Deepsum: a Novel Algorithm for Bitcoin Dice Roulette

[Biomass]

Deepsum: a Novel Algorithm for Bitcoin Dice Roulette - Safe, Customizable and Profitable

Corresponding author: Idea Nakamoto (DeepOnion team). Email: forthelastwords@gmail.com

Abstract:

In this context, a novel progressive strategy of roulette, regarding the house edge redemption, is first time introduced by DeepOnion team. Geometric progression is alternated with arithmetic progression to guarantee a slower bet size increasing. Comparison between DeepSum and classical strategies is briefly discussed. As a key point, within the increasing size of bet to redeem for the house edge payment, the law of large numbers is the key to expect with a sufficiently long run, a final winning result is reached. The algorithm has several customizable factors with its code written in Lua programming language to run with Seuntjie Dicebot. Simulation tests were carried out and showed that the strategy can help players gain tens or hundreds of percentage of profit. This is the safest roulette strategy that ever known, as a present of DeepOnion team to the cryptocurrency community.

1. Introduction

Roulette gambling is a fantasy dream of the human history. Millions of people even sold their houses to play roulette with an illusion that they could win with large enough finance balances in long runs. But most of time, no one could do it. It was explained as the rule of house edge exhaust their pockets.

Many progressive strategies were introduced to make winners, including Martingale, D'Alembert, La Bouchere, Fibonacci, etc. But none of them could be safe enough for a long run. The profit rate is low, while the balance must be very large.

DeepSum is introduced here as the safest algorithm so far. DeepSum has several customizable factors. Many tests were run by simulation on Seuntjie Dicebot via Primedice server. Millions of bets were carried out successfully with steadily increasing balance.



2. DeepSum algorithm

Principle: instead of increasing bet size after each bet, which makes a differential geometric progression by bet, DeepSum runs a group of bets without increasing bet size. The number of bets in a group is abbreviated as "g" in the Seuntjie Dicebot Lua scrypt enclosed.

During the betting, if the balance increases as a positive profit is obtained, the bet is reset to initial value. Otherwise, after g bets, a sum of loss is added to the next bets, which means only a win then can bring back all the loss.

There are two reasons DeepSum is much safer than the classical Martigale:

- A large number g of bets statistically helps gainning probability of winning (say herein 50%) more safely than only a single bet.
- A sum of loss after g bets (arithmetic progression) can be statistically much smaller than an increasing bet value after every each bet in Martingale by geometric progression.

Although Fibonacci strategy is not based on a geometric progression, the increasing rate of bet size is also very tremendous. Similarly, although Labouchere strategy is a smarter arrangement of betting, using this strategy, one can quickly end up with a very large bet size exceeding the balance.

Comparing with D'Alembert strategy ,which is also based on arithmetic progression, DeepSum adjusts the house edge redemption after each group of g bets. With D'Alembert strategy, the increase of one unit of bet size after each bet will not be able to sufficiently redeem for the house edge payment when the size is accumulated that a house edge exceeds the unit increased.

DeepSum procedure explanation:

Step 1: Place a bet, say 1 unit. If win, reset the bet for the next bet.
Step 2: If lose, repeat the bet. The repeating is for g times. Once the profit is obtained, the bet is reset back to Step 1.
Step 3: After g bets, the bet size increases to be equal to all the previous loss (which includes house edge payment).
Step 4: looping from Step 1...

The betting is looped until a profit is obtained.

For an illustrative example (regardless house edge):

For an illustrative example (regardless house edge):
g = 10
1 (lose)
1 (lose)
1 (win)
1 (lose)
1 (lose)
1 (win)
1 (lose)
1 (lose)
1 (lose)
1 (win)
(Finish g bets, total loss = 4)
4 (lose)
4 (win)
4 (win) (reset)

If the probability of win is set 50%, a fallacy can make one expect that finally, the number of lose would approach the number of win and the bet size would not increase too high. When the number of wins exceeds the number of lose only 1 bet, the profit is obtained.

With customized factors, DeepSum can guarantee either high profit rate, or slow increasing bet size. In the Seuntjie Dicebot programming script enclosed with this context, the following factors can be manually input.

g: number of bets in a repeating group.
ratio: the number of divided bet unit from the balance.
timecount: the order of the current bet in a group g.
level: the number of group g that has stepped up so far.
houseedge: houseedge (%) of the roulette site.

The following values are monitored (in Console section of Seuntjie Dicebot)
levelmax: the highest levels in the history of the session.
bets: number of bets been placed of the session.
betmax: the highest bet size in the history of the session.
balancemax: the highest balance reached in the session.

3. The number "g"
Statistically, when the chance of win a bet is 50%, one can expect to get g/2 number of win and g/2 number of lose.

