Pages:
Author

Topic: Pay On Target: New High variance payout System Offered by Ozcoin - page 6. (Read 36502 times)

vip
Activity: 980
Merit: 1001
had my whip out, Blaksmith has worked up a sweat pounding code on his anvil
new formula running Cheesy

round((1 - $a) / (1 - $a * pow($wd, (1 - $a)) * pow($X, ($a - 1))), 10) * ($wd * $B / $D) * pow(($sd / $wd), $a)

a = 0.8
X = bitcoin_difficulty * 1.5
B =  Currently set to 24.25 (take out 3% fee)
D = bitcoin_difficulty
wd = work from pool
sd = work returned to pool from miner

Lets see how this goes Smiley
Good luck to anyone playing Smiley
Graet


One lucky blocksolver (the only POT block so far) 2012-12-21 01:26:19 share diff 19 work diff  5218578 paid 0.61359400 BTC
congrats Cheesy
legendary
Activity: 4354
Merit: 3614
what is this "brake pedal" you speak of?
Im 2.4% under. seems ok to me. most shares are under but the over ones do a pretty good job of evening it out so far.. and it IS a gamble of sorts.

Ill stick it out some more. makes coming home to see the stats fun again Smiley
vip
Activity: 980
Merit: 1001
A few winners starting to appear Smiley
Site stats are getting better and thanks to kano for the calculator - I can see it getting some lots of use Smiley

We will be changing to the latest formula after Blaksmith finishes work today

Hopefully soon we can get something works for both sides

Please be aware that the nature of gambling is you can win or lose - this payout method is putting the luck back onto the miner - the shares you submit are what you are paid on, if your luck is bad you will lose, if good you will win.


Thanks for the feedback and help, apologies for the delays (at least you didn't have to pre-order - joke) it is taking a bit longer than expected to get sorted, I feel we are getting closer, thanks to miners that have persevered during the trial Smiley

Best wishes
Graet
legendary
Activity: 952
Merit: 1000
As expected unless you hit some 1M+ shares...

LOL, 17% less than PPS? .. I dont think thats, "expected" .. "expected" would be hovering around 100% pps.. and deviate from that up n down .. if I can maintain a 117% pps for just as long a period, then I will retract my statement that the formula is broken...till that time, its broke =P
I can confirm the opposite. Running POT nearly 2 days with 4 Ghash/s now and I am at 18% above PPS rate. This is because I got an 3.18M share at diff 7.

Great idea with this new payout method, definitely bringing some fun into mining again Smiley - And I hope my luck won't faint...
Right now I'm at 12% under PPS, but a single share of 1million would bring me up to 45% over PPS. Damn you PoT for stringing me along! Tongue
legendary
Activity: 2450
Merit: 1002
I hear what you guys are saying... but to me, if I take the "avg share diff 1" column value and plug that into my spreadsheet w/ a wdiff of 1 and submitted shares value of "avg share diff 1"(it shows me the pot PPS based on my "avg share diff 1" then multiply  by the value shown in "valid vardiff shares" column - that should be "ON AVERAGE" close to the cumulative paid amount per block row. Which, for nearly all my rows its about 20% short.

Meaning, the spreadsheet says, my calculated avg share diff 1 need to be a value of 8 to = ozco pps - fees.
But according to stats page, its looking like avg share diff 1 needs to be at least around 10+ to = ozco pps - fees....
newbie
Activity: 14
Merit: 0
As expected unless you hit some 1M+ shares...

LOL, 17% less than PPS? .. I dont think thats, "expected" .. "expected" would be hovering around 100% pps.. and deviate from that up n down .. if I can maintain a 117% pps for just as long a period, then I will retract my statement that the formula is broken...till that time, its broke =P

I can confirm the opposite. Running POT nearly 2 days with 4 Ghash/s now and I am at 18% above PPS rate. This is because I got an 3.18M share at diff 7.

Great idea with this new payout method, definitely bringing some fun into mining again Smiley - And I hope my luck won't faint...
donator
Activity: 2058
Merit: 1054
As expected unless you hit some 1M+ shares...

LOL, 17% less than PPS? .. I dont think thats, "expected" .. "expected" would be hovering around 100% pps.. and deviate from that up n down .. if I can maintain a 117% pps for just as long a period, then I will retract my statement that the formula is broken...till that time, its broke =P
The Pay-on-Target payout distribution is highly asymmetric. You don't get roughly equal numbers of above-average and below-average shares; rather, you get (with a = 0.8 ) 87% of the shares to be lower than average, then once in a while you hit the jackpot. Maybe you will soon get that big win that will bring you back to the black, maybe not.

This of course does not rule out the possibility of an issue with the implementation.
legendary
Activity: 2450
Merit: 1002
As expected unless you hit some 1M+ shares...

