figvam
Newbie
Offline
Activity: 42
Merit: 0
|
|
July 12, 2011, 09:31:24 AM Last edit: July 12, 2011, 09:46:56 AM by figvam |
|
I suggest to lower the minimum payout at least for a few days so the guys who are moving on (like me) could cash out. I'm just a bit under 0.05 BTC and it's impossible to retrieve that sum now.
*edit* thanks for taking care of that, Wuked.
|
|
|
|
Reikoku
Full Member
Offline
Activity: 140
Merit: 100
firstbits: 1kwc1p
|
|
July 12, 2011, 01:55:56 PM |
|
I'm back with 800MH, in a few weeks 2GH.
Over an infinite time period, PPLNS is not unfair to part-time miners UNLESS they are deliberately hopping at or before the expected halfway point consistently.
Sure, sometimes you will do 4 hours mining at the start of a long block and be credited for nothing, but sometimes you will do 4 hours mining at the end of a block and be credited as if you had been mining the whole block.
No that's not correct. If you join at the end, you'll get your percent of 750000 shares depending on pool hash rate. If it takes the pool 4 hours to hash 750000 shares, and you join at the last 4 hours, then you will be credited as if you were mining the whole block. If the pool takes 8 hours to generate 750000 shares, then you get half of what you'd normally get for a while block - but this is same as proportional so that is okay. But if you left early, you'd lose there too. I don't think this is really going to stop pool hoppers...instead of hopping in at the beginning, now just hop in NOT at the beginning - such as at 75% of difficulty and staying to 125%. That doesn't work. Your chance of the next hash finding a block is the same after 0 shares as it is after 10,000,000. All that means is that you'd lose the opportunity to get the gains from shorter blocks.
|
Rei | 1Kwc1pqv54jCg8jvnm3Gu1dqFQYhS34Bow Trades So Far: 7
|
|
|
HeroMiner
Newbie
Offline
Activity: 28
Merit: 0
|
|
July 12, 2011, 03:27:26 PM |
|
I rethought the PPLNS system last night and think I overlooked an important aspect. Say the pool is in the middle of a block of usual "length". When you leave at this point, you will get not credit at all for that block. But if you join the next day, the pool is in the middle of such a block again, you will get full credit for that one, even though you haven't mined since the beginning of that round. Thats exactly what Reikoku said a few posts back, I must have overlooked that too, sry . So this method shouldn't be disadvantageous for non 24/7 miners after all.
|
|
|
|
flower1024
Legendary
Offline
Activity: 1428
Merit: 1000
|
|
July 12, 2011, 03:32:23 PM |
|
it just adds an additional variance.
as i said: as this pool does not have that much hashing power (it's a shame... i really like this pool) it will took very long to even out.
it could be that i am lucky... but who knows?
|
|
|
|
OCedHrt
Member
Offline
Activity: 111
Merit: 10
|
|
July 12, 2011, 04:56:10 PM |
|
I'm back with 800MH, in a few weeks 2GH.
Over an infinite time period, PPLNS is not unfair to part-time miners UNLESS they are deliberately hopping at or before the expected halfway point consistently.
Sure, sometimes you will do 4 hours mining at the start of a long block and be credited for nothing, but sometimes you will do 4 hours mining at the end of a block and be credited as if you had been mining the whole block.
No that's not correct. If you join at the end, you'll get your percent of 750000 shares depending on pool hash rate. If it takes the pool 4 hours to hash 750000 shares, and you join at the last 4 hours, then you will be credited as if you were mining the whole block. If the pool takes 8 hours to generate 750000 shares, then you get half of what you'd normally get for a while block - but this is same as proportional so that is okay. But if you left early, you'd lose there too. I don't think this is really going to stop pool hoppers...instead of hopping in at the beginning, now just hop in NOT at the beginning - such as at 75% of difficulty and staying to 125%. That doesn't work. Your chance of the next hash finding a block is the same after 0 shares as it is after 10,000,000. All that means is that you'd lose the opportunity to get the gains from shorter blocks. You're right on that. You can jump in whenever and be either within 750000 shares of solving the block or not getting any credit at all. With PPLNS, when you join the hashing, when averaged out, has no impact on how much you get. So rather than basing on how many shares the pool has had since the last book, you could hop basing on the pool's current statistical luck.
|
|
|
|
zapeta
|
|
July 12, 2011, 06:25:14 PM |
|
Yeah, I have to say, I like PPLNS. Especially that it already affects the current round Your Shares 1177 (0.0842 %) Your Reward: 0.0 BTC Exactly, what a bunch of BS to change mid round. I had to have my computer off for a day, and now my shares are worthless. Thankfully my current payout is greater than your minimum, so I'll be taking that and will probably return if you institute a scoring method that isn't obviously broken.
|
|
|
|
ottodv
Newbie
Offline
Activity: 59
Merit: 0
|
|
July 12, 2011, 06:57:18 PM |
|
PPLNS is simple to understand and it appears that it can't be gamed like the old system. Especially it makes joining at the beginning of a round useless - something that was costing us permanent miners. It also seems to me that this system doesn't demotivate people from joining during a drought. I also really like the principle that two or more blocks found quickly are rewarded not just over the round shares but over all the last N shares. So while some shares get you nothing others get you double or triple - for the permanent miner this balances out.
Even if you don't mine 24/7, what you loose at one end you gain at the other end. If you mine little you're gambling as your shares may be worthless or be worth double or triple.
