Pay On Target: New High variance payout System Offered by Ozcoin

[Blaksmith]

Hey Guys.
        So I decided to try out the new platform, been changing my difficulty around and such.  I did notice that cgminer is showing a best share of 9.1k vs ozcoins POT stats of 830.  Does the error in cgminer's 'best share' happen often, just one of those blips?  No rejected shares.

Thanks.

We have been tracking it down with Conman, and we think it is a cgminer bug at this point in time.

[1714124836]

1714124836

[1714124836]

1714124836

[makngeerwork]

Had a look at the cgminer thread, seems to be some sort of best share bug.  Restarted cgminer to try and repeat the problem, had a false reading again. I am ignoring cgm.  

Thanks.

This is why Ozcoin is best in class.

[philips]

Argh, this is really frustrating. I had several best shares (reported by cgminer) that never appeared in Ozcoin stats.
I just had one at 58.5k and the higher reported share is 278.

Few hours ago I had (at least) one at 61k, one at 40k and one at 980k, all of them lost...I mean come on, what´s the point?

Is cgminer still reporting false best shares...or something broken in between...?

(running cgminer 2.10.2 with the best share bug fixed)

[-ck]

Argh, this is really frustrating. I had several best shares (reported by cgminer) that never appeared in Ozcoin stats.
I just had one at 58.5k and the higher reported share is 278.

Few hours ago I had (at least) one at 61k, one at 40k and one at 980k, all of them lost...I mean come on, what´s the point?

Is cgminer still reporting false best shares...or something broken in between...?

(running cgminer 2.10.2 with the best share bug fixed)

Alas the best share bug is NOT fixed, sorry.

[Blaksmith]

Argh, this is really frustrating. I had several best shares (reported by cgminer) that never appeared in Ozcoin stats.
I just had one at 58.5k and the higher reported share is 278.

Few hours ago I had (at least) one at 61k, one at 40k and one at 980k, all of them lost...I mean come on, what´s the point?

Is cgminer still reporting false best shares...or something broken in between...?

(running cgminer 2.10.2 with the best share bug fixed)

From what we can tell so far, that somewhere inside cgminer, it is setting it... I have been doing some personal tests following a share from when it arrives on the server to where it gets credited, and I have seen the best share go up on cg, but not see it even enter the stream on ozcoin.  Conman has also confirmed that as well.

[philips]

Oh, OK thanks...at least I know what´s going on.
(crap, another false best share at 675k)

Is that fix from...the Unnamed one gonna solve the issue? (cgminer thread)

[-ck]

Oh, OK thanks...at least I know what´s going on.
(crap, another false best share at 675k)

Is that fix from...the Unnamed one gonna solve the issue? (cgminer thread)

I'm investigating, but I will not be pulling in his code.

Should be fixed in git now.

[philips]

@ckolivas

OK thanks, installed the latest git source tarball and see how it goes.

[mdude77]

I admit it, I didn't read all 6 or 7 pages of this.  It seems like a neat idea, but I don't follow the math.

I've been using this for almost 2 days now, give or take, and according to oz's POT stats page, I'm about 0.05BTC behind where I'd be if I was using PPS. 

Using the calculator on the first post, if you solve a block, or get the largest allowed share, your reward is:

POT Value    0.39800980 BTC

Yes, that's significantly larger than what you'd normally get, but it doesn't really seem like much for solving a block.  I get more by solving a block on my own p2pool node with NMC merged mining (at current BTC <--> NMC exchange rates)!

Am I missing something here?

M

[philips]

@ckolivas

OK thanks, installed the latest git source tarball and see how it goes.

From what I see the issue is gone, no more false high shares since last night.

However, now I can see that higher shares are quite rare in fact, and the whole POT thing kinda lost its appeal to me.
I get that is about luck and gambling, but since I started I kept going and going downhill and occasionally higher shares were not enough to get me back on track.

Maybe I am not lucky enough or most likely not patient enough...but there is too little time left until my GPU rig is going to scrap to gamble with it.

So long, and thanks for all the fish, it was really fun though

[Graet]

Thanks for the feedback and giving it a try guys
The trial period is not yet over and with the festive season upon us coders have been a little busy with non-Bitcoin related stuff.
We will assess the numbers and possibly tweak some more.

