>>> drug detection dog in Illinois, alerted for narcotics 93 percent of the time during roadside sniffs, but was wrong in more than 40 percent of cases.
Wow - ignoring the point of the article, that still means police officers were pulling over people to search at better than 1 in 2. Cops were good enough to look at a car and get it right that often. I mean what are the odds of randomly pulling a car and getting drugs in it? 50:1? 20? 100?
The point is, the cop with the dog stopped a selection of cars. The dog "found" drugs in 93% of cars, but was wrong 40% of the time. Roughly speaking, out of every 2 cars the cop stopped, 1 had drugs in. Unless this was a main gate at Glastonbury, that means the cops is amazingly good at choosing cars likely to be containing drugs.
It does not matter if the dog is any good or not, the cop is.
(plus if you can find out what "tells" the cop is using, you can stop giving off the tells - unless its really dumb like clouds of sweet smelling smoke pouring out of the windows)
Edit: Or do you mean the cops pull over lots of cars, and then only sometimes decide to use the dog? Well, the same applies, only this time its less impressive as the cop gets to lean in the window, look over the occupants etc etc. But still, a 50/50 decision to call in the dog is a very high hit rate.
I am not sure you are reading the comments and applying yourself to understand the point being made.
When a decision is taken to use that dog (no matter who or where), the decision was 'right' 1 in 2 times. This is way better than random searches of a car population.
Wow - ignoring the point of the article, that still means police officers were pulling over people to search at better than 1 in 2. Cops were good enough to look at a car and get it right that often. I mean what are the odds of randomly pulling a car and getting drugs in it? 50:1? 20? 100?