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

[Shaban]

Sorry, I'm not going to go through all the pages, but can someone tell me if this pool uses VARDIFF or if it's set at 512?

It's set at 512 to reduce server load. Unfortunately with the crazy growth experienced recently there are still some server overloads/disconnects - hence the need for EU Server

So after the EU server, would it be lowered? Setting it at 512 would cause less shares going through, wouldn't it?

I believe somewhere in the 100+ pages there is a discussion about that and the higher difficulty does not ultimately affect the pool profitability.

But it causes less shares to go through. For example: I get 3,000+ kh/s, and when I monitor it at say 128 diff, I'm constantly sending accepted shares, when it's 512, I'll see it accept a share, but sometimes it takes too long that we already move on to the next block, so my hashing power during that time gets voided.

I hope I'm making sense.

edit: FWIW, I normally get over 3,000kh/s, but on this pool the most I've seen is 2,500 (i'm assuming this is because of the difficulty being at 512)

[envy2010]

But it causes less shares to go through. For example: I get 3,000+ kh/s, and when I monitor it at say 128 diff, I'm constantly sending accepted shares, when it's 512, I'll see it accept a share, but sometimes it takes too long that we already move on to the next block, so my hashing power during that time gets voided.

I hope I'm making sense.

edit: FWIW, I normally get over 3,000kh/s, but on this pool the most I've seen is 2,500 (i'm assuming this is because of the difficulty being at 512)

It increases your short-term variability: sometimes you have fewer shares submitted, sometimes more. However, it doesn't effect overall results. Over many days of mining it evens out.

If you keep mining you will see hashrates well above and well below your nominal value. There are no lost shares and no long-term opportunity cost associated with 512 difficulty. In fact, because it improves server reliability, the higher diff IMPROVES your average profitability.

[Shaban]

But it causes less shares to go through. For example: I get 3,000+ kh/s, and when I monitor it at say 128 diff, I'm constantly sending accepted shares, when it's 512, I'll see it accept a share, but sometimes it takes too long that we already move on to the next block, so my hashing power during that time gets voided.

I hope I'm making sense.

edit: FWIW, I normally get over 3,000kh/s, but on this pool the most I've seen is 2,500 (i'm assuming this is because of the difficulty being at 512)

It increases your short-term variability: sometimes you have fewer shares submitted, sometimes more. However, it doesn't effect overall results. Over many days of mining it evens out.

If you keep mining you will see hashrates well above and well below your nominal value. There are no lost shares and no long-term opportunity cost associated with 512 difficulty. In fact, because it improves server reliability, the higher diff IMPROVES your average profitability.

Thanks for clearing that up! I know sometimes I'll see it more and sometimes less, just depends on how many shares get sent to the pool. When I was on hashco.ws, I would see 4,700+ sometimes, but other times I saw as low as 1000

[Doom4535]

The pool may be getting too big for its own good. H2O really needs to split and point to different coins soon.

Without reading all 100+ page, any chance you could tell me in short why the pool becoming bigger is a bad thing?

More miners means more hashrate meaning:
-more coin swapping as difficulty adjustments hit harder
-hurts dedicated miners of the coin when we leave them with high difficulty
-more miners here likely means more everywhere, meaning lower profits across the board

If we split the miners across two or more coins this could mean greater stability in the difficulty of individual coins, and possibly better long term profits. We would have to have some kind of method of determining who gets to mine the best coins, and who mines the subpar coins though. Maybe round-robin style of approach, or maybe just based on raw hashrate (aka you earn h20 more fees so you get the best work to hash)

I'm thinking the fairest way to split it would be to run an optimization algorithim that determines the best hashrate for say the top five coins, and then splits the pool up to roughly achieve this number.  Payouts would be distributed amongst the miners as they are now, each person get paid based on the percentage of the total pools hashrate that they represent.

[Spencer]

Yes, I've noticed that with the high difficulty, my hashrate jumps between 220 (when I get a lot of rejects) and 410+ kh/s on a 380 kh/s setup, from what I remember.  Most of time time it is reporting near or slightly above the actual hashrate, I believe.  

Over the course of a day or 2 it definitely evens out, so the high difficulty isn't an issue I don't think.  

Also, that idea about the algorithm sounds like a good idea in theory at least.  I think that we could perhaps mine the 2-3 most profitable coins and then split the earnings between everyone, if that's possible to do, if we grow huge(r) and start having issues with wrecking markets even more so than we probably do.

EDIT: At the current exchange rate, I got $10.83 in BTC last payout.  That's awesome!  Also I noticed that Multipool.us has a Dogecoin pool now.  I have a feeling that, if Doges continue to be insanely profitable, they're going to get destroyed by pools like ours and multipool's.  Kind of sad because I think the whole idea of it is hilarious.

