Beating Roulette with a Computer Program

[gizmotron]

It came up in another thread. Is it OK to keep track of ideas? Yes. I think it's OK.

On another forum I demonstrated how easy it would be to write a handler method that could allow itself to be re-used over and over to perceive a quality, expressed by a condition of being correct by a specified percentage, a specific pattern or dominance and then having it return a true or false if it fit the conditions that would be  acceptable enough.

Here is that handler (method):

Code Select Expand

on passThisArgument thisData, thisCondetion, thisPercentage
global conditionTrueFalse
  put "false" into conditionTrueFalse
  if thisCondetion = thisData then
    put "true" into conditionTrueFalse
    exit passThisArgument
  end if
  put the number of items in thisCondetion into rTimesToCheck
  put 0 into holdTempConditionCheck
  repeat with i = 1 to rTimesToCheck
    if item i of thisCondetion = item I of thisData then
      add 1 to holdTempConditionCheck
    end if
  end repeat
  put (holdTempConditionCheck / rTimesToCheck) into percentageTrue
  if percentageTrue > (thisPercentage - .1) then
    put "true" into conditionTrueFalse
    exit passThisArgument
  end if
end passThisArgument

Now all I have to do is write an attacking handler and an array of conditions for the software to pass through the two handlers. I should be able to write thousands of conditions that will execute in the blink of an eye.

The only reason this has not been attempted is that most people don't see situational awareness with respect to pattern recognition in a gambling game. Programmers don't create in a vacuum. They must know the topic. They must know what a real player would consider. They just aren't interested otherwise this would have been done years ago. I'll tell you what is more difficult than this. Writing a chess player would be no picnic. But that has been done for decades.

[gizmotron]

-