When g=100 and house edge =2%, statistically after 100 bets, the average losing is (100/2)×2%×Bet Unit = 1 Bet Unit. This implies that within 100 bets of the same level, if the number of win exceeds the number of lose only 1 bet, a positive profit is gained.

While a level is not enough for a positive gain, a larger g (than 100), the higher number of wins exceeding the lose must be. In contrast, the bet size in the next level is smaller. And vice versa.

4. Seuntjie Dicebot DeepSum (ver.1) Script

Details of the bot can be found here: https://bot.seuntjie.com/ (Seuntjies DiceBot)

The Code:

-- By Idea Nakamoto (DeepOnion team), email: forthelastwords@gmail.com
-- federal link for Prime Dice betting site: https://primedice.com/?c=ea341ae45b (Primedice: Bitcoin Gambling - The Original Crypto Dice Game)
-- Donation for further development:
-- with Bitcoin: 1HmnV8KokBJGAKCJLVQ7F3hxnnvsEQQ1pz
-- with Doge: DHtiEU1PYh2AnKBnDiCUrCBXsnYHtSUytp
--- with DeepOnion: DZ9kKm984WNATbsyixJHybVwRay5bsMdbF

chance    = 50 --sets your chance for placing a bet
ratio     = 100000  -- set the number of bet unit divided from your balance.
base      = balance/ratio -- sets your first bet.
nextbet   = base;
timecount = 0;
inibal    = balance;
maxbal    = balance;
g  = 10;
houseedge = 2 --(%)
level     = 0;
levelmax  = level;
betva     = base;
bets      = 0;

function dobet()
bets+=1;

local p1  = "COUNT=";
local p2  = timecount;
local p3  = "LEVEL=";
local p4  = level;
local p5  = " | ";
local p6  = "LEVELMAX=";
local p7  = levelmax;
local p8  = "BET=";
local p9  = previousbet;
local p10 = "BALANCE=";
local p11 = balance;
local p12 = "BETMAX=";
local p13 = betva;
local p14 = "BALANCEMAX=";
local p15 = maxbal;
local p16 = "BETS=";
local p17 = bets;
print(p1..p2..p5..p3..p4..p5..p8..p9..p5..p10..p11..p5..p6..p7..p5..p12..p13..p5..p14 ..p15..p5..p16..p17);

if levelmax<level then levelmax = level; end;
if betva<previousbet then betva=previousbet end;
if maxbal<balance then maxbal=balance end;

timecount+=1;

if (inibal>balance) then
   if timecount<g+1 then nextbet = previousbet;  end;
   if timecount==g+1 then
   level+=1;
   timecount = 0;
   nextbet   = inibal-balance; end;
end;

if balance>=inibal or previousbet==0 then
   nextbet   = balance/ratio;
   timecount = 0;
   inibal    = balance;
   level     = 0;
   --resetseed();          (optional)
   -- print("DONE"); (optional)
end;
end

4. Conclusion
DeepSum is tested on Seuntjie Dicebot at Primedice server for millions of bets. Tens or hundreds of percentage of profit were gained. However, risk of lose is certainly the contrast of profit. The customizeability of DeepSum is hoped to bring joy and discovery to bitcoin dice roulette players. And this is a present from DeepOnion team to the crypto-community as a message of "the coming big next thing"
The above context is only for its initial stage. Further versions of DeepSum can be discussed herein or via private contact with the author

(Idea Nakamoto, email: forthelastwords@gmail.com)

Donation for further development of the game.

Bitcoin (BTC): 1HmnV8KokBJGAKCJLVQ7F3hxnnvsEQQ1pz



DOGE : DHtiEU1PYh2AnKBnDiCUrCBXsnYHtSUytp



DeepOnion (ONION): DZ9kKm984WNATbsyixJHybVwRay5bsMdbF





--------------------------------------------

Some testing results

[USBitcoinServices.Com]

It would be very interesting to see this project grow and bring more people to support it with new development, the game industry is looking for new games.

[Biomass]

A sustainable profit + safety is better than high profit rate with risk.
If one day, you increase only 5% of your balance, after 365 days, you can get:

1.05^365 = 54,211,841 x your initial balance  

Or couse this is too much dreamy. However, try the above scrypt with a dicebot supporting Lua language, you may have a very exciting time of playing.

[Nico1994]

This project looks like a great idea! Very interesting for all the people who is into the gambling world.

Looking forward to see a full version of this product. 

[nakamura12]

As what I understand in the statement. The bot you are trying to introduce is a bit the same as the martingale. Not that I don't like bots or like bots but many have use bots and never be heard if they earn good profit. What I mostly heard is that the bot made their bets easy and less time consuming just by changing the bettings that you want the bot to run.

[Biomass]

As what I understand in the statement. The bot you are trying to introduce is a bit the same as the martingale. Not that I don't like bots or like bots but many have use bots and never be heard if they earn good profit. What I mostly heard is that the bot made their bets easy and less time consuming just by changing the bettings that you want the bot to run.