LOL, 17% less than PPS? .. I dont think thats, "expected" .. "expected" would be hovering around 100% pps.. and deviate from that up n down .. if I can maintain a 117% pps for just as long a period, then I will retract my statement that the formula is broken...till that time, its broke =P
legendary
Activity: 1666
Merit: 1000
As expected unless you hit some 1M+ shares...
legendary
Activity: 2450
Merit: 1002
Well as it stands now, POT is horribly broken in my case. Im seeing negatives in the PPS difference column on almost every block(on the site stats)
My overall is like 83% pps currently. So, the formula (hopefully new generations of it) will help fix this. Seems like my luck should be avg'n out to at least 100% ozcoin pps overall =P not a deficit of 17% =(
donator
Activity: 2058
Merit: 1054
Oh, I'm only using a lone BFL Single, but my varr diff is manually set to 2. Would that affect my earnings much? From what I can tell, a manually set work_difficulty makes for higher-paying high-diff shares, but fewer low-paying low-diff shares. Is this correct?
Sort of. You get paid the same as if the shares you submitted were only half the difficulty the are(since vardiff = work_difficulty =2), but multiplied by 2.

Think of it like this:

Reward per share = function(share_difficulty/work_difficulty) *  work_difficulty

So playing around with Kano's calculator (Thanks, btw! Just what I was looking for), it seems that manually setting work_difficulty to a higher value does increase the payout per share by a bit, at almost any share_difficulty. How is this offset? Just the fact that fewer shares are submitted?
Yes. By upping the work_difficulty, shares which are less than the new value will not be paid at all. This offsets the higher payment for the shares that are paid.

I understand your first formula (the on in the OP) pretty well, the second I'm having a harder time wading thru. At this point, I'm just trying to figure out practical applications.
The second formula is what allows you to maintain the same average fee while capping the payout. It's more complicated, and that's part of the reason I'm not so enthusiastic about capping.


heh well I guess I need to check what OzCoin is using at the moment - coz that's what matters
(I'm not sure they've updated it yet to whatever the next version of the calculation is)
I put the equations I'm using at the moment at the bottom of the web page
(and the cap I'm using is simply a cap on 'Your Share Difficulty')

Edit: i.e. I'm using the OP post equation
Well if Graet is capping the shares but hasn't updated the formula, he's effectively taking 4% extra fee on top of the advertised fee. I hope this will be fixed soon.
legendary
Activity: 952
Merit: 1000
Oh, I'm only using a lone BFL Single, but my varr diff is manually set to 2. Would that affect my earnings much? From what I can tell, a manually set work_difficulty makes for higher-paying high-diff shares, but fewer low-paying low-diff shares. Is this correct?
Sort of. You get paid the same as if the shares you submitted were only half the difficulty the are(since vardiff = work_difficulty =2), but multiplied by 2.

Think of it like this:

Reward per share = function(share_difficulty/work_difficulty) *  work_difficulty

So playing around with Kano's calculator (Thanks, btw! Just what I was looking for), it seems that manually setting work_difficulty to a higher value does increase the payout per share by a bit, at almost any share_difficulty. How is this offset? Just the fact that fewer shares are submitted?

I understand your first formula (the on in the OP) pretty well, the second I'm having a harder time wading thru. At this point, I'm just trying to figure out practical applications.
legendary
Activity: 4592
Merit: 1851
Linux since 1997 RedHat 4
heh well I guess I need to check what OzCoin is using at the moment - coz that's what matters
(I'm not sure they've updated it yet to whatever the next version of the calculation is)
I put the equations I'm using at the moment at the bottom of the web page
(and the cap I'm using is simply a cap on 'Your Share Difficulty')

Edit: i.e. I'm using the OP post equation
donator
Activity: 2058
Merit: 1054
Meanwhile ... Smiley
Rather than typing numbers into my desktop calculator over and over ... I wrote this:
http://tradebtc.net/potcalc.php

The defaults (which you can change before you press 'Calculate' are a=0.8 Cap=1.5 Fee=3%)
I'll change the defaults to whatever Graet sets them to ... when I notice they've changed Smiley
I don't think you used the constant factor that corrects for share capping.
Yeah I was applying it to share count Tongue
I've changed it to cap the share value ... hopefully that's correct.
I'm not sure we're on the same page, the formula for payout is

[(1-a)/(1-a*wd^(1-a)*X^(a-1))]*(wd*B/D)*(min(X,sd)/wd)^a

So letting a = 0.8, wd = 1, sd = 10, D = 3370181.7992778, X = 1.5D you should get
9.42407*10^-6
Rather than 9.08*10^-6 as in the site.

