Roulette Beating Random by Betting Random

By the way, I just noticed that I had the marty set to 6 instead of 7 in my code. That explains why the win ratio was so low. With a marty of length 7 it should be around 3.6 : 1, so 3.6 winning sessions for every bust. That's still quite a long way from 6:1 though.

Here is the output from the modified program which simulates your system, David. I increased the number of played sessions to 30,000.

I added some code to compute the edge. As you can see, the edge is just about what it should be : -2.7%

Number of busts : 6477
Number of +50 sessions : 23523
Ratio of successful sessions to busts = 3.632
Computed Edge (PL / Total Staked) : -0.0274

Code:

type
  wheel = set of 0..36;
var
  conloss,
  spin       : byte;
  stake,i    : integer;
  numbusts,
  numplus50,
  totstakes,
  profitLoss,
  bank       : longint;
  win,
  bust       : boolean;
  nextbet    : wheel;
  reds       : wheel = [1,3,5,7,9,12,14,16,18,19,21,23,25,27,30,32,34,36];
  blacks     : wheel = [2,4,6,8,10,11,13,15,17,20,22,24,26,28,29,31,33,35];
  odds       : wheel = [1,3,5,7,9,11,13,15,17,19,21,23,25,27,29,31,33,35];
  evens      : wheel = [2,4,6,8,10,12,14,16,18,20,22,24,26,28,30,32,34,36];

begin
  randomize;
  numbusts := 0;
  numplus50 := 0;
  totstakes := 0;
  profitloss := 0;
  for i := 1 to 30000 do            {play 30000 sessions}
    begin
    stake := 1;                     {new session - reset stake}
    bank := 0;                      {new session - reset bank}
    bust := False;
    win := False;
    conloss := 0;
    spin := random(37);             {get initial spin for session}
    while bank < 50 do              {keep going until +50u}
      begin
      {bet selection}
      if win then                   {previous win : follow last odd or even}
        begin
        if spin in odds then
          nextbet := odds
        else
          nextbet := evens
        end
      else                          {previous loss}
        begin
        if conloss <= 3 then        {if consecutive losses less than 4}
          begin
          if spin in reds then      {bet R => O; B => E}
            nextbet := odds
          else
            nextbet := evens
          end
        else                        {consecutive losses greater than 3}
          begin
          if spin in reds then      {bet R => E; B => O}
            nextbet := evens
          else
            nextbet := odds
          end;
        end;  
      {end bet selection}
      spin := random(37);           {get next spin}
      if spin in nextbet then       {win}
        begin
        conloss := 0;               {reset consecutive losses}
        win := True;
        inc(bank, stake);           {add profit to bank}
        stake := 1;                 {reset Marty}
        end
      else                          {loss}
        begin
        inc(conloss);               {increment consecutive losses}
        win := False;
        dec(bank, stake);           {subtract loss from bank}
        stake := stake * 2;         {double up}
        if stake > 128 then          {bust}
          begin
          bust := True;
          inc(numbusts);            {add to bust count}
          break;                    {quit session - busted!}
          end;
        end;
        inc(totstakes, stake);      {cumulative stakes for edge calc.}
      end;
    if not(bust) then
      inc(numplus50);               {add to +50 session count}
    inc(profitloss, bank);          {cumulative P/L for edge calc.}
    end;
  writeln('Number of busts : ', numbusts);
  writeln('Number of +50 sessions : ', numplus50);
  writeln('Ratio of successful sessions to busts = ', numplus50/numbusts:3:3);
  writeln('Computed Edge (PL / Total Staked) : ', profitloss/totstakes:5:4)
end.

 

A red or black streak happens once every 256 spins average, so about 4 times each in 1000 spins. As for changing the patterns, as I said before, it doesn't make any difference because each pattern of the same length has the same chance of hitting. I know it's tempting to believe that it's harder for losses to occur if you keep changing the pattern, but it doesn't happen like that precisely because all patterns are equally likely.

 

That is reassuring.

 

Basically life can gift you with a lucky streak for 30k+spins with a basic martingale.

 

Your ratio of 3.632 successful sessions to 1 bust was established after 30,000 sessions were played. My data of 6 to 1 comes from only 231 sessions played. Can you see why there might be a discrepancy? So what is the bottom line from your testing? Is strategy successful or not?

 

Thats a nice font, was it called?

 

Out of 231 what is profit?

 

Yep. Seems like you've hit a very lucky streak. I set the number of sessions to 231 and ran the simulation multiple times. The best result was this :

Number of busts : 34
Number of +50 sessions : 197
Ratio of successful sessions to busts = 5.794
Computed Edge (PL / Total Staked) : 0.0223

Quite close to your 6:1 ratio, and with further runs there could be an even better result. Players who have hit a great winning streak often think they've found a Holy Grail, but long term tests always give the true picture. So sorry, not a HG. Maybe you should think about quitting while you're ahead.

I was misled by the first simulation because I hadn't set the progression length correctly.

 

I have already posted that on page 5 of this thread. 8,716 unit profit.

 

Bitstream Vera Sans Mono. You can download it here.

 

If you play the same pattern, a seven step marti as an example...if you keep playing it,

It goes 1/128, then 2/128, then 3/128, then 4/128, then 5/128, then 6/128 etc....until the shit happens

But if you change the pattern every time the odds stay at 1/128...

Am I right or wrong?

 

Im Probably wrong...

But anyways....how about first putting a big amount say £500 on either red or black and chase the loss with pattern betting.....

Or using pattern betting to recoup loss from a different strategy....

 

Could be the quickest £500 ever, or a not too long recoup in event of loss

10
10
20
40
80
160
320
640
1280

A 9 stepper

 

Better for smaller amounts.. .

Aim to make £50...so bet that first, if you win awesome, if lose you are protected by an 11 step marti pattern or more depending on table limits to recoup loss...

1
1
2
4
8
16
32
64
128
256
512

 

Because I already told you it probably won't work. But you seem to be invested in the idea so I was trying to help you find the best thing you could.

Truth is, betting red/black with a martingale progression is doomed, and everybody knows it. But still we try to help.

I also suggested you try a modified d'Alambert instead of the ridiculous martingale.

Good luck.

 

So Jerome, what you are saying here is, that after 30,000 sessions played which would amount to an average of 125 spins per session X 30,000 = 3,750,000 spins (give or take) and it shows a handsome profit after all of that. Let's see, 23,523 - 50 unit sessions won = 1,176,150 units won, minus 6,477 sessions lost X 128 per session = 829,056 units lost. So then, 1,176,150 units won minus 829,056 units lost = 347,094 units profit.

But that's not all. You ran another 231 session test; 197 sessions - 50 unit wins = 9,850 units won. 34 busts of 128 units = 4,352 units lost. So therefore, 9,850 units won minus 4,352 units lost = 5,498 units profit. Add that to the profit above and you have a total of 347,094 + 5,498 = 352,592 total profit from the two.

And even that's not all. Mac also tested the strategy showing his results after 1,926 spins with a profit of 284 units.

You know what's really ironic, you made the statement: "Players who have hit a great winning streak often think they've found a Holy Grail, but long term tests always give the true picture. So sorry, not a HG. Maybe you should think about quitting while you're ahead."

Wait just a minute here - great winning streak? You and Mac are the ones who ran those tests. The results have nothing to do with my luck or winning streaks. You ran these tests on computer software which doesn't allow for luck. So what is a long term test, a trillion sessions. I think you math boys should give it up, because there is no software that will ever be able to calculate the probabilities of randomness. I never, ever claimed this to be the Holy Grail, but your own result show that it may be damn close.

 

I'm looking into using pattern betting as a recouping mechanism for other strategies, the less it is used , the better....I reckon

 

How about a a report from the action with this method from a B & M casino . ( proof please )


.

 

Take the the table limits out and we would take them to the cleaners be it you have the capital.lol

 

No you wouldn't.

Every bet must inevitably be resolved. The expectation at each spin is negative. Therefore you would lose.

It is not possible to turn a negative expectation game into a positive one by simply raising or lowering your bets.