Author

Topic: [1200 TH] EMC: 0 Fee DGM. Anonymous PPS. US & EU servers. No Registration! - page 211. (Read 499491 times)

donator
Activity: 2058
Merit: 1007
Poor impulse control.
The block only had 4.8 million shares, are you sure you're looking at the right numbers?


here's some copypasta for you. I do hope I'm wrong, but anyway:

21   136839   2011-07-18 04:53:08   14:46:41         673019       47/120    3293    0.24342476    0.00122324
20   136732   2011-07-17 14:06:04   1 d 20:39:07   1930164   Confirmed    1758    0.04531329    0.00022771
19   136430   2011-07-15 17:26:54   3 d 21:30:06   4984469   Confirmed    9003    0.12257784    0.00123816
legendary
Activity: 1260
Merit: 1000
The block only had 4.8 million shares, are you sure you're looking at the right numbers?
legendary
Activity: 1260
Merit: 1000
We should have been on scoring since block 19.  I'm not sure what accounts for what you are seeing - unless you submitted your shares at/near the end of the round as opposed to the start.  

I have on idea that might account for it, I will have to look into it and it would only effect the blocks in question because of the script errors.  It may not have been resetting the score properly (I should have checked that) at round start.  If that's not it, I will dig into it further.  Please feel free to keep up hopping and testing it out.

*EDIT*

Never mind, the scoring reset properly so that wouldn't account for it.

I think we need to conduct some more testing on this.  Can you guys continue to report on your hopping?  I will keep looking at this end.  But as already has been stated, the hopping algorithm only affects you if you are hopping from the round after the start, not near the end.
donator
Activity: 2058
Merit: 1007
Poor impulse control.
If that's the case, let this be a case in point for those of you who don't care about hopper proofing - any extra money a hopper earns by pool hopping comes out of *your* pocket.

That's interesting, but you didn't tell us during which part of the block you were hopping. Meni's algorithm rewards miners who were present in the past 5-10 minutes *when* the block is actually found. So it doesn't matter how much you were hopping before, if you're not present when the block is found your reward will be significantly (in math language, exponentially) reduced.

i was hopping from some time after the start of the block until a max of 688000 total shares into the block. But i don't think thats actually how the algo works.
member
Activity: 64
Merit: 10
If that's the case, let this be a case in point for those of you who don't care about hopper proofing - any extra money a hopper earns by pool hopping comes out of *your* pocket.

That's interesting, but you didn't tell us during which part of the block you were hopping. Meni's algorithm rewards miners who were present in the past 5-10 minutes *when* the block is actually found. So it doesn't matter how much you were hopping before, if you're not present when the block is found your reward will be significantly (in math language, exponentially) reduced.
full member
Activity: 182
Merit: 100
Just looks like there's a lot of hoppers?
donator
Activity: 2058
Merit: 1007
Poor impulse control.
I moved away after about 700000. Inaba, if it's the scoring algo let me know and I'll refund some coinage. If we were still on proportional, well maybe not  Wink
member
Activity: 101
Merit: 10
I wanted to stay hopping here to see how your algo went. I hopped for blocks 19 and 20, one monster and one longer than average block, to see how it would affect my earning. If all went ok, it should significantly reduce my earnings (see what I do for you guys? the sacrifices i make? Wink  )

Block 19: My shares: 9003
Total shares for pool: 4984469

expected earning: 9003*50/4984469 = 0.0903
actual earning: 0.12257784

I also moved off the pool after about 2.5 million shares or so during that long block, and yet I ended up with a higher payment than expected.  I suspect something with the scoring algorithm is not working as intended.
newbie
Activity: 51
Merit: 0
Yay I'm allowed to post, this pool rocks thanks Inaba. One thing I don't seem to be able to get working right is the SMS feature, I get the messages but they are all blank any ideas?
sr. member
Activity: 444
Merit: 254

The top miners info is off by 1 decimal point. I'm hashing at 24ghs..LOL
hero member
Activity: 798
Merit: 1000
Did we solve another one (136839?).   Not sure I trust the site right now Cheesy (and I didn't get an SMS, but I expect it's just delayed)
donator
Activity: 2058
Merit: 1007
Poor impulse control.
I wanted to stay hopping here to see how your algo went. I hopped for blocks 19 and 20, one monster and one longer than average block, to see how it would affect my earning. If all went ok, it should significantly reduce my earnings (see what I do for you guys? the sacrifices i make? Wink  )

Block 19: My shares: 9003
Total shares for pool: 4984469

expected earning: 9003*50/4984469 = 0.0903
actual earning: 0.12257784

Block 20: My shares: 1758
Total shares for pool: 1930164

expected earning: 1758*50/1930164= 0.0455
actual earning: 0.0453

In the second case I'm down a little, probably because the total shares were only a little over difficulty edit: no, it's because of donation. I'm at a loss to explain the first case though - unless scoring wasn't turned on yet? If that's the case, let this be a case in point for those of you who don't care about hopper proofing - any extra money a hopper earns by pool hopping comes out of *your* pocket.



sr. member
Activity: 444
Merit: 254

Nice! I'm only getting < 0.1% stales. I never knew that. Cheesy
legendary
Activity: 1260
Merit: 1000
I've added an experimental graph under the Block Stats page (https://eclipsemc.com/block_stats.php). Click on your shares for a given block to view the graph of our miners.

