Desperately Need A Programmer

[ScoobyDoo]

We need a coder to program something that will take a matrix for dozens (7-spins wide) to determine how many of the same dozen will line up vertically in any colume such as:

3 2 2 1 2 1 3
2 1 2 3 1 2 3
1 2 3 2 2 1 3
2 1 2 1 3 2 3<---The last colume has four in a row

We are interested in learning how many times a group of three and groups of four and higher happen.

If some of you are interested, please contact Twisteruk or Johnlegend in the thread:

MATRIX VERTICAL METHOD ONLY

Could really use your help.

Scooby Doo

[mr.ore]

Code Select Expand


// tracker.cpp
// WTFPL license: link:://sam.zoy.org/wtfpl/

#include <cstdio>
#include <cstdlib>
#include <ctime>

int main(int argc, char *argv[]) {

  srand((unsigned)time(NULL));
 
  int nlines = 1000;
 
  if (argc > 1)
    nlines = atoi(argv[1]);

  const int width = 7;
  int matrix[width] = {0};
  int length[width] = {0};
  int group[64] = {0};
  int max = 0;

  for (int spin=0; spin<nlines; spin++) {
    for (int i=0; i<width; i++) {
      int dozen = rand()%3+1;
     
      if (dozen == matrix[i])
length[i]++;
      else {
group[length[i]]++;
length[i] = 0;
      }
     
      if (length[i] > max)
max = length[i];

      matrix[i] = dozen;
    }
  }
 
  double sum = 0.;
  for (int i=1; i<=max; i++) {
    sum += group[i];
  }
 
  printf("max = %d\n",max);
  for (int i=1; i<=max; i++)
    printf("%d - %d - %f\n",i-1,group[i],group[i]/sum);

  return EXIT_SUCCESS;
}



-------------------------------
./tracker 100000000

max = 19
0 - 103714115 - 0.666700
1 - 34562919 - 0.222179
2 - 11524435 - 0.074082
3 - 3841407 - 0.024694
4 - 1280589 - 0.008232
5 - 427108 - 0.002746
6 - 141550 - 0.000910
7 - 47549 - 0.000306
8 - 15792 - 0.000102
9 - 5260 - 0.000034
10 - 1789 - 0.000012
11 - 612 - 0.000004
12 - 202 - 0.000001
13 - 64 - 0.000000
14 - 24 - 0.000000
15 - 9 - 0.000000
16 - 2 - 0.000000
17 - 1 - 0.000000
18 - 1 - 0.000000

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

In 7*100 millions spins, dozen will line up vertically max. 18 times. Counts of occurence and percentages are also there. I hope I have it right...

[mr.ore]

BTW matrix size have no effect on the number of occurences of groups, only number of trials.

[mr.ore]

Btw zero means that there were change, 1 means two in a column and so on.

[ScoobyDoo]

Thanks Mr.Ore.......Unfortunately I haven't got a clue what I should do with the info you gave.
                                                                                                                                                        I was hoping to have a finished software, Bot or whatever format it would be in because no one in that thread knows anything about code or coding or much of anything else other than just what they can see in front of them. Any chance I could get you to do that? Pleeaasseeeee?

Scooby Doo

[mr.ore]

The simulation shows obvious - there is no advantage in that matrix, if you bet a dozen, the probability that some other than previous shows is 2/3=0.666, the probability of a series of two is 1/3*2/3=0.222 and so on. Matrix does not seem to help anything, no matter how you change width.

What you want to do? Measure if it helps to "force" game to have a concentration of hits in certain limit and from time to time a huge lose, without something in "middle"? If only that were possible, the game were beaten already. I wish there would be some method, but as long as you can split choices of bet selection into independent events, there is no help, you can't run from that lose, even if you wished. I also tried ideas like playing a system on several location in virtual and as soon as a progression tanked on some location, switch there and hope it will not tank soon enough. It does not work in the end.

What seems to work best for me is to react to game by playing insides or outsides. If it's hitting play inside numbers and capitalize, if not, move to dozens or even even chances. If not hiting and already up, try to risk some units on streets or splits to stop losing...

What you probably want is a graph of simulation, that would take some more time, and I already know, it will most probably not work. Anything can happen. If you see some regularity in 10000 spins, it means nothing. If you see on the graph, that after every x wins come a lose, it means nothing. Gaps between loses means nothing. You can change your system, count wins, wait for virtual loss, and the system will certainly win over that same spins, but feed it another 10000 spins, and it tanks. Frustrating, but true.

In roulette, there are only three things you can truly control - bet size, variance(=coverage) and target. Decisions alwayds depend on your balance and target. If you do not want lose now, you increase coverage and/or decrease bet, you win less, but sacrifice long term probability - that's a trade. You can also do it in reverse of course. That's only control you have over game.

[The_Force]

If you're interested in matrices, you're much better off looking into SVD to analyse them.

[mr.ore]

I can't see how could be SVD useful in this case. How could be that decomposition useful? I did not look much into way they use the matrix, but from what I saw, they just bet that cross won't be formed. So the matrix is really just a bet selection scheme.

[superman]

I did not look much into way they use the matrix

It's just visual stimulation, they think they see something but it's nothing to get so worked up about mr.ore, I have botted it and it fails at the usual rate as expected, oh well