I could only afford that at probably 15% fee, which is IMO ridiculous. MaxPPS is the only viable way to do anything even close to feeless PPS.

make ur offer :-P
bitcoins.lc sharevalue is atm @ 9,9972462987887*10^-5
ur's is atm @ 6,93413218269838*10^-5
deepbit.net with pps is @ 5,129803015564*10^-5



and whats with the 15% with pps, don't u "just" need a financial cussion to start with and after that just some % to keep it filled?

[Luke-Jr]

PPS is king for miners obviously, but that means I'd be left paying them out of pocket at a loss. I could only afford that at probably 15% fee, which is IMO ridiculous. MaxPPS is the only viable way to do anything even close to feeless PPS.

make ur offer :-P
bitcoins.lc sharevalue is atm @ 9,9972462987887*10^-5
ur's is atm @ 6,93413218269838*10^-5
deepbit.net with pps is @ 5,129803015564*10^-5

In human-readable decimal:
bitcoins.lc and (current) Eligius are proportional. So there is no "share value" to speak of.
Actual value of a share at the current difficulty is 0.000056996941566467068322407
deepbit.net is (in normal decimal number format) 0.00005129803015564
Eligius would be, if MaxPPS is used, about 0.000056996884569525500216354

and whats with the 15% with pps, don't "just" need a financial cussion to start with and after that just some % to keep it filled?

Yes, that % is 15.