Let me know if there are any bugs in the display!
hero member
Activity: 798
Merit: 1000
Yay, I can finally post in this thread Smiley

Er....  that's all I have to say right now Wink
donator
Activity: 2058
Merit: 1007
Poor impulse control.
Nope.. that's only first a fee calc how much loss one does once we hit over the diff share rate.. next calc is how much Hash Calc each share needs. In my ex: it's with 345MH's

My avg Share calc was each 14.5sec and that would make it 5000 MegaHashes In avg to Calc one Share..

You Try.. 160740 sec / #Blk20 Your Sent in Shares = Avg time sec per calc share. Then you put in your avg Rig HashRate  (Mine ~345MH)* the time sec result.  Should be around 5000MH (I guess that is if you run your rig 24/7)

Can we add an avg hashrate per miner for each block??

[Edit] @organofcorti: Great explanation, that was exactly what I was looking for.. If this info could be available in the pools mining area that'd be great. Perfect mining in theory is one thing.. We get all the random factors that pushes this back a bit.

4294967296 ~= 4300MH / 14.5 = ~300MH Per Sec, So my avg Hashing speed for BLK#20 was 300MH if my calc is correct. Still my mining program says it's about ~345MH. Can it be the Rejected %??

Assuming that your 'luck' was 1.0 for the entire block (by no means guaranteed), and that you contributed shares for the entire block at the same rate (no hopping), and that your time calcs are correct (didn't check) you can back calculate the amount of time eclipsemc was down:
 
(5000-4300)/5000* 160740 = 6.25 hours.

You can check at: http://mksv.ath.cx/bit/stats.php

Which unfortunately is down right now.
newbie
Activity: 31
Merit: 0
4294967296 ~= 4300MH / 14.5 = ~300MH Per Sec, So my avg Hashing speed for BLK#20 was 300MH if my calc is correct. Still my mining program says it's about ~345MH. Can it be the Rejected %??

You need to remember that there was a few hours of downtime for block # 20 so would explain why your average is less.
legendary
Activity: 1260
Merit: 1000
Yes, I plan on adding average hash rate per miner as a graph or output in some fashion.  It's definitely on the list, I just haven't had a chance to get around to it just yet.

The programs calculation of hashrate is based off of it's internal hashing mechanism, all pools hashrates (not just this one) is based off a mathematical formula: numshares * 2^32 / timeframe

So in all likelyhood what you see in your program is always going to be higher than your calculated hashrate as an average, since the hashrate displayed in your program is the best possible hashrate as an absolute value and any factors that affect that will lower your calculated hashrate.
member
Activity: 83
Merit: 10
Nope.. that's only first a fee calc how much loss one does once we hit over the diff share rate.. next calc is how much Hash Calc each share needs. In my ex: it's with 345MH's

My avg Share calc was each 14.5sec and that would make it 5000 MegaHashes In avg to Calc one Share..

You Try.. 160740 sec / #Blk20 Your Sent in Shares = Avg time sec per calc share. Then you put in your avg Rig HashRate  (Mine ~345MH)* the time sec result.  Should be around 5000MH (I guess that is if you run your rig 24/7)

Can we add an avg hashrate per miner for each block??

[Edit] @organofcorti: Great explanation, that was exactly what I was looking for.. If this info could be available in the pools mining area that'd be great. Perfect mining in theory is one thing.. We get all the random factors that pushes this back a bit.

4294967296 ~= 4300MH / 14.5 = ~300MH Per Sec, So my avg Hashing speed for BLK#20 was 300MH if my calc is correct. Still my mining program says it's about ~345MH. Can it be the Rejected %??
donator
Activity: 2058
Merit: 1007
Poor impulse control.
How high MHash does one need to make better profit then PPS?

Est. Reward: 0.28700100 / PPS 0.00002877132165184 (10% Fee) = Shares submitted 9975 Before it's lower then PPS value, My Valid shares 11079 Value Per Share ~0.00002590132165184 (~18.97% Fee)

1 d 20:39:07 = 44h 39min = 160740 sec / Shares (11079) = 14.5s per share * ~345MH = ~5000MH to send one Share..

Does anyone feel like comparing his numbers with mine? How many MH per share do you need?

Just having a math crunch munch.. ^_^

The probable number of hashes per share (which is what I assume you mean, from your working out) is proportional to difficulty.

P(hashes to win) = difficulty*(2^32)

at difficulty 1 (which most pools run on) = 2^32 = 4294967296 hashes per share.

This is regardless of whether it's proportional or pps.

However if you want to know how many extra hashes you need to make up for the fact it's pps (assuming pps tax is 10%) that will be simply:

2^32*(1/0.9)

= 4772185884.4 hashes = approx 4800 * 10^6 hashes or 4800 Megahashes.

Note that the result relies on nothing other than the pool's difficulty setting and and the PPS tax. Any time you do the figuring out you'll always get around 4800 MH.

NB: some confusion arose previously because you used MH to mean both 'megahashes per second' and 'megahashes'.
Jump to: