฿0.00 Low Hash Lotto Number 29 by Dabs. Provably Fair. Weekly Draws. ฿

[Dabs]

Edit: Dabs Lotto Number 3 has a winner. See later posts for next game.

[🏰 TradeFortress 🏰]

I think you should only retain 20% of the jackpot that was paid, because right now you are retaining 24%. (You're also retaining the retained parts).

I would be interested in betting but 1/4 retain is way too high for my tastes.

[dooglus]

I think you should only retain 20% of the jackpot that was paid, because right now you are retaining 24%. (You're also retaining the retained parts).

I would be interested in betting but 1/4 retain is way too high for my tastes.

Good thing this game is provably fair or questions might be asked about my suspicious win... 

It took me a while to see where you got that 24% from:

If 0.6 is bet in this round, that will make a total pot of 0.6 + 0.12 = 0.72.  20% = 0.144 will be retained, which is 24% of the 0.6 that was bet.  80% = 0.576 will be paid out.

TradeFortress is suggesting that only 20% of the newly bet money should be retained, with the previously retained funds going to the winner(s), so 0.12 + 0.8*0.6 = 0.6 would be paid out, and 0.12 would be retained.

In effect, in TradeFortress' scheme, if the number of tickets sold remained constant, the retained amount would also remain constant, and all received funds would be paid out (except for the house take, if any).  That seems fair to me.  In a steady state game we don't want the retained amount to grow.  Although in the long run, the retained amount would only be 25% of the bet amount.

Code:

>>> retain=20.0; print 1 / (1 / retain - 0.01)
25.0

So I don't know what's best.  It doesn't seem like a big difference to me, 20% or 25%.

[Dabs]

Ok. So, what do you guys suggest? I'd like to keep the rules wordings as simple as possible. Maybe I should change the 80/20 to a different number, like 85/15 ? Would that make the end result (including retaining the retained parts) the same or something? Let me do some quick math on my little spreadsheet to find out what is a good percentage, so I only change the numbers without confusing new players about retaining the retained parts.

I just wanted to start every game with something, instead of nothing.

And, we have one ticket.

Code: (Confirmed Tickets)

transaction                                                      ticket   bitcoin address                     nick
dad9b48d45b3e7c884a11cf4ce6bdbacc133b9a1c91646ec80e5b0958af653b8 00000001 1EVq8bzJXS5VZnfhLToR3bgoreSuotvEKP  unknown

Seems we have a repeat player. But he won't identify himself. That's perfectly alright Mr. Anonymous.

[dooglus]

Ok. So, what do you guys suggest? I'd like to keep the rules wordings as simple as possible. Maybe I should change the 80/20 to a different number, like 85/15 ? Would that make the end result (including retaining the retained parts) the same or something?

Here's how to work out what the retained percentage ends up as if the number of bets is the same each week:

Code:

>>> retain=20.0; print 1 / (1 / retain - 0.01)
25.0

ie. a weekly retain of 20% results grows week by week until it gets arbitrarily close to 25%.

If you wanted to end up retaining 20% each week...

Code:

20 = 1 / (1 / retain - 0.01)
20 * (1 / retain - 0.01) = 1
20/retain - 0.2 = 1
20/retain = 1.2
retain = 20/1.2
retain = 16.666666%

... you'd have to retain 16 and two thirds percent each week.  Which is ugly. 

Here are some alternatives:

Code:

>>> retain=16.0; print 1 / (1 / retain - 0.01)
19.0476190476
>>> retain=16.5; print 1 / (1 / retain - 0.01)
19.7604790419
>>> retain=16.66666; print 1 / (1 / retain - 0.01)
19.9999904
>>> retain=17.0; print 1 / (1 / retain - 0.01)
20.4819277108

I say go with 17% per week, which ends up around 20.5% in the long run.

[Dabs]

Mr. TradeFortress, kindly check my math. I think I would achieve the same goal by changing the percentages to 83.33% / 16.67%.

83.33% to winner
16.67% to next game.

While keeping the rest of the rules the same. The sole ticket buyer I don't think will complain if this rule is implemented, as if he wins, he gets a higher percentage.

*edit* Dooglus posted. Testing out 83% / 17%

Ok. That one works nicer. I didn't use any formula. I just made a simulation of the next 20 games with those parameters. The retained amount about becomes close to 20% as intended. But it changes when bets are not constant, like if there are more this game, or less next game.

All in favor of amending rules for Dabs Lotto Number 3 to 83% / 17%? I will update and also post here.