Hi Nakamura san,

Almost web buit-in bots are only for Martingale strategy as you say. But there are professional bots like Seuntjie, which let you customize not only Martingale, but also other strategies like Labouchere, d'lembert, Piroli, Fibonacci ... It is must more interesting and if you use the bot well, you can get profits in short-terms. Beside the built-in strategies, such bots also allow you to setup your own strategies with scripts. In the context above, I created a Lua script for DeepSum strategy. You just copy and paste it to Seuntjie Dicebot in Programming Mode and enjoy.

DeepSum helped me to earn some hundreds of $USD during the Covid-19 off-work days. Try it. This game is very interesting.

[nakamura12]

Well let's see. I been busy with modules so I don't have much time to test your bot but I have something to ask about what sites that you can use this bot?. As the thread title said that it's for bitcoin dice roulette. I have read and learn that some bots can be customized as you have mentioned and I did know about it. I also found that some bots are only applicable for the site that it based. Example: 777coin bot for 777coin site only.

[lightlord]

No matter how you dress it up, from fancy betting systems, to innovative betting strategies, to regular Martingale, D'Alembert, La Bouchere, Fibonacci, ...
Each bet you make in the long-run , say on a one zero roulette table, is on average 0.973x of what ever bet you place. If you wager 100 Coins, you can expect to lose 2.7 coins on average
in the long-run.

But then again, we are talking out to long-run infinity. House edge is just a cut, and gambling is a hopes of hitting it big, or getting lucky. House's fee is a cost for having a risk.
But as always, only play what you can afford to lose, and nothing wrong with having fun 

[Biomass]

Well let's see. I been busy with modules so I don't have much time to test your bot but I have something to ask about what sites that you can use this bot?. As the thread title said that it's for bitcoin dice roulette. I have read and learn that some bots can be customized as you have mentioned and I did know about it. I also found that some bots are only applicable for the site that it based. Example: 777coin bot for 777coin site only.

Some sites are very fast and some sites are very slow. But fast sites always require a minimum size of bet which is more than 1/100,000,000 of the currency.

You are right all, gambling is not for sustainable profit. Gambles are for entertainment. The feeling of conquering the house is very fantastic. Never play with the money you cannot afford losing.

[fiulpro]

Well let's see. I been busy with modules so I don't have much time to test your bot but I have something to ask about what sites that you can use this bot?. As the thread title said that it's for bitcoin dice roulette. I have read and learn that some bots can be customized as you have mentioned and I did know about it. I also found that some bots are only applicable for the site that it based. Example: 777coin bot for 777coin site only.

Some sites are very fast and some sites are very slow. But fast sites always require a minimum size of bet which is more than 1/100,000,000 of the currency.

You are right all, gambling is not for sustainable profit. Gambles are for entertainment. The feeling of conquering the house is very fantastic. Never play with the money you cannot afford losing.

Despite this there are many people in this world who do use gambling as an income source .. there are literally people known as the Professional Gamblers those people have all the strategies to earn millions !!

The person who discovered card counting was a brilliant mathematician and his techniques are still being used today.

-It is completely obvious that you should not gamble with the money you cannot afford to loose since it would not only be difficult for the long run but also cause addiction in many cases.

-one always needs to be responsible no doubt.

-*-

As far as the algorithm is concerned , did someone try it ?
By the way if you are using Onion , you should know that there are many places where it's banned .. but at many places it's considered a good thing to protect privacy as a whole.

[Swordsoffreedom]

~
You are right all, gambling is not for sustainable profit. Gambles are for entertainment. The feeling of conquering the house is very fantastic. Never play with the money you cannot afford losing.

Of course, There is no doubt that gambling is for entertainment, But the thing to remember is that the issue of victory and defeat is no less important here. Most of the time if you keep losing gambling it will definitely not be a happy thing anymore.

If I judge by myself, If I lose the gamble most of the time, that game will be less enjoyable for me. 

[Norkys.2020]

I think the idea is not bad, but there is something that does not fit me, what is the probability of profit? If I am a gambler with 0.1BTC I would like to at least win 10% playing, and I do not see that you have implemented Artificial Intelligence, it must be remembered that you are trying to beat the provably fair for the most part.

[michellee]

I think the idea is not bad, but there is something that does not fit me, what is the probability of profit? If I am a gambler with 0.1BTC I would like to at least win 10% playing, and I do not see that you have implemented Artificial Intelligence, it must be remembered that you are trying to beat the provably fair for the most part.

If you think like that, you need to think that how big your chance to win 10% because we are in the gambling games, which we know that the probability of winning will not be too big. Maybe for less than 5% is acceptable to win, but to rely on 10%, it is very hard to get it on the gambling. I don't know much about using software to predict the winning percentage because I think the gambling site will have a different algorithm that they use in their games. That will not give us more chances to win, but if you think that using that software can increase your chance, you can try it yourself, but you need to remember only use the money you can afford.

