Right, I cleared up the issue as to whether the roving 1-2 Ph/s was comprised of 13 rigs, aka the 10Ph/s thresh-hold being the major weight responsible for the measly 13% .... I hope you follow. Anyway, here are my interim conclusions.
1. I am not sure I can refer to the roving 1-2 Ph/s as such anymore, seeing th pool's natural rate has shifted to around 9.5 Ph/s (-+5%), else the pool would then be at 10.5-11.5 Ph/s, which it has not breached consistently.... still following ...?
2. I am convinced that the 10 Ph/s thresh-hold is not material in itself, rather that the offending (seemingly pool hopping) hash rate is strategically deployed using load balancing, resulting in work being discarded / omitted when it most possibly have the golden nonce. (I am actually not sure about the logic in this one, but hey, that's my thinking!)
3. Finally, I am also convinced, having monitored kano.is during our last long barren spell, that their luck drastically improved during that time. I have never seen them be so lucky (which they might have been), that I nearly pointed my rigs there again. The punchline to all this is that this roving hash-rate seems to be load-balancing between slush and kano.is and sucking all the luck from here.
Having said that, I can confirm that I had to come to terms with the fact that I had no chance of solving the immensly cryptic task of establishing with utter certainity whether the previously aptly named 1-2 Ph/s roving hash rate was being pumped out by 13 rigs. Having weighed all options up, I decided one conspiracy theory was enough to deal with than adding another on top.

Yep, thought so. You are now submitting shares more frequently so the pool has enough data to more accurately estimate and report your rate.

If you are mining other coins or the pool is setting your diff, then it may be OK, but 4096 certainly is too high for mining BTC on an S3 (especially on a pool). In the final analysis, you should earn the same as the earnings are weighted by the share size, but you will end up submitting fewer shares in the same timespan than if your diff was lower (e.g 400), and your rejects / stales are going to be un-necessarily expensive. Infact, that may explain your earlier lower pool reported rate.

Saying that, your diff is quite high for an S3 .... unless ofcourse you are solo-mining, or not mining just bitcoin.

Most likely when it was showing a lower rate on the pool you'd hit a patch of rejects (as your stales look normal)

@kabopar - what was it hashing at before the jump? It may be the case that bitmain are testing a new rig with the new BM1384 chips (could even be newer ones!) ..... I am expecting a new product from bitmain at the end of the Chinese new year celebration holidays on or soon after the 25th Feb.

I hope that is the AVG rather than the 5s rate you are stating there and would explain the difference as a result of stale / rejected shares and HW errors .... I'd hazard a guess that the difference is greatly accounted for by stales / rejects though. You most likely have a high latency connection to the pool which results rejected shares .... thus the difference in hash-speed

Now that's up I better add that to my computations .... 10.2 Ph/s .... 19hrs

Just noticed something else, 132 of the 199 blocks found (~66%) when the pool was hashing above 9.8 Ph/s (aka blocks with a difficulty of 40640955016 and 39457671307) were found between 17th Dec 2014 and 12th Jan 2015 (thats ~ a 3 week window).

You are so full of yourself and clearly full of self importance. And yeah, the ignore button is a good option for the feeble minded, like yourself, and those that believe its only their opinion that matters. But that you get enough stupidity in your day already should sum you up nicely (and I know you are reading this ..... you just can't resist!).

Anyway, to keep on with my number crunching, I thought I'd continue my comparison but this time for only the blocks found when the pool hashing rate was above 9 Ph/s as with the 13 rigs' roving 1-2 Ph/s there was a more realistic chance of crossing that 10 Ph/s thresh-hold that seems to suck the blocks from the pool, and here goes:

November 2014 - today above 9 Ph/s: 354 blocks)
155 blocks - Less than 9.8 Ph/s
199 blocks - Above 9.8 Ph/s

Don't be silly, I do what I want on my end .... (see, there you go again!)

Same here, had it running without a hitch for weeks, and a couple of days ago, I noticed it had halved hash-rate. Did a software reboot and all was OK until I noticed the same yesterday ... another software reboot and its fine thus far. A bit worrying if you are not keeping an eye on it though ....

At this point, I do not care what you think or know, even whether you might have a point. You've adequately demonstrated to me (thus far today), that your opinion is worth disregarding in its entirity.
If you can confirm either way whether the roving hash-rate is from 13 rigs, that would be your time spent well.

Seeing I have not yet been able to confirm that the roving 1-2Ph/s is generated from 13 rigs, I took another delve into the stats. Rather than analyse the data without considering the two factors that I have been led to believe are important, i.e hash power and difficulty (I discarded luck as un-important), here's what I have based on difficulty.

