I've got 8 S3s mining.  6 of them I have listed on MRR, and each of those 6 uses the NH/WH patched version of cgminer to provide renters with the option of pointing them to pools that require it.  The other 2 I have pointed at OgNasty's pool, and they are using the cgminer binary from -ck's site (4.6.1 dated from 20 Oct 2014).

All of the S3s use Bitmain's latest firmware update (antMiner_S320150109.bin).

I'm not sure why you don't see the data... here's what I see on one of mine (pool details intentionally blanked out):

They give similar numbers because they're both based on the same calculation, as described by the formula Steve posted.  Based on expectations (the formula) given a certain hash rate at a certain difficulty, you expect to find a block in a certain time.  Let's take a real example and see how it plays out.  A single Antminer S5 hashes at an advertised 1155GH/s.  Using the formula, I can calculate how long that should take to find a block:
According to expectations, that S5 should take 2138.75 days to find a block, giving me a 1 in 2138.75 chance of finding a block today.

Phil's calculation method uses the above formula as well, but he buries the numbers... he calculates the expected earnings per day, then just takes 25 and divides it by that number.  Using the output from above, I can see that it's 2138.75 days to find a block.  The current block reward is 25BTC.  So, I just take 25 and divide by the number of days to find a block and get my expected earnings per day:

Since that S5 expects to earn 0.01168905BTC per day, which is what the calculator at bitcoinwisdom.com shows as well (they round it off to 0.1169), I just take 25 and divide by it:

Gets me right back to the same 1 in 2138.75 chance.

They're missing BTC China because the string they look for in the coinbase script is wrong.  I found that out when I did my own empty block analysis.  BTC China now puts "BTCChina.com" in the coinbase.  blockchain.info looks for "BTCChina Pool" and "btcchina.com".  They also look for a specific payout address in the coinbase transaction, "152f1muMCNa7goXYhYAQC61hxEgGacmncB".  Apparently that part of their identifying code is broken, since BTCChina still pays out to that address.  For example, I correctly find that block 360755 is mined by BTC China, but blockchain.info does not.

I think it'll go up again, but not by much.  I'll take the +1.76% to +2.00% slot please

Forgive me if this has already been addressed... and if you've been asked like 1000 times already, I apologize for making this the 1001st query .

The latest S3 binary I see is from last year and is version 4.6.1.  I know that the NH/WH guys have some version they coded based on 4.7 for the extranonce subscription.  Am I missing a newer version for the S3, or is the 4.6.1 the latest?

OK... back on topic... gee, there's a surprise about needing to push empty blocks because it's faster /sarcasm.

Not to nitpick here...

Upgrading your service from 100A to 200A will get you only 22kW of total new power, or 17.6kW usable.  Just want to make sure you understand that you're not going to have 24kW available for use out in those sheds, unless of course you don't mind tapping into the original 100A you had in your home already.  You'll need just about 136A out to those sheds to get 24kW of usable power.... this is all assuming you've got 220V service.

I guess if I had to compare it to a lottery, I would state that it's like going to the store and continuously buying quick pick tickets, and you could determine whether or not that ticket was a winner immediately.  Even if you bought out the number of tickets to guarantee a jackpot win, you wouldn't necessarily have one simply because the quick pick might give you the same set of losing numbers multiple times and never give you the winning combination.  Or, you could get lucky and the first ticket you bought won you the jackpot.

I've posted quite a few times in their thread asking them to remove the references to p2pool.  They don't.  Thankfully one of the mods (I believe it was -ck) changed their thread title to get rid of the bogus "supporting decentralization" claim it spouted.  I completely disagree with dogie's statements about it being a "lost in translation" thing.  Bitmain took forrest's code and added some database stuff to it.  They didn't create their own implementation, or attempt to fix any of the variance problems.  Whatever they did was certainly not the $100,000s effort claimed, and they abandoned it since they realized they didn't know what they were doing.

Unless you've got virtually free power, or you get your mining hardware for virtually no cost, you've got almost no shot at ROI with any gear.  For example, let's take a look at the S5.  Right now, new from Bitmain it's $360 (1.434BTC at their current conversion rate).  Let's just pretend there are no other costs at all (you already have a PSU and no shipping costs).  At the current $250 per BTC, even with only a 2% overall increase in the network hash rate per adjustment you will not ever make ROI with $0.10 per kWh power.  Make that adjustment 1% and you'll start to see a profit in about 300 days.

So, I'm glad your primary motivation isn't ROI .  Also, I just wanted to point out something that may be obvious, and forgive me if you're already aware.  The U2 is a SHA256 *only* device.  I mention this because of your forum name.  If you plan to point those U2s to some alt-coin, ensure it's SHA256 or the miners won't work at all.

Yes, it's the close enough formula that most everyone uses.  I'm certainly guilty of using the easy-mode version .

WBF1, I try not to express mining chances in lottery terms (although I certainly have in the past).  With a lottery, you have a very specific set of numbers in play, which define a closed system that allows you to provide accurate chances at a jackpot win.  For example, to guarantee a jackpot win in the powerball, you'd have to purchase 175,223,510 tickets.  To guarantee that you just win *something* you need purchase only 35 tickets.

