Topic: Weekly pool and network statistics - page 18. (Read 91444 times)

Ho ooc
http://ozcoin.net/content/hall-fame-round-shares-bitcoin is by round (users PPS and pot are dummy users to keep DGM stats correct - probably shouldn't be displayed)
http://ozcoin.net/content/hall-fame-lifetime-shares-bitcoin is alltime since we moved to the new site (30 December 2011)
and everything is converted back to D1 equivalent for payout and stats calculations
hope thats a help

Keep up the good work

Glad you like!

CSV or JSON - doesn't really matter since I just need a vector of hashrates in Ghps, nothing else.

What I'd really like is the 24 hour average hashrate per user account. Is 24 hours a problem? Not many pools average for that long, but I suspect that has to do with what miners want rather than what I want

I've also meant to ask Graet about this - Ozcoin's Hall of Fame lists shares contributed per miner, but I'm not sure over what sort of time period, or whether they're D1 equivalent.

Any pool op that doen't yet publish these stats but wants to be added is welcome to post here or PM me. I want to (at some point) provide a chart of how each pool's hashrate distribution changes over time.

Nice work, as always.

I like the new violin plot of hashrate distribution among miners in a pool.

What kind of API/data do you prefer, to add more pools there?

14th April 2013 weekly pool and network statistics

Welcome, miners.

Changelog:
Figure 8 (Chart of fulltime miner earnings loss at a proportional pool caused by pool hoppers) has been removed since very little pool hopping has occurred. Instead, figure 8 is now the user hashrate distribution, as a follow up to post 12.1 The pre-ASIC network hashrate distribution.

Pools missing from results:

None, although MTRED (which sadly closed it doors this week) will be missing from next weeks' results. They're included this week since there were nine blocks solved by the pool during the current week before they closed.


"ASIC ready" pools:

BTCGuild
50BTC.com
Slush's pool
Ozcoin
HHTT
Bitminter
EclipseMC
Eligius
Bitparking
Itzod
Triplemining

Pool hopping:
None
BTCGuild is almost at 40% of the weekly average network hashrate this week (although only 36.7% of network blocks, thanks to some bad luck). The pool is keeping pace with the increase in weekly average network hashrate, ~ 6 Thps. This is good new for the pool and the network in general - on average, miners are choosing not to start at BTCGuild or move there from other pools.

Bitminter has become a top 5 hashrate pool and Slush's pool is vying with 50BTC.com to hit second spot - back to where it was when DeepBit was king.

As usual, please post comments if there's anything you don't understand, with which you disagree, or just think is wrong.

You can view this weeks charts at http://organofcorti.blogspot.com/2013/04/14th-april-2013-weekly-pool-and-network.html

You can view all previous charts and other posts and fun things at http://organofcorti.blogspot.com

Edit: Ooops - I forgot to post here last week. If you want to have a look, last weeks update was at  http://organofcorti.blogspot.com/2013/04/31st-march-2013-weekly-pool-and-network.html



7th April 2013 weekly pool and network statistics

Welcome, miners.

Changelog:
Figure 3 (top 10 hashrate history) is back to using a base 10 logarithmic y axis since some pools which once had a very high proportion of the network have a much lower proportion and visa versa. The only way to show a comparison is by using a log y axis.

Pools missing from results:

None


"ASIC ready" pools:

BTCGuild
50BTC.com
Slush's pool
Ozcoin
HHTT
Bitminter
EclipseMC
Eligius
Bitparking
Itzod
Triplemining

Pool hopping:

Slush's pool - was being hopped for about 5 days during the week. Still not much though, and I doubt it's having a significant impact on fulltime miner earnings.

The weekly average network hashrate increased by ~ 6 Thps this week. BTCGuild's hashrate only increased by another 2Thps, with 38% of the network blocks for this week. Pool op eleuthria has decided on a number of proposals to limit BTCGuild's influence over the network, including increasing the pool fee if the pool's proportion of the network exceeds 40%. This is a move I applaud, and is a good indicator of eleuthria's commitment to bitcoin and not just profits.

"Unknown" is down to 5% of the network. There are a few pools of which I know that I can't yet include in the weekly stats, and between them they account for about 2% of the netwrok - so there's still 3% completely unknown. So still no indication that any significant number of ASICs are going solo.

HHTT lost 90% of their hashrate in the past few days, in the wake of concerns about pool bankruptcy - an ever present possibility for a PPS only pool. Pool op fireduck is busy changing the payout method to PPLNS and has assured all miners that all owed earnings will be paid. Since fireduck also works for SatoshiDice, I'd have to guess he can make good on that commitment.

ecki, previously small pool ( run by a German fellow permanently work-vacationing on Koh Samui , and of whom I'm rather jealous ) just increased their hashrate by 1000% with the addition of someone with about 7 Avalon's worth of hashrate. Unfortunately I can't yet include them in the network stats since their page is a bit hard to scrape, but I hope to include them soon.

As usual, please post comments if there's anything you don't understand, with which you disagree, or just think is wrong.

You can view this weeks charts at http://organofcorti.blogspot.com/2013/04/7th-april-2013-weekly-pool-and-network.html

You can view all previous charts and other posts and fun things at http://organofcorti.blogspot.com

24th March 2013 weekly pool and network statistics

Welcome, miners.

Changelog:
Nil

Pools missing from results:

Maxbtc API not working


"ASIC ready" pools:

BTCGuild
Slush's pool
Ozcoin
HHTT
Bitminter
EclipseMC
Eligius
Bitparking
Itzod

Pool hopping:

Slush's pool

Many more ASICs came online this week - the average weekly network hashrate increased by about 10Thps to ~49 Thps, and many of those went to BTCGuild  who alone accounted for ~ 36% of the network. The only other pools to see a significant increase in their share of the network was HHTT who now are in the top five pools, and Bitparking. Many pools had an increase in their hashrate, but none of them significantly more than the network increase.

As usual, please post comments if there's anything you don't understand, with which you disagree, or just think is wrong.


You can view this weeks charts at http://organofcorti.blogspot.com/2013/03/24th-march-2013-weekly-pool-and-network.html

You can view all previous charts and other posts and fun things at http://organofcorti.blogspot.com

Yes, the hashrate distribution varies widely between pools.

If we could a majority of pools to provide a simple API - just a list of each user accounts' 24 hour average D1 equivalent hashrates - it would make the comparisons more valid, and possible to perform weekly.

pool users and workers are already in our API.
but
we had 287 users and 800 or so Gh a while back
right now we show
Pool: [ 4,858.78 GHash/s | 300 Users | 775 Workers and 1 user with 2.8Th/s
and a couple of days ago 7Th with 309 users

personally I don't think user numbers are a good indicator currently unless you also look at user hashrates.

It's a bit chicken-and-egg. Hard to know exactly what the data format should be until I've analysed it, and it's hard to know how to analyse it if the data format varies from pool to pool.

But from what I've seen so far, I'd need the 24hr average D1 hashrates, or proofs of work with average difficulty attached. The miners don't have to be in order or named, but it does need to be the total hashrate per account rather than per miner.

I wouldn't expect pools to go to the hassle of creating an API just for that until I can show some use for the data. Hopefully once I finish my current work and post it there'll be enough interest from enough pools that I'll be able to include the weekly pool user hashrate distributions in the weekly pool stats.

I have data on how much work each user did in each shift and each block round. I know the duration also so I can use that to calculate each user's average hashrate during a shift or round. It gets a little crazy for 1-second rounds though.

I'm not sure how to massage this into some useful metrics to make available through the API.

Perhaps if you let us know what you need, pools can deliver.

Well now, that's handy. But if you're worried about decentralisation, the number of miners isn't as important as the number of miners plus as the hashrate distribution of the miners. That's what is hard to get synchronously, and since some pools average the hashrates for 15 minutes and others for 24 hours, it's very hard to compare the pools. That's why it's taking a while.

