Quote from: divxrullz on February 21, 2014, 08:49:01 AM

Quote from: igroock on February 20, 2014, 10:30:38 PM

Quote from: byt411 on February 20, 2014, 10:05:33 PM

Quote from: igroock on February 20, 2014, 09:44:44 PM

Can someone explain to me how CleverMining is getting so many rejects considering its hash power has gone up significantly? Its a problem with Terks code it seems. Considering Waffle had 1GH last week and there was amost less than 1% of rejects. I am tired of seeing the phrase its normal, its not.

I am tired of seeing people like you criticizing something you don't even understand!
IT IS NORMAL. IT HAS NOTHING TO DO WITH HASHPOWER. It switches coins quickly, and mines very quick coins, so a higher reject rate IS PERFECTLY NORMAL!

What are you retarded? Compared to every other pool out there, NO ONE IS GETTING THAT REJECT %.

I want to know whether Terks profitability % is based on hashes WITH rejection or % of hash before the rejection.

Feel free to change the pool or stop blaming the admin Wink

Many times he mention that he's working to fix those reject %

Compared to other pools - clevermining has only one diff - 512 regardless of your speed or diff of mined coin. Waffle has varrdiff - IMO that is the reason for "high" reject %
simple math:
- clever - for 1 min you are accepting 20 shares with diff 512. If 1 of them is rejected, this is a 5% (due to switching and etc)
- Waffle - for 1 min you are accepting 200 shares with vardiff 8-512 (often 32-128 due to mining coins with low diff). If 1 of them is rejected, this is a 0.5%

p.s. This is my opinion, i dont want to engage with it no one, and high profit confirms this

Sigh. Difficulty has nothing to do with reject %. It's not simple math, it's faulty logic. If only 1 out of 200 shares is rejected, that makes the probability 0.5%. If you increase the difficulty and only submit 1/10th of the shares in the same period of time, that means it's 10 times less likely that one of them is submitted at the exact moment when the coin switch/stale block occurs. You'd have to go through the same scenario 10 times to randomly get that 1 reject. The percentage doesn't change.

It's like shooting bullets in random directions and hoping that one hits the bullseye. 1/10th of the bullets = 1/10th of the chance to hit. The only thing that matters is the size of the bullseye, i.e. the combination of latency to the server, intensity setting and block speed.

And just for you I'm currently mining on Wafflepool with d=1024, and guess what? Same reject rate as with d=256.