There is a guy in my area who has recently lost two larger scale planes. After the second one he was cursing his radio and laying the blame on it. Sure enough after sending away the Rx for testing it came back as bad.
Now there are quite a few of us in the area that dont believe the radio was the culprit, sure it tested bad, AFTER two crashes in large models. You only have to watch this guy flying to make the assumption that it wasnt the radio's fault, he seems to barely have control of a small fun fly at times.
Im not saying 100% it was pilot error, but . . . . . Im sure if we were able to have a black box type thingy, like airliners do, the statistics would come up about the same for model crashes. IE: most can be put down to being human error as the cause, either CFIT (controlled flight into terrain), error in maintenance procedures or some other factor that was preventable. Pure radio failure (not being dead battery or loose connection) would account for very few crashes.

However that doesnt mean we get complacent about our radios as others have said, do the range check, it will show up the loose connection or dead battery etc as well as complete failure.