Number of active users mining at BitMinter is available at http://bitminter.com/api/pool/users

We went from 900 to 1500 active users over the last two months.

It's possible to get the information, but hard to do synchronously since for most pools you have to request the data from the pool ops. I have actually been working on something along those lines for the pre-ASIC era, so keep and eye out for it.

Do pools publish information about the number of miners submitting work?  It seems like this would be an important number to estimate as decentralization is key to Bitcoin's security.  It would also be interesting to see how many people drop out when ASICs hit.

It's an interesting word - I heard it first when I was little. It was in some Seventies song (anyone name it for me? I 've forgotten).

I liked the word even more when I found out no one knows exactly what the etymology of the word is - it's the creature of the black lagoon of Americanisms.

Word of the day? I had to look up that one.

Absolutely, on both points. Hence the "can" - p2Pool can net you more if you set the donation to the dev lower than whichever pool at which you mine currently. For p2Pool to run well, you need to be a bit sysdmin-ish, (or be prepared to learn), and there are other geek requirements. I wouldn't recommend it to someone (like me) who wants to mine and just check once a day that everything is copacetic.

I personally find the idea of low maintenance, easy to manage solution (like yours) more appealing, but for miners who love fiddling with their set ups, p2Pool could be a good choice. I have other things with which I prefer to fiddle (mind out of the gutter, DrH!).

I think you meant to say that since true zero-fee pools like BitMinter, MtRed and P2pool will pay better in the long term, you may want to look into that. And on BitMinter you don't need to be geeky, just click a button.

17th March 2013 weekly pool and network statistics

Welcome, miners.

Changelog:
Nil

Pools missing from results:

Eclipse Mining Consortium - API not working.
p2pcoins.net,  Mkalinin - these pools were hacked and the owner has taken them off-line.


"ASIC ready" pools:

BTCGuild
Ozcoin
Slush's pool
Bitminter
EclipseMC
HHTT
Itzod
Bitparking


Pool hopping:

Slush's pool


Lots of drama this week. The hard fork had a significant effect on the midweek hashrate, so the retarget difficulty was a little less than estimated.  Figure 4 shows how pool's hashrates were affected by the downtime.

Earlier in the week while BTCGuild pool op eleuthria was upgrading a merged mining server to bitcoin 0.8.0, a bug allowed miners to credit their accounts with far more bitcoin than they were due. This personally cost eleuthria 1250 btc - about US $60 000 at current exchange rates. So far only one miner has returned any of the coins. CryptoCluster returned the 47.8 btc (US $2 300) without even a request email - that's certainly one way to increase your standing in a community that is mostly anonymous. On the other hand, a miner (whose details eleuthria has not released) responded to eleuthria's polite request to return the overpayment with scorn. I hope we learn his username, since I wouldn't want to do business with such a community member.

Eclipse Mining Consortium was DDoS'd for much of the week. I'm not sure whether his response to this has affected the API was using for his pool data, but I expect to have it fixed for next week. This means the "Unknown" section of the pir chart is about 1.5 Thps larger than it would be if Eclipse had been included. This means that "Unknown" is down to about 6%. It appears not much solomining mining is being done by ASICs.

Both Eligius and BitMinter overtook Deepbit this week, and thanks to some hefty ASIC power added Eligius' hashrate jumped to 5 Thps for a day or so.

Bitcointalk forum member gyverlb posted an excellent guide to mining on p2Pool. Since mining on p2Pool can net you more coin than mining on other centralised pools, this post is important reading for any geeky miner.

As usual, please post comments if there's anything you don't understand, with which you disagree, or just think is wrong.

You can view this weeks charts at http://organofcorti.blogspot.com.au/2013/03/17th-march-2013-weekly-pool-and-network.html

You can view all previous charts and other posts and fun things at http://organofcorti.blogspot.com

Interesting is an understatement lol