[BITLC] Change of payment scheme for Bitlc.net (Previously bitcoins.lc)

[Jine]

Hi!

It's about time to jump on the train and switch from proportional to some other payment scheme.
This is a a poll for our users (and non-users, and future users) to check what you guys actually want to have.

There is multiple pros/cons for all payment schemes, therefor we want to make this a democratic decision and let you guys decide.

Here's a short description of the alternatives (copy'n'pasted from wiki)

DGM - Double Geometric Method. A hybrid between PPLNS and Geometric reward types that enables to operator to absorb some of the variance risk. Operator receives portion of payout on short rounds and returns it on longer rounds to normalize payments. [1]
PPLNS - Pay Per Last N Shares. Similar to proportional, but instead of looking at the number of shares in the round, instead looks at the last N shares, regardless of round boundaries.
PPS - Pay Per Share. Each submitted share is worth certain amount of BC. Since finding a block requires <current difficulty> shares on average, a PPS method with 0% fee would be 50 BTC divided by <current difficulty>. It is risky for pool operators, hence the fee is highest.
Score - Score based system: a proportional reward, but weighed by time submitted. Each submitted share is worth more in the function of time t since start of current round. For each share score is updated by: score += exp(t/C). This makes later shares worth much more than earlier shares, thus the miner's score quickly diminishes when they stop mining on the pool. Rewards are calculated proportionally to scores (and not to shares). (at slush's pool C=300 seconds, and every hour scores are normalized)


Either way, we'll be starting to include the block fee's into the user reward when the new payment scheme is implemented.

What do you guys think? Something got to be done.

--
Regards, Jim
Bit LC Inc.
https://www.bitlc.net


EDIT: I hit enter to fast, submitted the thread before i even finished writing.
Fixed now...

[1715128510]

1715128510

[1715128510]

1715128510

[1715128510]

1715128510

[1715128510]

1715128510

[mcorlett]

Won't all the poolhoppers just vote for proportional just so that they can continue exploiting the system?

[phungus]

Why does something have to change, exactly?

-p

[Jine]

We will NOT stick with proportional due to a lot pressure from non-hoppers and regular 24/7 mining users have requested that we do actually start using something else.
So yes, things will change, but the question is - to what?

Personally I'd love to stick with proportional if there was something to be done against hoppers...
PPLNS is the closest scheme to what we have now, and will solve the problem for all miners that's mining more or less fulltime.

DGM is "better", but a lot harder to implement and isn't that similar to proportional (you get paid over a longer period of time etc)..

Score based is proved to work and scale, as slush is using it... seems stable enough to me.

I let you guys decide, but this was just some of my thoughts.
PPS comes with a risk for me, so even if it wins - I'll have to think that over quite a bit...

[film2240]

Stay with proportional please as none of the systems seemed right for me (I tried) plus the other systems seem to give me a lot less for the same efforts so please stay with proportional.

[eleuthria]

Jine - Contact me in IRC.  I can give you some numbers about PPS and why I would STRONGLY discourage you from choosing that option given your pool size.

[jjhall]

Many of us have discussed this at length, including you and I, in the past on IRC.  The biggest reason to switch was in order to discourage pool hoppers from leaving the pool.  After much discussion it was (I thought) decided that pool hopping doesn't actually hurt the pool or the workers in it.  Yes they reduce the payout for all of the shares since there are more of them, but they contribute to more frequent blocks so it all comes out in a wash.  It has been quite a while since I've participated in any discussions, so if the overall consensus has changed, then I digress.

To me determining why is it time to change the payout method affects the preferred method.  If it is "just because" then my vote is to leave it as-is.  "If it ain't broke, don't fix it."  If there is a reason other than an (arguably) irrational fear of pool hopping, choosing a method that best answers to that reason is the way to go.  Is it because the risk is unbalanced between the pool and the workers, then choose the method that shares that risk in the most fair manner.  Is it because there is too much confusion on how proportional payouts work?  Then choose the method that is easier to understand (is there one?)  Is it because proportional is too much load on the servers to calculate payout?  Choose the method with the least load to complete the payout calculations.

In the long run, assuming any fees/charged by the pool are reasonable, it all works out the same for people who have workers that sit and stay in the pool, and payout calculation method really doesn't matter.

Just my 2 bitcents worth.

