With LRLs and dowsing, "random chance" applies to randomized blind tests, not to field use. A randomized blind test does 2 things that a field test cannot do. First, it eliminates outside influences that might alter performance results, such as observable clues. Second, it provides a baseline from which to compare results, namely guessing.
Despite intentional attempts to mislead people, random chance doesn't apply to field use. You can't ask, "What are the odds of digging 10 holes in a park and recovering a gold coin?" There is no way to calculate that, because there is not enough information*. But in a randomized blind test, it is quite easy to calculate the odds. Depending on the design of the test those odds can vary, so it is not a fixed number that applies to every test, but it's not a "moving target" either.
The administrator has disabled public write access.