Stephen Gornick
Legendary
Offline
Activity: 1988


June 25, 2012, 10:51:53 PM 

Heh ... don't look now but SatoshiDICE just got even easier to use:  http://blockchain.info/walletFrom your Blockchain.info wallet, click Send Money, then on the left is "Games  Satoshi Dice". The balance will update with your win or loss, or you can click Transactions. It will update in realtime.






Advertised sites are not endorsed by the Bitcoin Forum. They may be unsafe, untrustworthy, or illegal in your jurisdiction. Advertise here.


evoorhees
Legendary
Offline
Activity: 994
Democracy is the original 51% attack


June 26, 2012, 06:24:31 PM 

OMG I just saw the Blockchain.info auto plugin for SatoshiDice!!! I didn't even request that he add it, but now I will have to send him money and hookers, cause that is the coolest thing ever. Way more fun that just playing from a desktop client.




runlinux


June 26, 2012, 06:26:14 PM 

omg... why... now its easier for me to blow money... ungh... trying to save up for my island!




ErebusBat


June 26, 2012, 06:26:46 PM 

OMG I just saw the Blockchain.info auto plugin for SatoshiDice!!! I didn't even request that he add it, but now I will have to send him money and hookers, cause that is the coolest thing ever. Way more fun that just playing from a desktop client. Hrm umm.... I suggested it... ya that is right it was my idea....




dooglus
Legendary
Offline
Activity: 1988


June 26, 2012, 06:56:52 PM 

Sum typoes: All transactions in that block and befor are included in the jackpot.
before So a players own winning bet gets included in their jackpot.
player's Winners are not payed until their bet transaction has at least 2 confirmations
'paid' is more commonly accepted as correct It is to the winners advantage to wait for confirmation
winner's So for the progress games we involve a trustable third party
progressive? allowing interested parties to eaves drop on that message stream
eavesdrop is a single word




sgravina


June 27, 2012, 12:23:04 AM 

Since we are correcting stuff the following text from http://www.satoshidice.com/bits.php is not correct: If you lose, the return is one half the house edge.
In actuality you are returning .005 * the bet. The amount of your winning payout is the bet times: ((1h) 0.01*(1p)/2)/p where h = house edge, p = probability of winning. According to your rules it should be: ((1h) h*(1p)/2)/p Since your house edge started out at 1% your winning payout was initially correct. But now it isn't. I figured this out by creating a SatoshiDice simulation in Excel and then in Matlab. It made a very nice demonstration of why I lost so much money playing SatoshiDice. As I already knew and everyone else should know also, SatoshiDice will make you poorer. Interestingly there is a significant probability that SatoshiDice itself will loose a lot of money, despite it's house edge. I think that is why they have had to decrease bet limits and increase the house edge.




dooglus
Legendary
Offline
Activity: 1988


June 27, 2012, 12:59:59 AM 

Interestingly there is a significant probability that SatoshiDice itself will loose a lot of money, despite it's house edge. I think that is why they have had to decrease bet limits and increase the house edge.
Look how they lost over 3000 BTC ($15,000) in a 7 day period (June 2nd to June 9th): It looks to me like a large amount of those losses stemmed from a single bet: http://blockchain.info/tree/7470279Follow that bet, its winnings, and the various change outputs, and you find a sequence of large bets with unusually high win rates. I found 45 "lessthan 32000" bets with 26/45 winning (58%) when it should be 48%, and 19 "lessthan 48000" bets with 15/19 winning (79%) when it should be 73%. The bets are mostly 50 or 100 BTC each. My knowledge of statistics isn't good enough to work out whether this is reasonable variance, or whether it's suspicious but I suspect that I don't have enough data points to draw any safe conclusions.