I'll be staying with Mineco.in.
|
|
|
|
Wuked (OP)
|
|
July 12, 2011, 08:43:33 PM |
|
Yeah, I have to say, I like PPLNS. Especially that it already affects the current round Your Shares 1177 (0.0842 %) Your Reward: 0.0 BTC Exactly, what a bunch of BS to change mid round. I had to have my computer off for a day, and now my shares are worthless. Thankfully my current payout is greater than your minimum, so I'll be taking that and will probably return if you institute a scoring method that isn't obviously broken. Please explain why the method is obviously broken.
|
|
|
|
Sukrim
Legendary
Offline
Activity: 2618
Merit: 1007
|
|
July 12, 2011, 10:57:37 PM |
|
Just make sure not to cycle out getworks too soon!
Take a look at this scenario: Current difficulty: 100, next difficulty: 400 (maximum increase = 4 fold) Last block of difficulty 100 and the first block of difficulty 400 found immediately after each other. This means if you just rotate all getworks after 1/2 difficulty out, you'd miss 3/4 of getworks for the next round.
You have to have at least (current_difficulty * 2) difficulty1 shares as a reserve to make sure you can always calculate payouts.
I just wonder if "negative hopping" could work - waiting until you are on a long round (e.g. current share count = current difficulty *1.2) and then going in. If a LOT of people follow this strategy, the added hash rate would even make it a self fulfilling prophecy (as it gets more likely that a block is found). Have you done some calculations/simulations based on this?
|
|
|
|
Wuked (OP)
|
|
July 13, 2011, 10:55:15 AM |
|
Yes of course, We are storing enough info to calculate it
|
|
|
|
Wuked (OP)
|
|
July 13, 2011, 10:49:12 PM |
|
So what features are you guys looking for that are missing? I'm all ears...
|
|
|
|
Digigami
|
|
July 13, 2011, 11:18:40 PM |
|
Ability to delete or edit/rename workers please
|
|
|
|
Wuked (OP)
|
|
July 13, 2011, 11:28:21 PM |
|
Ability to delete or edit/rename workers please Editing the name should be possible, but deleting them is a bit of a grey area, as they are linked to blocks and shares by their id
|
|
|
|
Digigami
|
|
July 13, 2011, 11:35:39 PM |
|
Yea good point.. alternatively you could allow someone to "delete" the worker through the web interface, but just set a flag in the database so it's hidden from the users view.
But, being able to edit them should work out alright just fine.
|
|
|
|
lll
Newbie
Offline
Activity: 26
Merit: 0
|
|
July 14, 2011, 09:57:10 AM |
|
Wuked, what happens? I'm in the pool 24 / 7. I see that you take a commission of 5-6% give an explanation or I will go to another pool
|
|
|
|
Wuked (OP)
|
|
July 14, 2011, 10:09:36 AM |
|
Wuked, what happens? I'm in the pool 24 / 7. I see that you take a commission of 5-6% give an explanation or I will go to another pool
I don't understand what you mean? We take no commission. How are you getting this 5-6% value ?
|
|
|
|
lll
Newbie
Offline
Activity: 26
Merit: 0
|
|
July 14, 2011, 10:18:09 AM |
|
145Ghash / 6.621Ghash = 21.90 50 BTC / 21.90 = 2.2831 BTC and projected 2.1667 BTC verified in the last three blocks - the loss of 5-6% of the predicted
|
|
|
|
Wuked (OP)
|
|
July 14, 2011, 11:20:00 AM |
|
145Ghash / 6.621Ghash = 21.90 50 BTC / 21.90 = 2.2831 BTC and projected 2.1667 BTC verified in the last three blocks - the loss of 5-6% of the predicted
The 145Ghash is an estimate of the current pool speed over the last 30 minutes, As you can see from the graph the speed fluctuates somewhat. Reward isn't calculated from the estimated speeds, but in fact from valid shares Your reward is calculated like this : Total Shares = Valid Shares in last (Difficulty / 2) Shares Your Shares = Your Valid Shares in last (Difficulty / 2) Shares Your reward = your % of your shares in total shares. You cannot make a simple calculation like 145Ghash / 6.621Ghash to get your exact reward. The projected reward is the most accurate. Hope that clears things up a bit.
|
|
|
|
Reikoku
Full Member
Offline
Activity: 140
Merit: 100
firstbits: 1kwc1p
|
|
July 14, 2011, 11:59:33 AM |
|
145Ghash / 6.621Ghash = 21.90 50 BTC / 21.90 = 2.2831 BTC and projected 2.1667 BTC verified in the last three blocks - the loss of 5-6% of the predicted
Unless the pool is at a constant 145 GH, you can't use such a simple algorithm to calculate what you expect.
|
Rei | 1Kwc1pqv54jCg8jvnm3Gu1dqFQYhS34Bow Trades So Far: 7
|
|
|
lll
Newbie
Offline
Activity: 26
Merit: 0
|
|
July 14, 2011, 12:11:22 PM Last edit: July 14, 2011, 02:09:31 PM by lll |
|
I was doing stats for 3 days !!!!! I came out, that gets paid 94-95% of wages. About BTC that must be shared between participants in a pool of 7 / 24, which comes from part-time participation in a round do not even want to raise the topic .... I think you should stop cheating and really show% commission, which you take each and decide he should take part in your pool or not. I look forward to one day return to 5% for 3 days. The trust is lost. This occurs in 3 days.
|
|
|
|
|