You can't make that same analysis with BTC.  Yes, you expect that if you produce 49,692,386,354 shares you will solve a block.  However, that's not guaranteed.  It might take 1 share, or it might take 100,000,000,000 shares or theoretically that miner may never ever produce a valid block solve.

Because each share is wholly independent and has a chance at solving a block, we can't truly state that "miner X has Y chance" like we can in lottery terms.  Furthermore, most every lottery has some kind of prize tiering.  Using the powerball again, as long as I match the powerball number, I will win $4.  That's why it only takes 35 tickets to guarantee a win.  There's no such concept in BTC.  You either solve a block or you don't.

Primarily, people setup a proxy and control their miners through it.  I don't have anywhere near 30 miners, but the 8 I do have I control through MRR as TheAnalogKid mentioned.  I also have a VPN server at home so I can log in and control my miners manually from the road.

Oh!  I forgot about my USB miners .  I've got 5 Antminer U2 that give me about 9GH/s.  I have them pointed to -ck's solo pool... because you never know.  They cost virtually nothing to run (they're plugged into an rPi).

Still got the same S3s I purchased last year from the early batches (1,2 and 3).  They're turning 1 soon, and I'm sure I'll have a nice party for them.

Of my original 9, 8 are still functioning within acceptable parameters.  The poor one that is no longer with us failed on me at the beginning of the year.  I get about 3.6TH/s from them.  Cost of electricity is around $0.10 per kWh.  They are still squeaking out a profit...

The fact that you even know what Linux is, and have successfully gotten p2pool and merge mining up and running puts you far ahead of a typical user.  I happen to agree with you in that getting a node up and running is a pretty trivial task with a bit of research and reading; however, people expect a simple plug-and-play operation and running your own node is not.

As for dogie's comments in other threads, they can be addressed in the appropriate threads.  I was simply responding to his reply here, and his statements aren't off the mark.

Not so much forgotten as not included in the tests.  OP chose the pools he wanted to test and p2pool wasn't one of them.  I invited OP to use the data I have been recording as part of my tests on OgNasty's Nasty Pool.  You can see the long-running results of that test here: https://bitcointalk.org/index.php?topic=891298.0

The summary is that my S3 running on the standard p2pool payouts has earned 99.36% of expectations (test started 12/26/2014).

No, it really isn't.

Inconsistent earnings are going to happen on virtually very pool with a low percentage of the network's hash rate.  P2Pool is no exception here.  It's called variance.

High pings.  This is only the case if you're mining on somebody else's node.  Most new users who don't have as much technical experience aren't going to bother setting up their own node (sadly) and will instead mine on other people's nodes.  This leads to the next point...

Higher complexity - setting up and running your own p2pool node requires a bit of technical know-how.  Ideally you've got a dedicated *nix server sitting somewhere, have installed the proper dependencies and have tuned things.  You also need to be running a full Bitcoin node.  Now, if you're mining on somebody else's node, it's just as simple as mining on any pool like kano.is.  However, you're at the mercy of that node operator and whether or not he's competent.

Lower performance / earnings.  Specific hardware will indeed have lower performance or not work at all on p2pool.  This is well known.  Earnings were addressed in the first point.

Now, if you don't feel like running your own node, there are plenty of operators that have proven themselves to run a tight ship.  In these cases, mining on p2pool is just as easy as any other pool.  If you've got the technical wherewithal or just feel like a bit of learning, running your own p2pool node is very rewarding.

There is something wrong with your setup.  I've been successfully running my own node for well over a year and have never experienced this type of behavior.  I have had the 0.10.x core since release candidate running without issue.

I've got the data for every block... I got lazy with lumping things together

I just got lazy there, to be honest.  Any time I couldn't figure out what the pool was, I just lumped it into "unknown" for the stats.  I could certainly break it down to show every address, but figured it was overkill.  I also just figured out why I didn't identify some of the ckpool solo blocks.  The first blocks coinbase script was simply "ckpool/".  I was looking for "/solo.ckpool.org/".  Doesn't matter though since the ckpool code doesn't allow for empty blocks.

Steve, if you want, I'll tweak things again to pump out all of the addresses.  The vast majority of the empty blocks by "unknown" were from the beginning of the blockchain.  Pretty much every block up to 50,000 is empty.  As a matter of fact, I think the first block I saw with more than 1 transaction was 49,076 (but that was just quickly scrolling through the data and I might have missed an earlier block or two).  It's not until the mid 60k range that we really start to see blocks with more than the coinbase transaction start to outweigh the ones with only the coinbase.

Also, there were a number of blocks found where the coinbase was paid to multiple addresses, but the first address in the list couldn't be decoded.  For example, check out block 157606.

Tweaked the code a bit more to run some stats.  Ran it again today.  The vast majority of the empty blocks are from the beginning, which is to be expected.  Here's the data broken down by pool:

There's a bit of inaccuracy in it, since I know that -ck has recorded 67 blocks found in his solo pool, but I only detected 54... not sure if the coinbase wasn't always signed as it is today, or what the root cause of the discrepancy is.

Anyway, from the data and ignoring the unknowns, AntPool is by far the worst offender with 4.62% empty blocks.  Next up is KnC with 2.91% and they are virtually tied with discus fish at 2.9%.  That's what I've got.  Not sure how useful it is, but I had fun putting it together