He's including the 3% fee, which makes it closer. When are you applying the fee, kano?
Sorry, I included the fee too, ETA.
The calculator simply uses the (1-a) factor rather than the [(1-a)/(1-a*wd^(1-a)*X^(a-1))] factor.
donator
Activity: 2058
Merit: 1007
Poor impulse control.
Meanwhile ... Smiley
Rather than typing numbers into my desktop calculator over and over ... I wrote this:
http://tradebtc.net/potcalc.php

The defaults (which you can change before you press 'Calculate' are a=0.8 Cap=1.5 Fee=3%)
I'll change the defaults to whatever Graet sets them to ... when I notice they've changed Smiley
I don't think you used the constant factor that corrects for share capping.
Yeah I was applying it to share count Tongue
I've changed it to cap the share value ... hopefully that's correct.
I'm not sure we're on the same page, the formula for payout is

[(1-a)/(1-a*wd^(1-a)*X^(a-1))]*(wd*B/D)*(min(X,sd)/wd)^a

So letting a = 0.8, wd = 1, sd = 10, D = 3370181.7992778, X = 1.5D you should get
9.42407*10^-6
Rather than 9.08*10^-6 as in the site.

He's including the 3% fee, which makes it closer. When are you applying the fee, kano?
donator
Activity: 2058
Merit: 1054
Meanwhile ... Smiley
Rather than typing numbers into my desktop calculator over and over ... I wrote this:
http://tradebtc.net/potcalc.php

The defaults (which you can change before you press 'Calculate' are a=0.8 Cap=1.5 Fee=3%)
I'll change the defaults to whatever Graet sets them to ... when I notice they've changed Smiley
I don't think you used the constant factor that corrects for share capping.
Yeah I was applying it to share count Tongue
I've changed it to cap the share value ... hopefully that's correct.
I'm not sure we're on the same page, the formula for payout is

[(1-a)/(1-a*wd^(1-a)*X^(a-1))]*(wd*B/D)*(min(X,sd)/wd)^a

So letting a = 0.8, wd = 1, sd = 10, D = 3370181.7992778, X = 1.5D, B = 24.25 you should get
9.42407*10^-6
Rather than 9.08*10^-6 as in the site.
legendary
Activity: 4592
Merit: 1851
Linux since 1997 RedHat 4
Meanwhile ... Smiley
Rather than typing numbers into my desktop calculator over and over ... I wrote this:
http://tradebtc.net/potcalc.php

The defaults (which you can change before you press 'Calculate' are a=0.8 Cap=1.5 Fee=3%)
I'll change the defaults to whatever Graet sets them to ... when I notice they've changed Smiley
I don't think you used the constant factor that corrects for share capping.
Yeah I was applying it to share count Tongue
I've changed it to cap the share value ... hopefully that's correct.
donator
Activity: 2058
Merit: 1054
Meanwhile ... Smiley
Rather than typing numbers into my desktop calculator over and over ... I wrote this:
http://tradebtc.net/potcalc.php

The defaults (which you can change before you press 'Calculate' are a=0.8 Cap=1.5 Fee=3%)
I'll change the defaults to whatever Graet sets them to ... when I notice they've changed Smiley
I don't think you used the constant factor that corrects for share capping.
legendary
Activity: 4592
Merit: 1851
Linux since 1997 RedHat 4
Meanwhile ... Smiley
Rather than typing numbers into my desktop calculator over and over ... I wrote this:
http://tradebtc.net/potcalc.php

The defaults (which you can change before you press 'Calculate' are a=0.8 Cap=1.5 Fee=3%)
I'll change the defaults to whatever Graet sets them to ... when I notice they've changed Smiley
donator
Activity: 2058
Merit: 1007
Poor impulse control.
Wlog we assume wd*B/D=1. The pdf of sd for sd>=wd is wd/sd^2 and the payout is (sd/wd)^a. So without a cap and without a constant factor, the expected payout is

\int_{wd}^{\infty}(wd/sd^2)(sd/wd)^a\ sd

This is 1/(1-a), and thus we need a constant factor of (1-a) to make this equal to 1.

With sd capped to X the integral becomes

\int_{wd}^{X}(wd/sd^2)(sd/wd)^a\ sd + \int_X^{\infty}(wd/sd^2)(X/wd)^a\ sd

The second term is (X/wd)^(a-1) and the first term (by subtracting the primitive function at the endpoints) is (1-wd^(1-a)*X^(a-1))/(1-a). Add to get [(1-a*wd^(1-a)*X^(a-1))/(1-a)], so the constant term should be [(1-a)/(1-a*wd^(1-a)*X^(a-1))].


Thanks for that. Nice clear way of showing how to think about the problem.
Pages:
Jump to: