Topic: [ANN][Pool][Profit-Switch][Optional Auto-Exchange per Coin][Vardiff] ~ Hashcows - page 170. (Read 347332 times)

I'm actually using load-balance with an LTC pool.

So I know we don't mine LTC due to high network difficulty and very low probability of finding a block, but would it be possible to mine it on p2pool? I have no idea how hard this would be to implement but i haven't seen this suggestion or any explanation as to why it wouldn't work posted anywhere yet.

And make sure you got Failover enabled... I failover onto BTB and hit 14 BLocks yesterday for the 30 seconds it drops out.

Nah, this pool is great. I was dealing with the quirks of multiple coins manually. That was an headache.

That said, there's one thing miners will have to consider. When switching coins, the connection is interrupted for some seconds or no work is available. You must take into consideration what those 50+ daily start/stop cicles (and so, temperature variations) do to your videocards in the long term, specially if you're using --auto-fan. If you mine here, I suggest you undervolt, don't push them too far or take a serious look at cooling

Thanks for the reply nearmiss.  Most of what you said I kind of figured and I understand.  I should know that 24hrs is a terrible test window.  You have always been a very stand up guy and I will probably give this another shot in a few days and give it a week or so for a full test.

Totally agree.

For miners want to spend time optimising their returns then go for it and good for them.  I've tried all types of mining and now I want a good return without much effort and hashcows is brilliant for that.  Keep up the good work!

CGB is taken back off.  We had one successful round, but then later hit it again and resulted in >30 orphans.

$comment++;

I am curious how would BTB react to 300MH/s hashing... Would we be able to make some good coin on it?

Could we put it in a try?

Yesterday was a bad day but I've been here for 2 weeks and the returns have been good over all. I don't have to bother checking whether Cryptsy is actually working anymore. More than anything else I'm impressed with the efforts the operators put in to keep improving it. I'm willing to trust them to make the best choices as to how they manage which coins are on the pool and how to calculate how to switch to them. They have all the data available, unlike the people who post here about how bad one particular round of any given coin was.

Good work Hashcows, keep mooing.

With 4mhs I have been doing the same or better than LTC here over the last 12 days.

I'm running with 5MH myself and yesterday was a poor day but I still got paid 0.0759169, the previous 24 hours I was paid 0.09445155 for my miners efforts.  If you are getting below 0.07, there must be something up with your rigs or settings or connection to the pool.

nearmiss/aTriz, is is possible to get a full historic view of all payouts at all?

1) Yesterday was our second worst day on record
2) A coin that says its most profitable on external site A, but we find would NOT be if we dumped our pool earnings, who is right? We could dump before anyone mining it gets a chance, in which case it wasn't the most profitable for them by the time they got any coins to trade
3) We had 2 coins disabled (CAP and CGB), both of which are often near the top of the charts throughout the day.  Both are re-enabled at this point.  The CGB was particularly costly, as we had a few rounds worth of it orphan out on us, meaning we lost any profits there.
4) If you are basing profitability solely on coinchoose, its not always as quick on diff changes (from my personal experience, nothing against the site). If you have your own or other sources, so be it.
5) We do take stalerate/past performance into account. Some coins, while showing as 'most profitable' ended up not being so once we got the final BTC back. Some examples, GDC, ALF, KGC, EMD, DMD...all of which we have since moved on from mining because of this.
6) Sometimes a small increase in profitability difference isn't worth switching for, so we don't. Not sure if this fits any of your examples though.
7) We do dump coins rather quickly (in my opinion), and rarely are things left over a matter of days.  I agree with you you'd likely end up ahead if you just kept the coins and traded yourself, and I think you'd find that with any auto-trade option out there. This is precisely why we give miners the flexibility to do it.

@ 5.5Mh/s you should earn 5.2LTC/day = .0878 BTC.  Still not a profitable comparison for us, but not over 50% either.  As I said, it was our 2nd worst day in 2 weeks of operation.

At any rate, I'd say in summary, try us out for a longer period than just 24hrs (a particularly rough 24hrs), but that is up to you.  I appreciate you checking it out, I know you hit up my other pools often as well.  

Well after using the pool for over 24hrs and previously posting about how I like the pool sadly this will be the last time I use it.  After A full day of mining with 5.1-5.7 Mh depending on temps I ended up with 0.04883982 BTC.  Thats over 50% less than just mining LTC alone.  Why is the pool spending all its time mining coins that are not anywhere near the highest profitability?  Several times that I checked we were mining coins less profitably than LTC which as we all know is nowhere near the top of the list.  If market depth on coins being held in your exchange hot wallet are preventing the pool from mining coins that are truly more profitable than this pool needs to work out its strategies.  The profitability switcher needs some serious work.  The auto trade feature is nice but if I just took the actually alt coins and traded them myself right away I would have made more BTC.   I think you guys need to dump the coins asap for whatever you can get for them instead of hoping to maximize the trade value over a few days since it doesn't look like the miners here see the benefit of you trying to sell high.

I've re-enabled both CGB and CAP.  CGB with updated addnodes as per dev, and CAP is running the 1.5.2 client.  Will keep an eye on how they go.

Yes, we add a penalty on top of the final profitability number to account for stales.   (anyone seen fastcoin mined in recent memory? ).  I've found with a rare few exceptions we only hit dgc when its diff drops, at which point it is in the range of 10-14 profitability, and more than worth the reject rate!

We are looking into that round, we switched to IFC three times last night, two of the times it was for less than 10 mintues and the payout's were great. All these LTC coins we are adding will be evaluated and be dealt with accordingly.

Seems that stratum was down during that IFC round, we will implement a fix for this today.

Also remember http://coinchoose.com/ is a wonderful site and is really great, but we use our own algorithms. Sometimes sal002 is extremely busy and the info is not always 100% updated/accurate, I can assure you IFC is not at "Profitability after adjusting for stales: 2.63%"

http://coinchoose.com/
Profitability after adjusting for stales: 2.63%

It's probably not an estimation error...
The question is: Why is the pool mining this coin?

Does the pool take into consideration stale rates when calculating profitability? DGC: 23479 Accepted, 4917 Rejected. That's 20.9% hashes wasted.
If it already does, then it's awesome

Keep in mind CAP too was disabled due to its issues the other night.  We'll hopefully have it back up this morning if its looking ok again.   The script checks fairly frequently, and will switch coins if the difference is significant, even after a short period, such as the 6 minute ANC round we just had.   Switching has its own implied cost of lost hashpower.

Also note, that the pool takes things like market depth, and its own outstanding balances into account when deciding a coins price.   ie: If the price of CAP is listed as .0014, but the pool has 15,000 CAP outstanding to trade, it will factor that 15k into its price calculations, which could affect its profitability number vs. other external sources.

Perhaps listing the profitability of each coin as the script updates would be useful (though not quite sure where to throw it, its a good sized list).