nimda


June 27, 2012, 01:26:45 AM 

Interestingly there is a significant probability that SatoshiDice itself will loose a lot of money, despite it's house edge. I think that is why they have had to decrease bet limits and increase the house edge.
Look how they lost over 3000 BTC ($15,000) in a 7 day period (June 2nd to June 9th): http://i.imgur.com/ru1Gz.pngIt looks to me like a large amount of those losses stemmed from a single bet: http://blockchain.info/tree/7470279 Follow that bet, its winnings, and the various change outputs, and you find a sequence of large bets with unusually high win rates. I found 45 "lessthan 32000" bets with 26/45 winning (58%) when it should be 48%, and 19 "lessthan 48000" bets with 15/19 winning (79%) when it should be 73%. The bets are mostly 50 or 100 BTC each. My knowledge of statistics isn't good enough to work out whether this is reasonable variance, or whether it's suspicious but I suspect that I don't have enough data points to draw any safe conclusions. For situations like these it's hard to set a threshold of probability. For example, according to our current understanding of quantum mechanics, it is possible but highly unlikely that Ella Fitzgerald will appear next to you right now. There's an overwhelmingly larger yet still incredibly small possibility that most of the particles in my nose will "teleport" somewhere else. I ignore this possibility in my everyday life and assume that my nose will continue to exist in its current form because the probability of mass localized longdistance quantum teleportation is so small. To find the probability of "beating the odds," we can use a binomial probability distribution function. It's basically Bernoulli's experiment. The probability of winning exactly 26/45 bets on 48% odds is binompdf(45, .48, 26) = 5.05%, which is fairly reasonable. The probability of winning at least 26/45 bets on 48% odds is binompdf(45, .48, 26)+binompdf(45, .48, 27)+...+binompdf(45, .48, 45) = 12.23%, which is between 1/8 and 1/9 odds. This makes sense when you consider that 48% is close to 1/2, and 26/45 is close to 1/2 as well. For the second data set: 18.35% for exactly 15/19 wins on 73% odds 38.71% for at least 15/19 wins on 73% odds I am happy to post more technical details, such as the source code of the program I used to sum up the density functions or the output answers to more decimal places, but I suspect it's not really wanted. TL;DR no, it's not suspicious, just unlucky that it happened on such large bets.

I recommend asking me for a signature from my GPG key before doing a trade. I will NEVER deny such a request.



dooglus
Legendary
Offline
Activity: 1988


June 27, 2012, 01:49:42 AM 

it is possible but highly unlikely that Ella Fitzgerald will appear next to you right now
She didn't. It was quite a relief because she's been dead since 1996. To find the probability of "beating the odds," we can use a binomial probability distributionfunction.
Oh, I was close then. I found myself at http://en.wikipedia.org/wiki/Binomial_distribution yesterday when I was trying to work out how to do this, but didn't have enough confidence that I had the right distribution to proceed. Thanks a lot for doing the maths. The probability of winning at least 26/45 bets on 48% odds is 12.23%
38.71% for at least 15/19 wins on 73% odds
So a combined probability of 38.71 * 12.23 / 100 = 4.73%, or about 1 in 21. it's not suspicious, just unlucky that it happened on such large bets.
If you look at blockexplorer you see that at 20120601 08:34:55 he received 100 BTC which he lost (lose 50, win 60, lose 50, lose 71) and then at 20120601 21:27:22 he received the 100 BTC which went on to win lots. It strikes me as unlikely that he would lose the whole of his first bankroll if there was anything suspicious going on. Perhaps this was just poor bankroll management on SatoshiDice's part. The only refunded bet I saw was 3 days after this guy started on his winning streak.




nimda


June 27, 2012, 02:03:54 AM 

Warning: heading more offtopic into probability theories. Oh, I was close then. I found myself at http://en.wikipedia.org/wiki/Binomial_distribution yesterday when I was trying to work out how to do this, but didn't have enough confidence that I had the right distribution to proceed. Thanks a lot for doing the maths. You were more than close. That formula is exactly what you need. In general, an understanding of binomial probability or "Bernoulli trials" will get you a long way. If n = number of trials r = number of specific events you wish to obtain p = probability that the event will occur then all you need is Where is the binomial coefficient, also denoted C(n, k), _{n}C _{k}, or ^{n}C _{k}(yes, I did steal the part beginning with "Where" from Wikipedia; it's nice, easy, and compact)

I recommend asking me for a signature from my GPG key before doing a trade. I will NEVER deny such a request.



