[ANN] profit switching auto-exchanging pool - www.middlecoin.com

[HumbleMiner]

So, help me understand what happens with "inactive" accounts. I've decided to switch from one wallet to another on 01/08, got last payment to my old wallet on that date and now it shows 0.00598752 balance. If I understand it correctly today is the day this account was supposed to get paid the remainder of the balance (7 days not being paid + today is Wednesday), right? The daily payout went through for all accounts with > 0.01 balance, but I haven't received mine. Am I missing something obvious? I know, it's about $5, but still ...

The chart (http://www.middlecoin.com/reports/1vgf8yeEpr3BtqiFzjMRjaD1eS8DD6Vsj.html) doesn't reflect the latest balance, but the list does.

[noegzit]

Hey all. I've been on this pool a long time, haven't compared numbers in quite a while.

However the recent "let's talk about payout" caught my interest, as people reporting .01/mh or even .015/mh are way outperforming me. Like, up to 2x, that is considerable and a little worrying. I'm averaging something like 0.0075/mh.

So I poked around a little.

Here's two different users, for example:

This guy, let's call him Bob, has a steady 19 mh/s or so, and has been around since Dec 16th-
http://www.middlecoin.com/reports/1ACruk5rrJmGTkP7UyyoqGDLNZTHzkqviY.html

And this dude, Let's call him Frank, has slightly less, maybe 18 average, and just started mining on the 7th-
http://www.middlecoin.com/reports/1FxziX77LiZLZxhdXMWREMQVMyjbDmCubw.html

So these are roughly comparable hashrates, with comparable reject ratios, but Frank (the newer, slower miner) is absolutely trouncing Bob on the payouts.

WTF is going on here?
(And yes, I know all the usual explanations, but I don't think any of them actually apply to this specific example)

If I'd have to guess, I'd say they are mining on different servers. Bob is probably mining on the uswest server and Frank on one of the beta servers. uswest being the biggest and having the most hashrate means it can't mine the smaller coins. The beta servers which are smaller, don't have this problem and can mine the smaller coins without problem.

This is just me guessing though that the servers work independently from each other and are basically their own pools.

Could be.

[dgross0818]

Hey all. I've been on this pool a long time, haven't compared numbers in quite a while.

However the recent "let's talk about payout" caught my interest, as people reporting .01/mh or even .015/mh are way outperforming me. Like, up to 2x, that is considerable and a little worrying. I'm averaging something like 0.0075/mh.

So I poked around a little.

Here's two different users, for example:

This guy, let's call him Bob, has a steady 19 mh/s or so, and has been around since Dec 16th-
http://www.middlecoin.com/reports/1ACruk5rrJmGTkP7UyyoqGDLNZTHzkqviY.html

And this dude, Let's call him Frank, has slightly less, maybe 18 average, and just started mining on the 7th-
http://www.middlecoin.com/reports/1FxziX77LiZLZxhdXMWREMQVMyjbDmCubw.html

So these are roughly comparable hashrates, with comparable reject ratios, but Frank (the newer, slower miner) is absolutely trouncing Bob on the payouts.

WTF is going on here?
(And yes, I know all the usual explanations, but I don't think any of them actually apply to this specific example)

Bob has a lot more rejected shares

Yes, however, the rejected shares shouldn't matter that much, as "Bob" still had a consistently higher accepted hashrate, yet was making less.

My payout tonight was very very close to 0.01 BTC per MH/s, which puts me in line with "bob"

I've also been around a while, and I'm on the main server, not one of the newer "beta" ones

I think the best explanation is the one already proposed - the smaller servers can more effectively "coin hop" for increased profitability while the older "main" server can pretty much only do LTC, DOGE, and a few others

Of course, others will say that H2O starts siphoning payout for himself after you've been around a while :-P

In order to test the server theory, you should switch all miners to a different address for a day (on same server), then add up the total made in 24 hrs (exchanged + unexchanged + immature) and compare it to what another person (who has been on the server for a while) with a very similar hashrate / rejects made during that same 24 hr period (remember to check the difference in the unexchanged balance as well as the payout)

As long as there's no "leeching" going on, the payouts should be identical...

Now take your miners and point them at one of the beta servers for 24 hours and again record what you made... compare this to the payout of the equal hashpower address and see if they still match.

[RickJamesBTC]

I know this will annoy some people. But to finally put to bed the rumour that MiMiMiner is running some sparkly fandangled ASIC miner....   read his post in this thread here:

https://bitcointalk.org/index.php?topic=355268.msg4371557#msg4371557

Where he says:
Hi,

Can somebody please PM me that has contacts with these people. We are very interested. Currently we have a pilot running with 80 GPU servers. We want to expand bigtime.

Thanks.


Now how one goes about getting over 900 MH/s from just 80 servers is another question - that's > 11MH/s per server.

Judging from here: https://litecoin.info/Mining_hardware_comparison#AMD_.28ATI.29 the best achieved is around 1.5MH/s per watercooled 7990, so with 80 servers would require 7 or 8 cards per server.

There are motherboards available with 7 PCI-E slots (e.g. http://www.newegg.com/Product/Product.aspx?Item=N82E16813131971) - probably more if I bothered to look harder

Quite a feat of juggling enough power supplies, dissipating heat and keeping 7 or 8 cards per board running in a stable manner.

he didn't say he was running the big server did he? I thought he was the SECOND big account.

[TazMan143]

I know this will annoy some people. But to finally put to bed the rumour that MiMiMiner is running some sparkly fandangled ASIC miner....   read his post in this thread here:

https://bitcointalk.org/index.php?topic=355268.msg4371557#msg4371557

Where he says:
Hi,

Can somebody please PM me that has contacts with these people. We are very interested. Currently we have a pilot running with 80 GPU servers. We want to expand bigtime.

Thanks.


Now how one goes about getting over 900 MH/s from just 80 servers is another question - that's > 11MH/s per server.

Judging from here: https://litecoin.info/Mining_hardware_comparison#AMD_.28ATI.29 the best achieved is around 1.5MH/s per watercooled 7990, so with 80 servers would require 7 or 8 cards per server.

There are motherboards available with 7 PCI-E slots (e.g. http://www.newegg.com/Product/Product.aspx?Item=N82E16813131971) - probably more if I bothered to look harder

Quite a feat of juggling enough power supplies, dissipating heat and keeping 7 or 8 cards per board running in a stable manner.

he didn't say he was running the big server did he? I thought he was the SECOND big account.

Yeah he is the second guy.

[Shaban]

Hey all. I've been on this pool a long time, haven't compared numbers in quite a while.

However the recent "let's talk about payout" caught my interest, as people reporting .01/mh or even .015/mh are way outperforming me. Like, up to 2x, that is considerable and a little worrying. I'm averaging something like 0.0075/mh.

So I poked around a little.

Here's two different users, for example:

This guy, let's call him Bob, has a steady 19 mh/s or so, and has been around since Dec 16th-
http://www.middlecoin.com/reports/1ACruk5rrJmGTkP7UyyoqGDLNZTHzkqviY.html

And this dude, Let's call him Frank, has slightly less, maybe 18 average, and just started mining on the 7th-
http://www.middlecoin.com/reports/1FxziX77LiZLZxhdXMWREMQVMyjbDmCubw.html

So these are roughly comparable hashrates, with comparable reject ratios, but Frank (the newer, slower miner) is absolutely trouncing Bob on the payouts.

WTF is going on here?
(And yes, I know all the usual explanations, but I don't think any of them actually apply to this specific example)

Bob has a lot more rejected shares

Yes, however, the rejected shares shouldn't matter that much, as "Bob" still had a consistently higher accepted hashrate, yet was making less.

My payout tonight was very very close to 0.01 BTC per MH/s, which puts me in line with "bob"

I've also been around a while, and I'm on the main server, not one of the newer "beta" ones

I think the best explanation is the one already proposed - the smaller servers can more effectively "coin hop" for increased profitability while the older "main" server can pretty much only do LTC, DOGE, and a few others

Of course, others will say that H2O starts siphoning payout for himself after you've been around a while :-P

In order to test the server theory, you should switch all miners to a different address for a day (on same server), then add up the total made in 24 hrs (exchanged + unexchanged + immature) and compare it to what another person (who has been on the server for a while) with a very similar hashrate / rejects made during that same 24 hr period (remember to check the difference in the unexchanged balance as well as the payout)

As long as there's no "leeching" going on, the payouts should be identical...

Now take your miners and point them at one of the beta servers for 24 hours and again record what you made... compare this to the payout of the equal hashpower address and see if they still match.

It doesn't matter what server he's on. H2o already said that all servers are mining the same coins, running the same algorithm.


Also, I STILL haven't gotten an answer to my question. I keep getting higher rejects + lower hash rates, going on 4 days now. I've tried switching servers and that doesn't help. I've been mining here for a month now, and my settings never changed. I tried PMing H2o, but no answer there, and no answer here.

[dgross0818]

I'm getting around 5-7% rejects or so on the main server - what are you getting?

My rejects are high than others probably because it goes Rig --> switch --> EOP --> switch --> modem, which unfortunately adds a bit of latency

As for the lower reported hash rate, how do your Worker Units compare to your reported hashrate?

My Accepted + Rejected is within 2% of what is should be, so unfortunately I can't help you much there.

You could always create another address and point your rig(s) to that instead and see what happens.

[gsrcrxsi]

What a complete shitty payout today.  Everyone leave this pool immediately.

My payout has been exactly the same for the past 3 days. About .01/MH, that's pretty much on par with the difficulty increases.

I don't see the issue. This isn't mid December anymore when everyone was pumping and dumping DOGE. It's back to normal.

[codon]

It'll be a long time before we say goodbye to GPU mining. Each chip gets 60 kh/s... they only made 400 (10 chips in each one) for the first batch, which already sold out. That's a total of 240,000 kh/s... that's not much.

Is it just me or would a scrypt asic not look like this at all?  It would either be a big, big die with lots of expensive as hell eDRAM and a memory controller, or go external and put some expensive GDDR5 chips on the PCB. Basically, it would wind up looking like a GPU, no. So... this is bullshit?

[gsrcrxsi]

It'll be a long time before we say goodbye to GPU mining. Each chip gets 60 kh/s... they only made 400 (10 chips in each one) for the first batch, which already sold out. That's a total of 240,000 kh/s... that's not much.

Is it just me or would a scrypt asic not look like this at all?  It would either be a big, big die with lots of expensive as hell eDRAM and a memory controller, or go external and put some expensive GDDR5 chips on the PCB. Basically, it would wind up looking like a GPU, no. So... this is bullshit?

If you read through that reddit link, it seems the scrypt/LTC USB sticks are fake or a scam anyway.

[Shaban]

I'm getting around 5-7% rejects or so on the main server - what are you getting?

My rejects are high than others probably because it goes Rig --> switch --> EOP --> switch --> modem, which unfortunately adds a bit of latency

As for the lower reported hash rate, how do your Worker Units compare to your reported hashrate?

My Accepted + Rejected is within 2% of what is should be, so unfortunately I can't help you much there.

You could always create another address and point your rig(s) to that instead and see what happens.

This is my numbers that I posted earlier:
572 shares - 152 Reject - 5.33 MH/s.

I have 2 7970's, 2 r9 280x's, and 6 7950's.

The 7950's are getting 600 Kh/s each (different rigs, different numbers)
the 7970's and 280x's are all getting 730 kh/s each.

So all my hashing power together gets me 6,520 Kh/s. Yet on middlecoin I keep going between 4.0-6.5 Mh/s. Usually I'm around 6.3-6.5. But for the last 4 days, I keep losing hashrate and getting more rejects.

[zSprawl]

What a complete shitty payout today.  Everyone leave this pool immediately.

My payout has been exactly the same for the past 3 days. About .01/MH, that's pretty much on par with the difficulty increases.

I don't see the issue. This isn't mid December anymore when everyone was pumping and dumping DOGE. It's back to normal.

hehe, lots of heavy sarcasm in this thread the last few pages :p

[juggs]

I know this will annoy some people. But to finally put to bed the rumour that MiMiMiner is running some sparkly fandangled ASIC miner....   read his post in this thread here:

https://bitcointalk.org/index.php?topic=355268.msg4371557#msg4371557

Where he says:
Hi,

Can somebody please PM me that has contacts with these people. We are very interested. Currently we have a pilot running with 80 GPU servers. We want to expand bigtime.

Thanks.


Now how one goes about getting over 900 MH/s from just 80 servers is another question - that's > 11MH/s per server.

Judging from here: https://litecoin.info/Mining_hardware_comparison#AMD_.28ATI.29 the best achieved is around 1.5MH/s per watercooled 7990, so with 80 servers would require 7 or 8 cards per server.

There are motherboards available with 7 PCI-E slots (e.g. http://www.newegg.com/Product/Product.aspx?Item=N82E16813131971) - probably more if I bothered to look harder

Quite a feat of juggling enough power supplies, dissipating heat and keeping 7 or 8 cards per board running in a stable manner.

he didn't say he was running the big server did he? I thought he was the SECOND big account.

Yeah he is the second guy.

I must have missed that - my bad. So the 2nd guy is running ~150 - 160MH/s. 2 MH/s per server given 80 servers - that's far more sane.

[willittobe]

Hey all. I've been on this pool a long time, haven't compared numbers in quite a while.

However the recent "let's talk about payout" caught my interest, as people reporting .01/mh or even .015/mh are way outperforming me. Like, up to 2x, that is considerable and a little worrying. I'm averaging something like 0.0075/mh.

So I poked around a little.

Here's two different users, for example:

This guy, let's call him Bob, has a steady 19 mh/s or so, and has been around since Dec 16th-
http://www.middlecoin.com/reports/1ACruk5rrJmGTkP7UyyoqGDLNZTHzkqviY.html

And this dude, Let's call him Frank, has slightly less, maybe 18 average, and just started mining on the 7th-
http://www.middlecoin.com/reports/1FxziX77LiZLZxhdXMWREMQVMyjbDmCubw.html

So these are roughly comparable hashrates, with comparable reject ratios, but Frank (the newer, slower miner) is absolutely trouncing Bob on the payouts.

WTF is going on here?
(And yes, I know all the usual explanations, but I don't think any of them actually apply to this specific example)

Bob has a lot more rejected shares

Yes, however, the rejected shares shouldn't matter that much, as "Bob" still had a consistently higher accepted hashrate, yet was making less.

My payout tonight was very very close to 0.01 BTC per MH/s, which puts me in line with "bob"

I've also been around a while, and I'm on the main server, not one of the newer "beta" ones

I think the best explanation is the one already proposed - the smaller servers can more effectively "coin hop" for increased profitability while the older "main" server can pretty much only do LTC, DOGE, and a few others

Of course, others will say that H2O starts siphoning payout for himself after you've been around a while :-P

In order to test the server theory, you should switch all miners to a different address for a day (on same server), then add up the total made in 24 hrs (exchanged + unexchanged + immature) and compare it to what another person (who has been on the server for a while) with a very similar hashrate / rejects made during that same 24 hr period (remember to check the difference in the unexchanged balance as well as the payout)

As long as there's no "leeching" going on, the payouts should be identical...

Now take your miners and point them at one of the beta servers for 24 hours and again record what you made... compare this to the payout of the equal hashpower address and see if they still match.

Yep, I already started a test at the beginning of the payout period: two identical rigs. One rig simply pointed to a new address with no other changes- so still on the main server. The other rig pointed to useast with yet another new address. Both will run the same 24 hour period.

Shall see what comes of it, though as someone else already mentioned, it appears that the coin switching is identical between servers, I watched it for a while.

So, I really have *no idea* what can account for this discrepancy, unless I really am being 'penalized' or 'leeched' for being an "old timer". Will be interesting to see the numbers tomorrow. I really hope there is some other explanation other than being stolen from.

[RickJamesBTC]

Hey all. I've been on this pool a long time, haven't compared numbers in quite a while.

However the recent "let's talk about payout" caught my interest, as people reporting .01/mh or even .015/mh are way outperforming me. Like, up to 2x, that is considerable and a little worrying. I'm averaging something like 0.0075/mh.

