As I said, Ideology is the primary motivation in my selecting a pool and/or reward system.  Both 0% PPS and SMPPS pools are doomed to collapse and, therefore, unattractive to me.

Also, there is a significant difference between 0% PPS and SMPPS.

With 0% PPS the risk lies with the pool.  With SMPPS the risk lies with the miners.

From a miner's perspective, a share submitted to a 0% PPS pool has the same value as one submitted to a SMPPS pool with a positive buffer.  However, with a negative buffer, the miner must accept the risk that the pool will never again achieve a positive buffer and that they will never be paid for their work.  When the buffer is just below 0 this risk is small (because there are so many irrational miners) but the further from 0 the buffer goes, the larger the risk.  Even if you owned all of the shares submitted after the buffer became negative and were the only one left at the pool it would be rational to give up on the pool and move to solo mining.  If you don't own all of the risky shares and/or there are other people are still mining with the pool the incentive to hop away is even greater.

Admittedly, I found that yourbtc.net doesn't have as much information on the site about the reward system as I'd have liked.  I cannot speak for eclipsemc.com.  You can often find more information on the forums.  The first page of the main yourbtc.net thread for example links to much more information.  The fact that this thread contains positive posts from respected and knowledgable members of the pool mining community (Inaba and Mein Rosenfeld) is a good sign in my book.

It is true that, when a block is found, more recent shares count for more.  You get a smooth descent rather than a clean drop-off.  I only suggested it because the parameters of the current implementations of DGM are such that they almost correspond to PPLNS with a relatively large value of N.  If the clean drop-off is important to you then DGM is not an option, there are advantages and disadvantages and it largely comes down to a matter of taste.  Personally, I prefer the clean drop-off too and have used PPLNS a lot in the past but I fancied a change and, as far as reward systems go, DGM (with non-trivial parameters) is a close second for me.  Anyway, it was just a thought.  Happy mining!

Note: My choice of reward system and pool is driven primarily by ideology.  If the only options were Proportional, SMPPS, and PPS+10%, I'd mine solo.  I'd happily pay 5% for PPS or 4% for a good PPLNS/DGM pool.

Agreed.  This establishes why the following thought experiment doesn't apply to Bitcoin.


In this example, keeping track of failed guesses is very helpful because you will expect the number of failed guesses for a round to be much more than the number of different solutions (one in this case).  But with Bitcoin, for every block the number of failed hashes (worldwide) is utterly insignificant compared to the number of potential solutions.  Even if everyone were using the same pool and using random nonces there would be practically no efficiency loss due to repeated work.

I nearly did once.  I tried to buy 100 BTC but misunderstood the order form on the site and ended up offering 100-fold what I wanted to!  Fortunately, the site owner was able to reverse the mistaken trade for me.

Dumb but lucky!

If you like the idea behind PPLNS but dislike the high share-variance inherent to using a small value of N then perhaps consider trying a pool with using the double-geometric method such as eclipsemc.com or yourbtc.net.  Both of these pools have merged mining and 0% fee.  I can't speak for eclipsemc.com but yourbtc.net has been enjoying good luck in the last few days on both blockchains.

If you check out http://yourbtc.net/content/overview-bitcoin you'll see that the actual donation is much less than 2%.


I don't doubt that, for DGM, the best approach in the long term would be to change a fee to offset the intrinsic pool operator risk.  In the reasonably short term (order months - sorry I wasn't clear) I think it will be tough for a small pool that needs income to survive.  In this context it might make sense to start requiring a fee but I feel such a fee should be used as efficiently as possible.  I'd rather see it all spent on the server and operator's time and have the parameters of DGM tweaked so that the risk is covered by donations.  This will mean higher share variance for a lower fee and I must admit I'm biased towards such models.

That said, if this pool started charging a fee to offset pool operator risk I'd still continue to use it so I'm probably best ignored.

There could be a number of reasons to mine at a pool with a 5-7% fee:
1) Low variance
2) Reputation
3) Server reliability
4) "They change a fee, they must be good!" thinking
5) Plausibly sustainable business model implies future planning.

It's not anything like downclocking.

If you are running into problems with pool risk at large hashrates then perhaps you can tweak the parameters of DGM (making it more PPLNS like) so that you have less risk.

I would imagine that a fee should be brought in at some point and it should be done purely to pay for server costs and for your time.  At the moment you (and arguably gigavps) are providing a free service and you're simply not going to be able to sustain this on donations alone.

I didn't know eclipsemc.com was using merged mining already.  That's good to know!


I fully agree that we don't want "pool police" and that pools should simply outline their reward system to attract miners.  However, SMPPS was originally pushed as a way of eliminating the essential problem with 0% PPS mining where, in fact, all that changed was that the risk was transferred from the pool operator onto the miners.

I have no problem with pool operators using the proportional reward system but do have a problem if they suggest that the system is, in any way, fair to 24/7 or random-intermittent miners.

What could be more honest than "10% PPS where each share is difficulty 1 and is worth x BTC".  It may not be very lucrative compared to other options but it's about as simple and clear as you can get when it comes to reward systems.

If you can understand the problem with running a 0% fee PPS pool (assuming your costs are covered) then, with some thought, you will see the similar problem with SMPPS.  If you don't see a problem with this then look up "playing the Martingale" for a foolproof way of gambling and beating the casinos.

If you don't see the relationship then:
Would you mine at the SMPPS if the buffer was 500 BTC in the red?
What if it was 10`000 BTC in the red?

What would happen if people started leaving the pool en masse when the buffer is negative (thinking that if the pool recovers they'll get their due reward anyway)?  Would you continue to mine there yourself?

If SMPPS pools will always recover then how about starting a SMPPS pool with a negative buffer of 1000 BTC.  If you are right and SMPPS is sound then the buffer will eventually hit 0 and you will have made 1000 BTC out of nothing!  Go for it!

Just as with simple pyramid schemes, SMPPS might look good at first but, if it goes on long enough, a lot of miners will lose is a big way.

Yes, beware of this.  It seems many of the small pools have no plan of action for when their pool dies.  In many cases, submitting a share is gambling that the pool will last sufficiently long.  For bigger pools or PPS pools this is not an issue.

"pool" should be "poll" and deepbit offers both PPS and Prop.

Honestly though, deepbit seems to be a solid pool and is a wise choice for a backup pool.

We may end up getting a few more shares in the near future.  RFCPool.com has just closed and this small pool was PPLNS.  I doubt all of these miners are going to suddenly change to PPS+fee or Proportional.  However, I expect many will be looking specifically for PPLNS and so will go to mineco.in or simplecoin.us rather than here.

Ah well, there's always hope.

I'm sorry to hear this pool is gone.  It was pretty good on paper and, were it not for the grace period for stales, might well have been my main pool.

Smaller pools come and go but many of them are easy to join and often provide better average income than the big pools.  If you like PPLNS then mineco.in is decent.  If your more interested in the merged mining then be advised that there are still teething issues so you should look carefully at the pools recent performance before joining.  Both simplecoin.us and yourbtc.net have merged mining with unhoppable reward systems at 40 GH/s or so (simplecoin.us offers PPLNS and yourbtc.net uses DGM).

Always consider recent uptime and luck before joining a new pool.

Very kind of you to use the word "We" there.

To be sure, I do not doubt urstroyer's integrity one whit.  I only meant to highlight what is almost certainly a technical problem.  I care far more about the reputation and apparent professionalism of the pool than I do about any personal lost income.

I'm glad it appears to be fixed and hope that this pool proves itself solid in the next few weeks.

I'm glad to see things seem to be back to normal.

However, Namecoin block 25574 was remarkably unlucky (1304912 shares all at difficulty 156504.39 puts this block in the top 0.024% of unlucky blocks).  Of course, such poor luck can happen but such an extreme example occuring at the same time as the reward system was changed and the pool was having problems with stats is too much to ignore.

Have you checked for any missing Namecoin blocks?  Do you keep a log of the nonces for shares (or is such a log automatically kept by PoolServerJ)?

Damn, there goes my stategy.

Hmm... Does "There have been votes" give users information on the poll?  Presumably if there had not been any votes that could imply that people are thinking very carefully about the poll.

Perhaps it's only fair to give cryptic information such as whether or not any of the options is currently correct (not mentioning specific options of course).