Diff: 44455415962 (total: 42 blocks)
16 - Less than 9.5 Ph/s
20 - Between 9.5 Ph/s AND 10 Ph/s
7 - Above 10 Ph/s

 ----------------------

Diff: 41272873894 (total: 61 blocks)
58 - Less than 9.5 Ph/s
2 - Between 9.5 Ph/s AND 10 Ph/s
1 - Above 10 Ph/s

 ----------------------

Diff: 43971662056 (total: 48 blocks)
46 - Less than 9.5 Ph/s
0 - Between 9.5 Ph/s AND 10 Ph/s
2 - Above 10 Ph/s

 ----------------------

Diff: 40640955016 (total: 69 blocks)
0 - Less than 9.5 Ph/s
1 - Between 9.5 Ph/s AND 10 Ph/s
68 - Above 10 Ph/s

 ----------------------

Diff: 39457671307 (total: 73 blocks)
4 - Less than 9.5 Ph/s
5 - Between 9.5 Ph/s AND 10 Ph/s
64 - Above 10 Ph/s

 ----------------------

Diff: 40007470271 (total: 69 blocks)
17 - Less than 9.5 Ph/s
16 - Between 9.5 Ph/s AND 10 Ph/s
36 - Above 10 Ph/s

 ----------------------

Diff: 40300030327 (total: 68 blocks)
63 - Less than 9.5 Ph/s
4 - Between 9.5 Ph/s AND 10 Ph/s
1 - Above 10 Ph/s

 ----------------------

Diff: 39603666252 (total: 77 blocks)
76 - Less than 9.5 Ph/s
1 - Between 9.5 Ph/s AND 10 Ph/s
0 - Above 10 Ph/s

 ----------------------

Diff: 35985640265 (total: 18 blocks)
18 - Less than 9.5 Ph/s
0 - Between 9.5 Ph/s AND 10 Ph/s
0 - Above 10 Ph/s

Finally a dicotomous comparison of the data from November 2014 to today total: 526 blocks.

298 - Less than 9.5 Ph/s
228 - Above 9.5 Ph/s

EDIT: looks like we had our time with the 10 Ph/s threshhold when the difficulty was 40640955016 and 39457671307. Since then, our luck has depleted at that rate ..... there must be a pattern to it, but seeing I only have a sample size of 526 (not enough to be relied on ... I think minimum should be 2000), I'll have to concentrate on establishing whether that roving hash-rate is from 13 rigs!

You are not a legend for nothing .... you truly are pushing your point on the pretext that I (even bigtime) do believe there is an issue here (be that with / without scaling or something else sinister) ... seriously? You even have the time to context this in the first half of 2014 (be it the 10Th/s rate)?
So tell me then, what does it take to be a 4 star general on bitcointalk?

If you do not care enough to bother, then what use would it serve for me (or anyone) to prove your percieved lack of a fundamental problem or otherwise, as the case may be? Or is it the case that you really believe that your opinion matters?

Your legendary status precedes you!
While you are still basking in it, pull your finger out and post your own numbers that, apparently, make sense (you patronising so and so!)

Fair enough if you are the quiet one ... (there must be a reason why but thats for another day).
So, if you have the numbers from Slush dating back from October (I assume that is last year), then you'd know without any doubt that my numbers are NOT claims at all, rather facts. That, of course, is on the premise that you engaged your brain before posting to rubbish my numbers, which in this case you certainly did not (and may explain why you'd chosen to be quiet previously!).
Honestly, get off your perch!

That you failed read the entire post in context and thus also failed to decipher the sacarsm in that is only testament to you being new here ...!


I could not get the December numbers, but that is neither here nor there as my numbers are NOT claims, but I can honestly say yours are just that ... because I could not get hold of them when I tried!

But then again, you might be onto something ..... if you look at the numbers from your assertion, they tell us:

142 Blocks where the Pool Scoring Hash Rate was below 10 Ph/s
23 Blocks where the Pool Scoring Hash Rate was above 10 Ph/s

That is, during the period 8th Jan 2015 - 18th Feb 2015, the pool found ~86% of the blocks when the pool was hashing below 10 Ph/s and only 13% when the pool was hashing above 10 Ph/s.
Now, I am not usually a supersticious person, but that 13 in the above 10 Ph/s does put the conspiracy theories on the back burner ..... what do you think? And if it turns out that the 1-2 Ph/s roving hashpower that seems to affect our block solving capacity is made up of 13 rigs, that would be something!