So I poked around a little.

Here's two different users, for example:

This guy, let's call him Bob, has a steady 19 mh/s or so, and has been around since Dec 16th-
http://www.middlecoin.com/reports/1ACruk5rrJmGTkP7UyyoqGDLNZTHzkqviY.html

And this dude, Let's call him Frank, has slightly less, maybe 18 average, and just started mining on the 7th-
http://www.middlecoin.com/reports/1FxziX77LiZLZxhdXMWREMQVMyjbDmCubw.html

So these are roughly comparable hashrates, with comparable reject ratios, but Frank (the newer, slower miner) is absolutely trouncing Bob on the payouts.

WTF is going on here?
(And yes, I know all the usual explanations, but I don't think any of them actually apply to this specific example)

Bob has a lot more rejected shares

Yes, however, the rejected shares shouldn't matter that much, as "Bob" still had a consistently higher accepted hashrate, yet was making less.

My payout tonight was very very close to 0.01 BTC per MH/s, which puts me in line with "bob"

I've also been around a while, and I'm on the main server, not one of the newer "beta" ones

I think the best explanation is the one already proposed - the smaller servers can more effectively "coin hop" for increased profitability while the older "main" server can pretty much only do LTC, DOGE, and a few others

Of course, others will say that H2O starts siphoning payout for himself after you've been around a while :-P

In order to test the server theory, you should switch all miners to a different address for a day (on same server), then add up the total made in 24 hrs (exchanged + unexchanged + immature) and compare it to what another person (who has been on the server for a while) with a very similar hashrate / rejects made during that same 24 hr period (remember to check the difference in the unexchanged balance as well as the payout)

As long as there's no "leeching" going on, the payouts should be identical...

Now take your miners and point them at one of the beta servers for 24 hours and again record what you made... compare this to the payout of the equal hashpower address and see if they still match.

Yep, I already started a test at the beginning of the payout period: two identical rigs. One rig simply pointed to a new address with no other changes- so still on the main server. The other rig pointed to useast with yet another new address. Both will run the same 24 hour period.

Shall see what comes of it, though as someone else already mentioned, it appears that the coin switching is identical between servers, I watched it for a while.

So, I really have *no idea* what can account for this discrepancy, unless I really am being 'penalized' or 'leeched' for being an "old timer". Will be interesting to see the numbers tomorrow. I really hope there is some other explanation other than being stolen from.

Leave them running more than a day, try a week at least. Good plan though, Can you throw some rigs onto the beta servers? I don't really have identical rigs that I can switch around, there are always differences.

[Panteraswift]

I have yet to see two absolutely identical mining rigs. The manufacturing variances between examples of individual components make it nearly imposible. They may be very close, but not quite identical.

[RickJamesBTC]

I have yet to see two absolutely identical mining rigs. The manufacturing variances between examples of individual components make it nearly imposible. They may be very close, but not quite identical.

Very close is easy, I'm just still setting up the majority of them. Very close should work for comparison testing.

[willittobe]

@RickJamesBTC

Yes will probably let them run a while, unless after 24 hours something is already becoming apparent. Yes, one of the rigs is on a beta server- useast.

@Panteraswift

As identical as humanly possible. Same parts, same hashrates. Very, very, very close.

[Panteraswift]

I have yet to see two absolutely identical mining rigs. The manufacturing variances between examples of individual components make it nearly imposible. They may be very close, but not quite identical.

Very close is easy, I'm just still setting up the majority of them. Very close should work for comparison testing.

Granted. I usually see a difference of around 1 to 5 % between otherwise identical rigs. I try to take the differences into account when comparison testing so the results make sense.

[willittobe]

I have yet to see two absolutely identical mining rigs. The manufacturing variances between examples of individual components make it nearly imposible. They may be very close, but not quite identical.

Very close is easy, I'm just still setting up the majority of them. Very close should work for comparison testing.

Granted. I usually see a difference of around 1 to 5 % difference between otherwise identical rigs. I try to take the differences into account when comparison testing so the results make sense.

On the two rigs that are being used for this test (running over 2 hours now) the average hashrate is exactly the same, down to the third decimal place.