Prohashing Mining Pool Forums

Last Monday, you may recall that we stated that we would be increasing scrypt profitability by 4% for one week, and then performing an evaluation to determine impacts on profitability, hashrate, and customer satisfaction. The research has now been completed.

After all the numbers had time to adjust, on Sunday I performed a test of a 15 MH/s Zeusminer at litecoinpool and here. At Litecoinpool, the miner earned 0.0001366473225216 LTC. On Monday morning, the same miner here earned an adjusted time amount of 0.0001337544731149189186 LTC. The amount was adjusted because it's impossible to get it to run for exactly the same length of time on both pools. The ratio of earnings here to earnings at litecoinpool was 0.9788298. This number significantly improves upon CSZiggy's records from earlier, where he was reporting much lower earnings with the previous rate. To me, it seems like earning 98% of what litecoinpool earns is very competitive against them, given that they don't offer payouts in other coins, they don't provide the same level of customer service, they are anonymous, and they are not following all regulations, to our knowledge.

I then compared hashrates over the past few weeks. One month ago, litecoinpool had a hashrate of 57 TH/s, and declined today to 33TH/s, or 57.8% their previous hashrate. The litecoin network as a whole declined in hashrate from 313TH/s to 230TH/s, or 73.4% of previous hashrate. Prohashing increased from 5.397TH/s to a current rate of 5.617TH/s, or 104.1% of previous hashrate. Prohashing had actually declined to 5.4TH/s in the interim but rose again, possibly indicating that the profitability changes had a positive effect on hashrate. However, a 31% total difference from the LTC hashrate is insufficient to fund the fee reduction, given how low margins are.

I then ran a query into our luck and payouts. From what I can tell, we had a good week, finding 1.222 of expected value for litecoins. According to the query, we paid 0.891 of expected value. That's surprising and needs investigation. I think it is caused by a bug in how Digibytes are being calculated.

Until we can determine the cause of the bug, I decided to manually increase scrypt profitability by another 3%, for a total of 7%. This change will allow me to determine whether making this pool outright more profitable than litecoinpool will double hashrate, which is what we need for the change to be profitable and sustainable.

The changes are effective immediately. Enjoy the extra money! I'll perform another evaluation next Monday.

the pps rate of scrypt was so bad at one point i'm now solo mining nyc with my scrypt machines, seemed to make slightly more on average, plus i think nyc has potential

I think I'll turn on my other 3 L3+ scrypt miners and enjoy the increase

Ran another test with the upgraded payout percentages, got new values, very happy to report PH is paying what LCP paid out.


.03953 / .03999 = 99% so PH and LCP are just about dead even for payouts, but PH offers ability to also select other coins for payout method adding higher value. As long as the payout percentages stay like they are then choosing a pool would just come down to personal tastes/preferences.
If after a week the 3% or 4% increases goes away then the payout differences would reflect that going forward.

Very well done, there was a large difference between the two pools for payout and PH took the steps needed to close that gap.
I'll be interested to see if the increased payout has any effect or if they get discontinued in a week or two.

Steve I ran the test on its own account, see ticket: UST-4987 for login info.
You can check the data recorded for that account, it only exists for the 24-hour test run I ran.
In 24-hours I had 97 rejected. If you can find anything close to 97(changes from 1 coin to anther, blocks found, stale blocks found) it could be the pool reports stales and the miner records it as rejected, or could be another reason 3% of accepted was rejected. On litecoinpool with no switching I get 4-6 rejected for 2-3 day runs.