[Doom4535]

I just put up a *very beta* european server. Connect to eu.middlecoin.com:3333. It's in Ireland.

Tell me if it works, or doesn't. If you mine on it, make sure your hashrate on the webpage doesn't drop to zero, or do anything funky.

The main server can't fully handle the new hash power, and is shutting down for 30 seconds every hour or so. So I'm hoping this will alleviate that.

I just switched over to the new pool, I noticed it's not stratum?  My rejects will probably be higher on this server (different continent); if so, I'm going to switch back to the US one tomorrow.  It is nice having an extra server that I can set it as my failover now, really appreciate your hard work h20.

Here's my address if anyone wants to watch it (I'll probably switch back to the US server at about 7ish EST): 1EGww9a4VMLGEBZdmAKvBptgWeLMRNs9zQ

[phrozenspite]

such an algorithm would have to factor in miners dropping out, or sudden spikes of hash from other pool members joining.  I don't see it as really viable.  It helps the chains and possibly the markets, but if that was our goal, then we wouldn't be on a profitability pool.  

we haven't had any issues exchanging all of the doge that we've made so far, and having 5 different coins going at once wouldn't fix load issues on the server

[Doom4535]

such an algorithm would have to factor in miners dropping out, or sudden spikes of hash from other pool members joining.  I don't see it as really viable.  It helps the chains and possibly the markets, but if that was our goal, then we wouldn't be on a profitability pool.  

we haven't had any issues exchanging all of the doge that we've made so far, and having 5 different coins going at once wouldn't fix load issues on the server

You're right about the load issues, they'd probably actually go up because of the additional calculations.   The main benefit would be to allow us (or rather h20) to trade more subtly by using smaller volumes of a larger variety of coins (we'd be harder to predict and estimating the value of our coins would be easier because they'd be less likely to crash while cashing out).  The changing hashrate from people joining and leaving the pool could be accounted for by only using the algorithm for an approximation, maybe re-adjusting everything once every 24 hours and adding new miners to the coins that are below their target hashrate.

[envy2010]

You're right about the load issues, they'd probably actually go up because of the additional calculations.   The main benefit would be to allow us (or rather h20) to trade more subtly by using smaller volumes of a larger variety of coins (we'd be harder to predict and estimating the value of our coins would be easier because they'd be less likely to crash while cashing out).  The changing hashrate from people joining and leaving the pool could be accounted for by only using the algorithm for an approximation, maybe re-adjusting everything once every 24 hours and adding new miners to the coins that are below their target hashrate.

On normal coins, especially liquid ones e.g. LTC, I don't think trade volume is that big of a problem. h20 trades on at least 5 exchanges, and auto-trading takes place all day, not in one giant dump at the end of the day like DOGE.

Splitting part or most of the pool off DOGE when it's proven (at least for the time being) to be immensely profitable doesn't seem like a good idea either.

[ratty]

A "high risk" pool sounds fun. Then we can choose how much hash power to put on the more stable coins vs the higher risk coins.
Will we start labeling coins with AAAA in terms of price stability?

[envy2010]

A "high risk" pool sounds fun. Then we can choose how much hash power to put on the more stable coins vs the higher risk coins.
Will we start labeling coins with AAAA in terms of price stability?

Isn't that what Middlecoin is already though?

Also, 60 BTC just went from unexchanged to exchanged. 3rd straight record payout coming up.

[juggs]

I just put up a *very beta* european server. Connect to eu.middlecoin.com:3333. It's in Ireland.

Tell me if it works, or doesn't. If you mine on it, make sure your hashrate on the webpage doesn't drop to zero, or do anything funky.

The main server can't fully handle the new hash power, and is shutting down for 30 seconds every hour or so. So I'm hoping this will alleviate that.

I just switched over to the new pool, I noticed it's not stratum?  My rejects will probably be higher on this server (different continent); if so, I'm going to switch back to the US one tomorrow.  It is nice having an extra server that I can set it as my failover now, really appreciate your hard work h20.

Here's my address if anyone wants to watch it (I'll probably switch back to the US server at about 7ish EST): 1EGww9a4VMLGEBZdmAKvBptgWeLMRNs9zQ

What makes you think it is not stratum?

Looks and smells like stratum from here....

Code:

[2013-12-18 02:48:20] Starting Stratum on stratum+tcp://eu.middlecoin.com:3333
[2013-12-18 02:48:20] 1 miner threads started, using 'scrypt' algorithm.
[2013-12-18 02:48:23] Stratum detected new block

[Doom4535]

I just put up a *very beta* european server. Connect to eu.middlecoin.com:3333. It's in Ireland.

Tell me if it works, or doesn't. If you mine on it, make sure your hashrate on the webpage doesn't drop to zero, or do anything funky.

The main server can't fully handle the new hash power, and is shutting down for 30 seconds every hour or so. So I'm hoping this will alleviate that.

I just switched over to the new pool, I noticed it's not stratum?  My rejects will probably be higher on this server (different continent); if so, I'm going to switch back to the US one tomorrow.  It is nice having an extra server that I can set it as my failover now, really appreciate your hard work h20.

Here's my address if anyone wants to watch it (I'll probably switch back to the US server at about 7ish EST): 1EGww9a4VMLGEBZdmAKvBptgWeLMRNs9zQ

What makes you think it is not stratum?

Looks and smells like stratum from here....

Code:

[2013-12-18 02:48:20] Starting Stratum on stratum+tcp://eu.middlecoin.com:3333
[2013-12-18 02:48:20] 1 miner threads started, using 'scrypt' algorithm.
[2013-12-18 02:48:23] Stratum detected new block

Because when I manually add it to cgminer and enter it as stratum+tcp://middlecoin... cgminer considers it dead I have to add it as http://middlecoin... Disable the US middlecoin pool (and any other pools for that matter) and see if you continue to get the stratum updates (cgminer will check all the pools for new blocks).

*edit:  I stand corrected (and very confused).  I'm getting stratum updates, but am connected with http, any ideas?

[lorix]

It's pretty clear that H2o has a good handle on managing the manual trades to get the best result - his effort the other day offloading the DOGE without flooding the market and tanking the price was a good example.

This pool is excellent with how it converts altcoins into BTC through the day and cashes out, however there are opportunities in the longer term that makes it beneficial to hold a coin for a longer period before cashing it out, like any good managed fund.

My suggestion (unless he's doing it already) would be to maintain a "Strategic Reserve" of say 10 - 15% of each altcoin mined every day aside to be converted to BTC at H2o's discretion when a market opportunity arises. That way we won't be caught out with an empty bag when the really big swings come around.

H2o has my earned my confidence with his track record to date and I'd be more than happy to sacrifice a little short term profit for a long term jackpot under his care.

Thoughts people?

[Doom4535]

It's pretty clear that H2o has a good handle on managing the manual trades to get the best result - his effort the other day offloading the DOGE without flooding the market and tanking the price was a good example.

This pool is excellent with how it converts altcoins into BTC through the day and cashes out, however there are opportunities in the longer term that makes it beneficial to hold a coin for a longer period before cashing it out, like any good managed fund.

My suggestion (unless he's doing it already) would be to maintain a "Strategic Reserve" of say 10 - 15% of each altcoin mined every day aside to be converted to BTC at H2o's discretion when a market opportunity arises. That way we won't be caught out with an empty bag when the really big swings come around.

H2o has my earned my confidence with his track record to date and I'd be more than happy to sacrifice a little short term profit for a long term jackpot under his care.

Thoughts people?

Agreed

[h2odysee]

For today's payout, I only paid out people above >0.01 BTC. In the future, I'll make it so that you will get paid either if you have >0.01 BTC, or if you haven't been paid in 7 days and have 0.001 BTC.

[pindis]

It's pretty clear that H2o has a good handle on managing the manual trades to get the best result - his effort the other day offloading the DOGE without flooding the market and tanking the price was a good example.

This pool is excellent with how it converts altcoins into BTC through the day and cashes out, however there are opportunities in the longer term that makes it beneficial to hold a coin for a longer period before cashing it out, like any good managed fund.

My suggestion (unless he's doing it already) would be to maintain a "Strategic Reserve" of say 10 - 15% of each altcoin mined every day aside to be converted to BTC at H2o's discretion when a market opportunity arises. That way we won't be caught out with an empty bag when the really big swings come around.

H2o has my earned my confidence with his track record to date and I'd be more than happy to sacrifice a little short term profit for a long term jackpot under his care.

Thoughts people?

great idea!

[phreaker]

For today's payout, I only paid out people above >0.01 BTC. In the future, I'll make it so that you will get paid either if you have >0.01 BTC, or if you haven't been paid in 7 days and have 0.001 BTC.

Makes perfect sense to me.
Also thanks for all the hard work, I know you aren't doing it for free, but all the same I and I am sure all of us appreciate your efforts 

it has inspired me to get MH's cranking now.

[daynomate]

My reported hashing rate is 0.15Mh/s yet I was getting 480kh/s on a regular LTC pool :/ I'm sure there's nothing different here... any ideas?

nevermind - my derp.

[imafish2002]

For today's payout, I only paid out people above >0.01 BTC. In the future, I'll make it so that you will get paid either if you have >0.01 BTC, or if you haven't been paid in 7 days and have 0.001 BTC.

Seems sensible.

Make sure you change the note regarding payouts on the main page or I am sure you will be flooded with "wtf no payout scam" posts