DNA is affected by parts of the cell (i.e. physical conformation), but to suggest that the genetic code directly mutates as a result of contact with external stimuli is just plain incorrect.
Positive/beneficial/bad/worse are all terms that are fairly poorly used when it comes to talking of evolution (i.e. Is any mutation that is passed on to successive generations beneficial??). Sickle cell anemia is "positive" in areas of say, Africa, where malaria is abundant and this mutation makes humans impervious to the disease. Anywhere else that cases of malaria are few and far between, this is obviously a negative issue - for us.
As far as the odds dictating more negative changes, why? Quite simply, most mutations are benign.
The rest of your argument really seems to slide down to the semantic argument of directionality and the usage of positive/negative mutations. I'm sure you can lookup any number of sources to verify that your claims are a little off-base.