ThePok


June 27, 2012, 11:44:44 AM 

i dont understand the pogressiv pots: it says: Llama Dance 1DiCeLLhySbEQxWMGo84NbGbZsYwMjpBYs 1.0000 1 One in 16 1.500% 3.1916 so i pay 1 btc and one in 16 times i win 3.19 BTC....whos ever gone play that??? Its only profitable if Pot is at 16 BTC but it will never get there?? (Only if silly people play?!?) Whats the Math behind this




nimda


June 27, 2012, 01:15:06 PM 

The math is that SatoshiDice carries a house edge, so it's a negativesum game for the gambler.

I recommend asking me for a signature from my GPG key before doing a trade. I will NEVER deny such a request.




Otoh
Donator
Legendary
Offline
Activity: 1918


June 27, 2012, 03:01:17 PM 

Heh ... don't look now but SatoshiDICE just got even easier to use:  http://blockchain.info/walletFrom your Blockchain.info wallet, click Send Money, then on the left is "Games  Satoshi Dice". The balance will update with your win or loss, or you can click Transactions. It will update in realtime. I've refreshed & relogged in to my wallet but can't see any games link on the Send Money page, just this as before...

Node40.com is a leader in DASH hosting, dedicated exclusively to fully managed masternode hosting. Professional, organized, and responsive. I have many dozens of nodes with them. BTC = $c² BTC = 1otohotohMoQoxHuxLBveQiZcV3Pji3Tc DASH, Digital Cash = www.dash.org CHARITY  MY REP  DICE



Fredyy


June 27, 2012, 03:49:55 PM 





ErebusBat


June 27, 2012, 04:16:54 PM 

I've refreshed & relogged in to my wallet but can't see any games link on the Send Money page, just this as before...
Otoh, Blockchain.info is not affiliated with SatoshiDICE, you should post in their thread: https://bitcointalk.org/index.php?topic=40264.0




Otoh
Donator
Legendary
Offline
Activity: 1918


June 27, 2012, 05:50:52 PM 

I've refreshed & relogged in to my wallet but can't see any games link on the Send Money page, just this as before...
Otoh, Blockchain.info is not affiliated with SatoshiDICE, you should post in their thread: https://bitcointalk.org/index.php?topic=40264.0Thanks, have done

Node40.com is a leader in DASH hosting, dedicated exclusively to fully managed masternode hosting. Professional, organized, and responsive. I have many dozens of nodes with them. BTC = $c² BTC = 1otohotohMoQoxHuxLBveQiZcV3Pji3Tc DASH, Digital Cash = www.dash.org CHARITY  MY REP  DICE



dooglus
Legendary
Offline
Activity: 1988


June 28, 2012, 02:55:47 AM 

I think I just saw a bug in the Progressive bet detail page. It says the bet won, but that the payout was zero. Also the bet tx and the payment tx are the same: Then, a few seconds later, it fixed itself:




dooglus
Legendary
Offline
Activity: 1988


June 28, 2012, 04:23:25 AM 

so i pay 1 btc and one in 16 times i win 3.19 BTC....whos ever gone play that??? Its only profitable if Pot is at 16 BTC but it will never get there?? (Only if silly people play?!?) Whats the Math behind this Talking of silly people, look at this!Someone bet 100 BTC to win that 3.19 BTC. 70% of the 100 BTC gets added to the available pot, so they win a total of 73.19. Actually, they won it 4 times, and split the 73.19 with themselves 4 times. Still a loss of 27 BTC though...




mem


June 28, 2012, 05:13:59 AM 

so i pay 1 btc and one in 16 times i win 3.19 BTC....whos ever gone play that??? Its only profitable if Pot is at 16 BTC but it will never get there?? (Only if silly people play?!?) Whats the Math behind this Talking of silly people, look at this!Someone bet 100 BTC to win that 3.19 BTC. 70% of the 100 BTC gets added to the available pot, so they win a total of 73.19. Actually, they won it 4 times, and split the 73.19 with themselves 4 times. Still a loss of 27 BTC though... satoshidice due to the 0 confirmations makes for an awesome bitcoin laundering system