I also feel the cgminer incorrect best share in display has caused too much confusion - looking forward to next release

[philips]

Thanks for the feedback and giving it a try guys
The trial period is not yet over and with the festive season upon us coders have been a little busy with non-Bitcoin related stuff.
We will assess the numbers and possibly tweak some more.

I also feel the cgminer incorrect best share in display has caused too much confusion - looking forward to next release

I still think that this is a brilliant idea with a future.
Also, do you have any plans to implement this on your Litecoin pool (if possible) ?

[Graet]

Thanks for the feedback and giving it a try guys
The trial period is not yet over and with the festive season upon us coders have been a little busy with non-Bitcoin related stuff.
We will assess the numbers and possibly tweak some more.

I also feel the cgminer incorrect best share in display has caused too much confusion - looking forward to next release

I still think that this is a brilliant idea with a future.
Also, do you have any plans to implement this on your Litecoin pool (if possible) ?

Thanks, I think the idea has merit too, we will persevere

Most focus has been on the Bitcoin pool, we are currently investigating ecoinpool or stratum with DGM and PPS for Litecoin, if we go with stratum it would be easy to implement this payout system too

[-ck]

New version of cgminer, 2.10.3 should finally have fixed the "best share" displayed bug  . It also offers a Zero statistics button in the Display menu allowing people to reset stats.

[Graet]

Thinking about continuing trial and lowering a=0.8 to 0.75 or 0.7 reduces variance a little and raising cap to 5*D instead of 1.5*D

feedback welcome

[mdude77]

Thinking about continuing trial and lowering a=0.8 to 0.75 or 0.7 reduces variance a little and raising cap to 5*D instead of 1.5*D

feedback welcome

I'd consider coming back if a stayed at 0.8 and the cap increased.  Lowering it lowers the max payout.

Honestly, though, with difficulty decreasing significantly for both btc and nmc, I'm leaning towards staying on p2pool for a while.

M

[Graet]

Thinking about continuing trial and lowering a=0.8 to 0.75 or 0.7 reduces variance a little and raising cap to 5*D instead of 1.5*D

feedback welcome

I'd consider coming back if a stayed at 0.8 and the cap increased.  Lowering it lowers the max payout.

Honestly, though, with difficulty decreasing significantly for both btc and nmc, I'm leaning towards staying on p2pool for a while.

M

cool, lowering lowers max payout but raises minimum payout, also increasing the cap will raise maximum payout.

I don't mind where people mine - there has been demand for a payout system based on difficulty of submitted share, we are trying to fill that demand, I have deliberately pointed out it is a high variance payout system and also called it "gambling "
Also with difficulty lowering and D being a number in the equation the rate per share will go up when difficulty next drops.
cheers
Graet

[philips]

New version of cgminer, 2.10.3 should finally have fixed the "best share" displayed bug  . It also offers a Zero statistics button in the Display menu allowing people to reset stats.

Wow, so there really is a Santa!

[-ck]

Thinking about continuing trial and lowering a=0.8 to 0.75 or 0.7 reduces variance a little and raising cap to 5*D instead of 1.5*D

feedback welcome

I'd consider coming back if a stayed at 0.8 and the cap increased.  Lowering it lowers the max payout.

Honestly, though, with difficulty decreasing significantly for both btc and nmc, I'm leaning towards staying on p2pool for a while.

M

The higher the cap goes, the more risk the pool takes. The miner's per share income drops ever so slightly when the cap is raised, but almost a non-noticeable amount. So really, the miners should have nothing against raising the cap, it's up to the pool op to decide how much risk to take. On the other hand though, the a value changes dramatically how much variance the miner is likely to see (and then how high the reward would be at the max cap). I'm quite happy with a high a and the cap being equal to difficulty since I'm actually unlikely to find a block at my hashrate, but others may see things differently

[vapourminer]

Thinking about continuing trial and lowering a=0.8 to 0.75 or 0.7 reduces variance a little and raising cap to 5*D instead of 1.5*D

feedback welcome

sounds good to me.

just as a fyi. running about 12% behind here, been on it since the 19th. sorta a slow slide downwards, as it was about 2-3 % under for the 1st few days. not complaining, just spouting figures.