Due to the weird math, all results will be rounded to the closest satoshi.

[dooglus]

All in favor of amending rules for Dabs Lotto Number 3 to 83% / 17%? I will update and also post here.

Aye.

Due to the weird math, all results will be rounded to the closest satoshi.

Are you attempting to pull a Superman 3 on us?  Spoiler alert: the giant computer you build ends up trying to kill you when you attempt to unplug it.

[Dabs]

All in favor of amending rules for Dabs Lotto Number 3 to 83% / 17%? I will update and also post here.

Aye.

Due to the weird math, all results will be rounded to the closest satoshi.

Are you attempting to pull a Superman 3 on us?  Spoiler alert: the giant computer you build ends up trying to kill you when you attempt to unplug it.

Ok. I need one more "vote" to concur, specifically Mr. TradeFortress, since he was the one that complained.

I don't quite get the Superman 3 reference, although I know about the giant computer. Lots of giant computers tend to kill humanity when humans attempt to unplug them. Skynet (Terminator), Seth (Universal Soldier). Maybe even Eddy (Stealth, the movie about the plane with AI).

Bitcoin currently only allows 0.00000001 as the smallest unit, which most people call a satoshi. The weird math can make all sorts of values come up.

[🏰 TradeFortress 🏰]

I think 16.67% retain would be much better! Not sure why you need my 'vote' through. It should be that if no existing ticket buyer (proper to the official announcement of the rules change) objects, then rules can be changed. But enough yaddering about this, I sent 0.03 BTC so I should get 4 tickets

[Stunna]

Bought 5, good luck all   

f575ea840cc9e23d39bcd07b3b2aeadd10f0c761668227c3c7d7d67a0d7883e0

[NLNico]

Bought 3+1 tickets.

Lower % to next sounds good to me.

Good luck

[Dabs]

Code: (Confirmed Tickets)

transaction                                                      ticket   bitcoin address                     nick
dad9b48d45b3e7c884a11cf4ce6bdbacc133b9a1c91646ec80e5b0958af653b8 00000001 1EVq8bzJXS5VZnfhLToR3bgoreSuotvEKP  unknown
03acdb02f8396bda9db25d3f809a43a545271e3dd5e04e80154bec40e47e3de7 00000001 16H2KaaCkYiEbsNcVaFjbKuEaGj4itWRi9  TradeFortress
03acdb02f8396bda9db25d3f809a43a545271e3dd5e04e80154bec40e47e3de7 00000002 16H2KaaCkYiEbsNcVaFjbKuEaGj4itWRi9  TradeFortress
03acdb02f8396bda9db25d3f809a43a545271e3dd5e04e80154bec40e47e3de7 00000003 16H2KaaCkYiEbsNcVaFjbKuEaGj4itWRi9  TradeFortress
03acdb02f8396bda9db25d3f809a43a545271e3dd5e04e80154bec40e47e3de7 00000004 16H2KaaCkYiEbsNcVaFjbKuEaGj4itWRi9  TradeFortress (Bonus)
f575ea840cc9e23d39bcd07b3b2aeadd10f0c761668227c3c7d7d67a0d7883e0 00000001 1CvZQG9GRfS8DEsDX6AwZ8XJQ8iG1tLnq8  Stunna
f575ea840cc9e23d39bcd07b3b2aeadd10f0c761668227c3c7d7d67a0d7883e0 00000002 1CvZQG9GRfS8DEsDX6AwZ8XJQ8iG1tLnq8  Stunna
f575ea840cc9e23d39bcd07b3b2aeadd10f0c761668227c3c7d7d67a0d7883e0 00000003 1CvZQG9GRfS8DEsDX6AwZ8XJQ8iG1tLnq8  Stunna
f575ea840cc9e23d39bcd07b3b2aeadd10f0c761668227c3c7d7d67a0d7883e0 00000004 1CvZQG9GRfS8DEsDX6AwZ8XJQ8iG1tLnq8  Stunna
f575ea840cc9e23d39bcd07b3b2aeadd10f0c761668227c3c7d7d67a0d7883e0 00000005 1CvZQG9GRfS8DEsDX6AwZ8XJQ8iG1tLnq8  Stunna
f575ea840cc9e23d39bcd07b3b2aeadd10f0c761668227c3c7d7d67a0d7883e0 00000006 1CvZQG9GRfS8DEsDX6AwZ8XJQ8iG1tLnq8  Stunna (Bonus)
7d8dc567809ac0d2a4d6eb1be498481b3f35b83bf8a6452b2cd23e1a3b7d620a 00000001 1BjMmTBGsuKmh43uq9pDXMJ87KKSAxs85L  NLNico
7d8dc567809ac0d2a4d6eb1be498481b3f35b83bf8a6452b2cd23e1a3b7d620a 00000002 1BjMmTBGsuKmh43uq9pDXMJ87KKSAxs85L  NLNico
7d8dc567809ac0d2a4d6eb1be498481b3f35b83bf8a6452b2cd23e1a3b7d620a 00000003 1BjMmTBGsuKmh43uq9pDXMJ87KKSAxs85L  NLNico
7d8dc567809ac0d2a4d6eb1be498481b3f35b83bf8a6452b2cd23e1a3b7d620a 00000004 1BjMmTBGsuKmh43uq9pDXMJ87KKSAxs85L  NLNico (Bonus)

Amended Rules to 83.33% / 16.67%.

Short Link to Block Chain:
http://blockchain.info/fb/1lotto3

Less than 85 blocks to go for the Early Bird Bonus!

P.S. I might be out for the weekend. The links and graphic should update in almost real time. (And, in case of a bus, ... but let's not think about that.)

P.P.S. I just noticed, that Stunna sent 0.05 so he got 1 bonus. But if he had sent 0.06, he'd get 2 bonus tickets for a total of 8. But rules are rules. Bonuses are only applied on a per transaction basis. I can't combine two different transactions into one.

[dooglus]

Due to the weird math, all results will be rounded to the closest satoshi.

Are you attempting to pull a Superman 3 on us?  Spoiler alert: the giant computer you build ends up trying to kill you when you attempt to unplug it.

I don't quite get the Superman 3 reference, although I know about the giant computer.

August "Gus" Gorman (Richard Pryor) is an unemployed ne'er-do-well looking for a job. In the opening scene, when he is at a bank for a loan and told that is is no longer eligiable for unemployment benefits, he decides to take a job as a computer programer at Webster Industries. During his first week, Gorman discovers that he has a talent and a natural-born knack for computer programming. Disastified with the low amount of money he receives for his week's pay, he embezzles $85,000 from his new employer's payroll (through an early computer technique known as salami slicing).

ie. he was rounding the satoshis. 

[patricktim]

Bought 2 tickets.

Good luck 

[GueurK]

Quick go check the CoinWin lotto.
It accepts all kind of wallets !

[Dabs]

August "Gus" Gorman (Richard Pryor) is an unemployed ne'er-do-well looking for a job. In the opening scene, when he is at a bank for a loan and told that is is no longer eligiable for unemployment benefits, he decides to take a job as a computer programer at Webster Industries. During his first week, Gorman discovers that he has a talent and a natural-born knack for computer programming. Disastified with the low amount of money he receives for his week's pay, he embezzles $85,000 from his new employer's payroll (through an early computer technique known as salami slicing).

ie. he was rounding the satoshis.  

Oh. But that's done across hundreds of millions of accounts. I don't need to do that since I'm clearly taking a percentage once the total pot exceeds 1 BTC. The thing with bitcoin, there is currently nothing smaller than 1 satoshi. So I can't hide that. I'd just openly say I'll take the 1 satoshi instead.

I think the idea is to take fractions of less than 1 cent, because no one but the bank sees it. People aren't going to miss what they can't see. You can also make it larger, such as a few cents (or whole dollars) of the interest. People aren't going to notice.

Here in the Philippines, one of our highways (the SLEX and NLEX) has an e-pass system (RFID tag for prepaid toll). One time, I noticed a PHP 8 transaction in my monthly statement. Upon calling the customer service hotline, they said they will refund the charge. He slipped and also mentioned that everyone got that same charge, but they were only refunding those who called to complain.

100,000 cars pass that highway system with these e-pass tags. How many are complaining? Few. They can easily get a cool million.

I've seen it taken to a larger extreme in another movie, Firewall (Harrison Ford) but the baddie took $10,000 chunks from 10,000 accounts.

Bought 2 tickets.

Good luck  

patrickim, I don't see your 2 tickets. Did you send to the correct address?

[Dabs]

I just did a UTC Midnight test again. It is now 3 minutes after midnight of UTC.

Here are my results

Here are the strings, generated with randomization from 2013-06-01:

VZWe49s1h1lyMG1R
0Uww0lUNTRwtZoz6
FAfvToE4oO2EM5ES
V25pAEkJSvavaOHt

Timestamp: 2013-06-01 00:01:51 UTC

2013-05-31: e5c3ad1cdc7a92b0e931e04e722d9ea13579cf68

PrimeDice was actually early by a few minutes when I checked at 23:50 UTC

May 31st 2013    Secret: 4URmN28CN46MdSBCCDKCgmf2DvEtGPgSD5xVWJWTQSipGLrO2tiGrHMlnkDcav9t
Hash: 00dd496b47c77ad69a921da5b07b059ac72a5d92296338e1ba76ccd441d64a72

20130601 00:00    4008ea9d82e43e36b53494b70af7b9913d92ab153af9ae04f48d6ad5ed172204
server time: 2013-06-01 00:00:38 UTC

298e27dde49ce0984e237f975f83fadcc458894f89ca671678e275ea9d3c8e7b0849be52f1c3331 c7746b9e411c70ac2a17006ccfc57f518d629c4a502f6131f

SRoulette uses Host Gator. hehe.

2013.05.30 KV7mQGKIchTfbIH7HnWaGoxCe6mU7s4P3EYjueXkblr1evayjgFW0KXsRs91GaNY

SD is the last one that still hasn't revealed it's secret. But it's only been 10 minutes past midnight.

So I will wait a bit and see when they publish the last day of May or 2013-05-31.

All other previously mentioned sites with secrets take a few more hours or another full day before revealing their secrets.

[Dabs]

2013.05.31 JxbOyqYPAjiA3mYFDxUpVIK2yBKfDuU8tda7Y5AUodCUZnYvP3UcDCABICcJWzlR

Time: That's about two and a half hours. Okay. I can still use them.

[Dabs]

Using Visual Basic 6 (Classic) I made a slow, but working program. Using a simulation of 600 character strings to account for the combined concatenated length of the transaction hash, the ticket number, and the 7 secrets, I ran sha-256 in an iterative loop with results as follows.

' interpreted (in IDE)
' 10   seconds for 1,000
' 94   seconds for 10,000
' 230  seconds for 100,000

' compiled
' 1.17  seconds for 1,000 10 BTC
' 11.5  seconds for 10,000 100 BTC
' 142.5 seconds for 100,000 (less than 3 minutes) 1000 BTC
' 5675  seconds for 1,000,000 (1.5 hours) 10,000 BTC

These results are automatically sorted as the array (listbox with 1 million rows, actually) is added to.

The 1 million ticket simulation took that long. It would have been faster to just create the hashes and then sort them all afterwards. As it is, my code was sorting as each item was added into the list. Near the end of the simulation at around 800,000 to 900,000 the speed of computing the hashes and sorting was significantly slowing down.

This is all on my 2009 year model T4400 @ 2.2 Ghz Intel Pentium laptop.

I then redid the simulation using actual previous secrets (that don't necessarily match to any date, I just picked them up from whatever I saw.)

In theory, I can "upgrade" this capability to sort up to 2 billion tickets, and make it significantly faster, so it's not something I'm going to worry about for awhile. 2 billion tickets is about 12 million BTC.

12 million bitcoins
1.2 billion regular tickets
400 million early bird bonus tickets
400 million volume bonus tickets
Total: 2 billion tickets

I can also make this faster by optimizing the sha256 code I'm using, but 3 minutes for 100k tickets works for now.

I will make the final program / calculator available, but for now it's still super beta version, has no error checking, and will certainly crash if you feed it invalid input (such as strings where it expects a number), negative or zero values, etc. And it's in VB, so maybe most people think it's butt-ugly or something.

Actually, all I have is this simulator. For the few tickets I've been getting, I just do all the hashing manually on a spreadsheet. I've since upgraded from hashing each string one by one (back in Dabs Lotto Number 1) using online sha-256 javascript function.

I've also experimented a bit with blocknotify and walletnotify on my bitcoin-qt client.

Using 8 digits as the ticket number, the game allows up to 99 million tickets per transaction. If I stuck to 10 digits then it becomes 9 billion tickets.

Again, not something I am going to worry about now, no matter how sloppy or slow my code is. There are only 11 million bitcoins in existence at this point in time. I'd be happy if someone bought 100 BTC or even 10 BTC worth.

[Dabs]

Another ticket from unknown player.

tx id: 5cf24f1e4e642b6cea06f4f6eca6014b1f3130ed6c6d9a31956fd8627be9d6ac
ticket number: 00000001
bitcoin address: 12Qq3np8EkBZdNjrpmfHU7NaJ7tMpX54qx