Jeremy

[Jine]

The reason for me wanting to switch payment scheme comes from the anger and frustration of pool-hopping-pools. (Abcproxypool for example)
It's not a single users that hops us, it's entire pools (which results in slowdowns and frustration from my side) - i know that eleuthria agrees on this.

So yes - we're switching cause of pier pressure and to "escalade" and in the long run - be a better pool.
As I've said before - we're here for the long run.

--
Regards, Jim
Bit LC Inc.
https://www.bitlc.net/

[RoloTonyBrownTown]

We will NOT stick with proportional due to a lot pressure from non-hoppers and regular 24/7 mining users have requested that we do actually start using something else.

Shame.  P2pool is looking better by the day.

[joulesbeef]

It's not like the anti-prop folks dont have any pools to choose from.. are you losing users? or are people just being loud for the sake of being loud?

Vote with your feet I say, or in this case with your hashes.

[martychubbs]

Won't all the poolhoppers just vote for proportional just so that they can continue exploiting the system?

nope

[malavita]

Sorry for the newbie question, but can someone explain to me how you can exploit the system by pool-hopping?  Thx!!

[jkminkov]

The reason for me wanting to switch payment scheme comes from the anger and frustration of pool-hopping-pools. (Abcproxypool for example)
It's not a single users that hops us, it's entire pools (which results in slowdowns and frustration from my side) - i know that eleuthria agrees on this.

So yes - we're switching cause of pier pressure and to "escalade" and in the long run - be a better pool.
As I've said before - we're here for the long run.

--
Regards, Jim
Bit LC Inc.
https://www.bitlc.net/

is that hard to ban 10Gh+ hopper or review 10+ btc per 24h payments by hand?

I want ESMPPS -tx fees

[Starlightbreaker]

10gh?

more like 400gh 

[mcorlett]

Won't all the poolhoppers just vote for proportional just so that they can continue exploiting the system?

nope

You're only confirming my theory by doing that.

[jkminkov]

my theory is that regular miners are sheep, you can put 10% pool tax(see btcguild hashrate) and they won't figure it out for months even if you put it on front page with size 24 font...

[padrino]

I'm not sure exactly what the best method is but I think something needs to be done, if anything to help stabilize and work to grow bitlc. If miners that are unhappy vote with their feet the pool will definitely loose some real hashing power but I suppose that must be balanced with the fact that right now he attracts a large number of hoppers, is that good for the consistent users, I say no. Is it good for the pool operator, I can't say but I assume because of the 0% fee the answer is no.

I do like the point about hoppers earlier in the thread though, if I were one of the 200+ Gh/s that hops I guess I would vote to stay on the current method.

[jjhall]

Well I'd voted for keeping it as-is before learning that hopping-pools are actually causing undue stress on the system and/or operators.  Unfortunately you can't go back and change a vote, so please disregard one of the proportional votes.  I think the best way to choose is to compare and contrast what the various methods actually mean to the pool and its members.

DGM - I haven't gone and looked up how the calculation is actually done, but on the surface it sounds a bit complicated.  The pool absorbing some of the fee from shorter rounds and paying it out on the longer rounds seems to add a bit of accounting overhead.  I would think there'd be risk of pool members accusing the pool of trickery if they don't understand precisely how it works.  Unfounded negative publicity would be a very bad thing considering how paranoid some bitcoin users are to begin with.  Wouldn't this also expose the pool to risk that a large number of subsequent long rounds causing the pool to constantly pay out without any income to compensate for it?  Long term this should average out, but the pool would need enough capital to survive shorter-term ruts.
PPLNS - This is a very transparent methodology, and is simple to calculate.  The negative here is the shares become completely worthless over time.  I think many users that only mine part-time (not hoppers, but those that mine when not gaming, etc...) would tend to not like this method for the same reason hoppers don't.  This is a non-issue for those of us who mine 24/7 since we're constantly replacing the shares that roll off with new shares.
PPS - I wouldn't even consider this for the pool.  Right now BitLC is built on being a fair pool, not charging fees to the miners, and being open and transparent.  Switching to PPS would cause undue risk to you as the pool operator, and/or would require the pool to charge fees.  The first is not fair to you, and the second goes against the philosophy of the pool (at least how I perceive it as a pool member.)
Score - This may be the best choice, since it eliminates the "worthless" share issue with PPLNS.  A part-time miner still gets "something" if they stop mining during a round.  It would ease their minds getting a miniscule "token" amount for the work put in.  Hoppers would still get paid some too, but they're in it for the quick reward and the token payment they receive should not be enough to have them view this pool as a valid target.

For full-time 24/7 miners the payout method really means nothing in the long haul; it all averages out.  The change in payout method needs to be used as a method to attract part-time individual miners that come and go based on individual PC habits, but not give enough advantage to make it a viable pool for hopping purposes.  Score-based payout is probably a good balance and would be my first choice, but if you wanted to follow the KISS principal, PPLNS is the way to go, and I'd consider that my second choice.

No matter what you decide, thank you Jim for all of the hard work you put into running the pool.  I know it isn't easy, and to do what you do without charging fees is saying a lot.

Jeremy

[martychubbs]

Won't all the poolhoppers just vote for proportional just so that they can continue exploiting the system?

nope

You're only confirming my theory by doing that.

Very scientific.

Jine, prop is how you are going to grow ATM...let the market decide.  Add a fee to prop like nmcbit?  If your avg hashrate drops, the market will confirm your decisions as foolish. Like Masterpool, Triple, and MtRed...  

Notice how pool ops are moving to  giving members a less that 100% return  on their hash power and then use the aggregate hash power to pool hop or kill crypto currencies.  Just like a corrupt politician, pacify the decent while you are milking everyone! ABCproxyPool and Eligius using bH's port "http://mining.eligius.st:8337"

I love how Pool Ops get persuaded to change things by some members at the risk the members that made the pool grow.  

McDonald's still makes burgers and fries (what made them,) not only pseudo bbq rib sandwiches, coffee or salads...   So, make a PPLNS or DGM pool along side Prop configured via the worker and see where the hash power gravitates to.  That way you don't lose members and you give members options...win/win and you can test your theory in production

In marketing, you should study the "new coke" mistake...  Pool ops should run the pool as a business and they would be less likely to change things without understanding ALL the costs to the business.

[deepceleron]

My preference is in my sig...

A rewards system is best and most transparent if your payout still comes when the pool finds a block, as long as the pool hashrate is high enough to give miners several daily block finds. It lends to enthusiasm about finding blocks and doesn't hide how Bitcoin works. The rewards system simply determines how the 50BTC block generation reward is split between pool miners, and we want it to be fair and easy. We know that proportional is not an option because it can be gamed by pool hopping.

Geometric formulas that alter or "decay" the value of shares give a level of obfuscation where you can't determine what your shares will earn. A few other pools already have gone to geometric rewards (and PPS with fee), so you wouldn't be offering anything unique by using those. PPS pool ops can be attacked by block withholding. Slush's formula is not perfect and still hoppable. SMPPS you don't have in the poll - good, because it is a PPS, but lowers the reward when the pool gets unlucky (this variability makes it still hoppable).

The only miner complaint (from those who don't understand the system) is "expiring shares". PPLNS, however, also allows for multiple block wins from the same submitted share, so you can be unexpectedly surprised and rewarded by pool luck (on a proportional pool, all shares expire once a block is found, of course). When you submit a share, it will earn you a fixed payout, 1/N * 50BTC, for any block finds during the next N pool shares after submission. The N number can be set higher than difficulty if users prefer a longer window, so during an unlucky stretch there are fewer shares that expire with no reward, or can be set lower, if users want faster bigger payouts. Any value of N still gives both full-time and part-time miners the same expected reward. (Note: N should not be time-based, expiring after 12 hours for example, as this breaks the payout system by making share value variable and exploitable.) Mining is really a lottery anyway - if you solo mine, for example, every hash that didn't find a block instantly "expires"; when you mine and don't find a block, you really contributed zero work to the pool.

Consider any votes for proportional as coming from the pool hoppers. When the posting history is in many pools and pool hopping threads, we know they are the pool exploiters. Pool hoppers don't contribute anything by their straw-man argument that they "help find blocks"; the hashrate of the pool doesn't affect earnings (I was very happy with mineco PPLNS even at 80ghash). When everyone but the pool hoppers leave, there will be one last time that the pool crosses 43% of difficulty shares, and then it will be dead.

PPLNS worked very well on mineco. It doesn't have any pool-op risk requiring extra fees. Miners still get joy when a block is found.