Weekly pool and network statistics

[organofcorti]

I could do that but it would be cumbersome however many people make the their p2pool nodes public.

Here is forrestv (p2pool developers) node:
http://forre.st:9332/static/

It might be easier to just grab the data when you need it.

Wasn't aware of that, I'd only seen the graphs on p2pool.info. Graphs on http://forre.st:9332/static/graphs.html are much more useful. Thanks, D&T.

[DeathAndTaxes]

No problem.

[organofcorti]

Weekly average pool hashrates, 2nd May 2012:

Code:

          Pool    Time.in.days    Av.Hashrate.Ghps   %.of.network.hashrate
1      Deepbit          7.0             3516                 31.34
2        Slush          7.0             1227                 10.94
3       Ozcoin          6.7              661                  5.89
4      Eligius          6.7              454                  4.05
5      Eclipse          6.6              432                  3.85
6       P2pool          7.0              333                  2.97
7        Bitlc          6.5              171                  1.52
8    Bitminter          6.7              149                  1.33
9       Maxbtc          5.2              149                  1.33
10 Bitclockers          6.8              111                  0.99
11       Itzod          6.1               28                  0.25

Weekly average % of network hashrate to date:

Code:

               Apr.17.2012 Apr.24.2012 May.02.2012
Bitclockers        1.01        1.05        0.99
Bitlc              1.53        1.46        1.53
Bitminter            NA        1.18        1.33
Deepbit           33.66       32.29       31.34
Eclipse              NA        3.83        3.85
Eligius            4.01        3.94        4.05
Itzod                NA        0.37        0.25
Maxbtc             1.33        1.37        1.33
Ozcoin             6.06        6.35        5.89
P2pool               NA          NA        2.97
Slush             11.62       11.51       10.94

[SgtSpike]

It'd be interesting to somehow compare the average hashrate to the pool payouts.  See if there are any pools that are consistently paying out less than the others with regards to a payout/hashrate ratio.

I suppose it would require a miner to be pointed at each pool, and assume all else being equal.

Maybe I'll do it when I get my 10 BFL's in. 

[organofcorti]

If you ignore the losses due to strategic mining, this could be done as a percentage of network reward for the time period:

Pool blocks solved / Network blocks solved

Sort of a luck measurement. Good idea, Sgt Spike

Now, I don't suppose you'd mind parting with 10% of those BFL singles, would you?

[SgtSpike]

If you ignore the losses due to strategic mining, this could be done as a percentage of network reward for the time period:

Pool blocks solved / Network blocks solved

Sort of a luck measurement. Good idea, Sgt Spike

Now, I don't suppose you'd mind parting with 10% of those BFL singles, would you?

I'll part with any of them if the price is right. 

[organofcorti]

If you ignore the losses due to strategic mining, this could be done as a percentage of network reward for the time period:

Pool blocks solved / Network blocks solved

Sort of a luck measurement. Good idea, Sgt Spike

Now, I don't suppose you'd mind parting with donating 10% of those BFL singles, would you?

I'll part with any of them if the price is right. 

Should have been more specific. Sorry.

[SgtSpike]

If you ignore the losses due to strategic mining, this could be done as a percentage of network reward for the time period:

Pool blocks solved / Network blocks solved

Sort of a luck measurement. Good idea, Sgt Spike

Now, I don't suppose you'd mind parting with donating 10% of those BFL singles, would you?

I'll part with any of them if the price is right. 

Should have been more specific. Sorry.

That would be the wrong price.  :p

[check_status]

It'd be interesting to somehow compare the average hashrate to the pool payouts.  See if there are any pools that are consistently paying out less than the others with regards to a payout/hashrate ratio.

I suppose it would require a miner to be pointed at each pool, and assume all else being equal.

Doesn't require mining, only a calculator, provided the pools post enough stats to make calculations.

1 share ≈ 2^32 hashes or 4294.967296 megahashes
60*60*24 seconds in a day = 86,400 seconds/day
1010MH/sec = 87,264,000 Megahashes/day

Shares/day = (MHash/day) / (Megahashes/share)
Shares/day = 87264000 / 4294.967296
Shares/day ≈ 20317.73
Payout/day = 20317.73 * 0.00003 Bitcoins/share
Payout/day = 0.60953199 Bitcoins

Above, the Payout Per Share is .00003

Eclipse has a 12% better payout currently over Ozcoin, while Eligius doesn't publish, that I could find, average shares round.

{Pools total mega hashes} * 86,400 / 4294.967296 * nPPS / {Pools total Gigahashes} = ⊅BTC per GH/s per day
nPPS can be any reasonable arbitrary value, .00003 can be used. It won't represent actual payouts just differences between pools.

[SgtSpike]

It'd be interesting to somehow compare the average hashrate to the pool payouts.  See if there are any pools that are consistently paying out less than the others with regards to a payout/hashrate ratio.

I suppose it would require a miner to be pointed at each pool, and assume all else being equal.

Doesn't require mining, only a calculator, provided the pools post enough stats to make calculations.

1 share ≈ 2^32 hashes or 4294.967296 megahashes
60*60*24 seconds in a day = 86,400 seconds/day
1010MH/sec = 87,264,000 Megahashes/day

Shares/day = (MHash/day) / (Megahashes/share)
Shares/day = 87264000 / 4294.967296
Shares/day ≈ 20317.73
Payout/day = 20317.73 * 0.00003 Bitcoins/share
Payout/day = 0.60953199 Bitcoins

Above, the Payout Per Share is .00003

Eclipse has a 12% better payout currently over Ozcoin, while Eligius doesn't publish, that I could find, average shares round.

{Pools total mega hashes} * 86,400 / 4294.967296 * nPPS / {Pools total Gigahashes} = ⊅BTC per GH/s per day
nPPS can be any reasonable arbitrary value, .00003 can be used. It won't represent actual payouts just differences between pools.

But isn't the pool hashrate estimated based on the number of shares?  Couldn't the pool misrepresent the number of shares?  Or the payout per block?  Or etc, etc, etc?

I think there's a lot of wiggle room for pools to potentially skim or scrap coins off the top without anyone noticing.  That's why I'd be interested in a direct comparison on the same hardware between 10 of the most popular pools.

[organofcorti]

The easiest way to do something like you suggest that's also hard to detect is to simply keep round rewards for very short rounds.

Even if you're blatant about recording abnormal round lengths - like bitclockers, although I'm not saying they stole any round rewards, just faked rounds lengths - it will take a while to record enough data to prove it unless you're deepbit.

If you go to the Neighbourhood poolwatch blog (see my sig) there's more info about these types of problem.

[check_status]

But isn't the pool hashrate estimated based on the number of shares?  Couldn't the pool misrepresent the number of shares?  Or the payout per block?  Or etc, etc, etc?

I think there's a lot of wiggle room for pools to potentially skim or scrap coins off the top without anyone noticing.  That's why I'd be interested in a direct comparison on the same hardware between 10 of the most popular pools.

I expect the hashes per second displayed by a pool to be honest numbers.
If a pool advertises 500 GH/s, I would expect 500 * 1 GH/s rigs are mining at that pool, for example.
Calculating hashing power any other way is deceptive because the value is altered to appear other than it is.

[organofcorti]

Also, it's easy to record the number of shares you submitted locally. If your valid shares minus your stale/invalid shares don't add up to the number of shares you're credited with, or if your stales (as recorded on the pool website) are much higher than you record locally, you'll move to another pool.

It's easier just to credit miners with shares correctly, and just fake round lengths. Or just take the whole reward for short blocks. Neither of these scams work on PPS pools of course - for PPS you just have to keep your eye on submitted vs accepted shares and on your stale/invalid share to valid share ratio.

Also won't work at Slush's pool since you don't get to know your accepted shares, and I'm not sure if you can back calculate from the score since it get normalised hourly - that is intraround sometimes. Lucky Slush is a trustworthy guy.

[organofcorti]

Thanks for the pool 'luck' suggestion, Sgt Spike.


Weekly average pool hashrates, 8th May 2012:




Weekly average % of network hashrate to date:


Weekly luck index to date:

[Graet]

good work again
thanks

[organofcorti]

Nice pool luck there, Graeme. Ozcoin's lookin' good.

[Graet]

Nice pool luck there, Graeme. Ozcoin's lookin' good.

heh thanks , we had a few ugly days, but yesterday caught us up nicely

* Graet tries not to anger the luck gods by talking about luck

eg

luck has been excellent here recently 

*knocks on wood*

Since you talked about it, you have automatically killed it. Never discuss luck because that breaks things.

That did break it for 2 blocks in a row at the time he said that

lmao

[organofcorti]

Nice pool luck there, Graeme. Ozcoin's lookin' good.

Fixed that!

[Graet]

Nice pool luck there, Graeme. Ozcoin's lookin' good.

Fixed that!

LOL - literally
<3
and thanks . things are looking good atm, more good stuff to come - stay tuned

[DrHaribo]

What, 19% above expected? BitMinter did 42% above expected.

Watch us do it again next week!