[leea-1334]

A sustainable profit + safety is better than high profit rate with risk.
If one day, you increase only 5% of your balance, after 365 days, you can get:

1.05^365 = 54,211,841 x your initial balance  

Or couse this is too much dreamy. However, try the above scrypt with a dicebot supporting Lua language, you may have a very exciting time of playing.

If you play dice often enough even at the lowest possible starting bet of 1 satoshi and on a bankroll of 1 BTC you will probably see that eventually you will bust. It seems impossible at 1 BTC because it is roughly 1 in a billion bets to hit the losing streak (much less because of doubling and of house edge) but variance is variance:)

[LUCKMCFLY]

OP has no proof of the use of it? For the use of this type of bots with how much balance in Bitcoin do you recommend using? since a person can use it and see the results after certain numbers of launches.

[Biomass]

A sustainable profit + safety is better than high profit rate with risk.
If one day, you increase only 5% of your balance, after 365 days, you can get:

1.05^365 = 54,211,841 x your initial balance  

Or couse this is too much dreamy. However, try the above scrypt with a dicebot supporting Lua language, you may have a very exciting time of playing.

If you play dice often enough even at the lowest possible starting bet of 1 satoshi and on a bankroll of 1 BTC you will probably see that eventually you will bust. It seems impossible at 1 BTC because it is roughly 1 in a billion bets to hit the losing streak (much less because of doubling and of house edge) but variance is variance:)

I never think we can make sustainable profit with gambling, due to the house edge rule. But I did get some profit on Primedice with this strategy. I started with 3000 Doge then got finally about 32,000. The longest time I let my bot run on a virtual machine (Google Cloud) was about 3 weeks, about 2 millions bets, if I remember correctly.

This DeepSum strategy make the progression rate very close to be sufficient to gain positive profit. And you can customize the factors to either increase the profit rate (vs. high risk) or safety (vs. low profit rate). Martingale or other welknown strategies are not very customizable. In fact, their progression rate is too high.

[Biomass]

OP has no proof of the use of it? For the use of this type of bots with how much balance in Bitcoin do you recommend using? since a person can use it and see the results after certain numbers of launches.

The factor "ratio" is the number of base bets that our balance can be. I mean that if your base bet size is 1 unit, your balance is equal to "ratio" unit. Say ratio = 1000000 as I recommended, you can play on Primedice with 0.1 Litecoin and a base bet is 0.00000001 Litoshi.

For a experienced gambler, I consider this game for fun, never for money. So should you all. But if you are "professional" and wise, you still can earn money from this strategy. It is safe and profitable in short and mid terms.

[michellee]

The factor "ratio" is the number of base bets that our balance can be. I mean that if your base bet size is 1 unit, your balance is equal to "ratio" unit. Say ratio = 1000000 as I recommended, you can play on Primedice with 0.1 Litecoin and a base bet is 0.00000001 Litoshi.

For a experienced gambler, I consider this game for fun, never for money. So should you all. But if you are "professional" and wise, you still can earn money from this strategy. It is safe and profitable in short and mid terms.

If there is no minimum amount to bet, I guess people will try to use the smallest amount, such as 0.00000001 coins, to gamble. That is what I will do if I want to play gambling. But some gambling sites have a minimum amount that we can use to bets, so we need to use that to gamble.

If you consider using gambling for a fun thing, you can select anything you want, including using many methods that you know to see how good your luck. We can earn money from all strategies, but that will depend on your luck because gambling needs to have the luck to win any gambling games.

[BlackFor3st]

If it is indeed profitable in the long run, I am sure that many will be interested to it but may I know if it is already available?
I mean for free or you will be asking for some fee in order to use it?

I am having a difficulty in understanding the algorithm of your DeepSum as I am not that very techie person but I will be looking forward
that I can also use it and test it for myself whether it is really helpful or not.

[wxa7115]

As what I understand in the statement. The bot you are trying to introduce is a bit the same as the martingale. Not that I don't like bots or like bots but many have use bots and never be heard if they earn good profit. What I mostly heard is that the bot made their bets easy and less time consuming just by changing the bettings that you want the bot to run.

It has been demonstrated many times in the past that betting patterns no matter how advanced do not work, this just seems like a slower martingale in which case it will take you long to be busted by the casino but eventually it is going to happen, it is better to just accept the house edge is there and bet responsibly, it is true that variance can allow us to win from time to time and this gives the impression that the roulette could be beaten with the application of patterns but this is not true, besides the kind of patterns described on the OP can be applied successfully by a bot but if a person tries to do this and losses several times in a row most likely they will revert to martingale negating the strong points of this system.