But what i found weird at your pool, is that it shows a flat hash rate, same as reported in miner and pool speed should be calculated by shares found not by miner itself...
Well, I have to disagree. Hashrate is constant, depends on the gpu and settings only, does not depend on luck. So it's a stable metric for the miner software, when calculated in a fair way.
How is it weird to report what the miner effectively computes? Every fair miner does that.
On the pool side, found shares are what counts, and this metrics is shown on the pool, with historical graphs for each miner address.
So, I don't get the "transparent" thing. I do report both accurate hash from the miner and accurate shares from the pool.
All data is there, raw, for everyone. It's not on others.
So when in some rounds i found very few shares(low speed at pool), lots of blocks where found. And rounds i found more shares(higher speed calculated) less blocks where found.. This could be solved by 2 hour rounds to even it out better, or maybe lower pool difficulty. It is more dependent with luck at current state of diff and rounds i believe, and not a 100% fair distribution of blocks found. As this happens a lot... I might be wrong man , not to old in this game but to me it looks unfair and luck are involved at payouts for each round.. But when in some rounds 12 shares found and other up to 24 its a lot with up to 50 % different for each round in shares found, would like to hear you opinion of this.. With lower diff the % would be way less if more shares where found..
So, here you're talking luck and variance.
Yes, there is a large variance from round to round (1 round = 1 hour).
This variance is caused by 2 things :
- network variance (I added the graph of the net variance on the dashboard a few weeks ago). When netdiff varies, so do the blocks per hours for the whole net.
- your personal variance. The more hash you have, the less variance you'll experience. On average, you may be unlucky on some rounds, but then more lucky than average on others. On average, everyone luck is the same, so in the long run you can't loose every time.
I do agree that
- Lowering pool diff
- having rounds longer than 1 hour
would make the luck factor less visible.
I may lower the pool diff later on, but my priority it to have a stable and reliable pool, and I have somehow to adjust to the global net hashrate and to the average miner hash.
So yes, small miners see the luck factor more. But on average, it's the same.
I do not see a transparency issue there. The shares are what they are, I report what the miner sends, I can't influence your miner on sending less shares on a good round
.
pool shows graphs and share counts, miner has logs with found shares, all is transparent. Some others pools do not even provide that data.
I think I'm the pool providing the most data.
On the other pool, you don't even have the historical data to compare and see your luck from hour to hour...
You could also lower your fees
I could, but this is no transparency issue. Pool fees are displayed on the pool website.
They are not on the one you say is "most transparent".
I also got significant improvment on other pools due to my main hash is AMD cards which your miner isn't very good at.
Bis-pool miner may indeed be better suited for Amds and computes the hashrate in the miner the same way, so can be compared to EggPool.
The other one does not compute the hash rate in the same way, so the hashrate improvement you saw is not a real one. You were not comparing the same metric.
(Not a transparency issue neither from Eggpool: I say how the hashrate is computed - number of tested solutions per second - coinsaurus computes a different thing and does not say what nor how)
So, I do understand your concerns, mainly Amd efficiency and variance because of low hash.
I get it that another pool may give better results for your case (but do your tests again, I'm 100% sure it's not the one you believe it is) and that's ok with me.
Honestly, I just don't get the lack of transparency you hold against EggPool , and why you say the other pool, the most opaque one, is the "most transparent".
. 
It seems the node sometimes happens to report about some error in line 2436 though I'm not familliar with dat python thang to see the reason what's going on there 
