Quote:
Originally Posted by 14cobster
What's this can't prove a negative business?
If you're serious---
In this case, the problem is twofold.
First, you can test for one type of rigging (say 'action flops'). When that doesn't show rigging, then the riggies will way 'no action flops'. When that's tested, then the riggies will say KK v AA. When that's tested then the riggies will say boomswitch after deposit. When that's tested then the riggies will say doomswitch after withdrawl. When that's tested ... (I hope you see the pattern)
Second, the tests that have been done are statistical. In this case, you are trying to reject your null hypothesis. That is, you never 'prove' your hypothesis of 'not rigged'. You either reject the null hypothesis (at some statistical level) or fail to reject the null hypothesis.
Now, if you actually had the code, you could possibly get around both of these issues by 'proving' the code does what it's supposed to do. But, another nasty version of the first problem shows up--the riggies can now say 'well, that's not the code that's currently used, they swapped the code out for a rigged code', and we're back to square one.