Using the second concept of "average," yes I do. I have seen lots of car accidents, and lots of people pulled over by police (presumably for speeding or reckless driving). I have never experienced either, other than being pulled over once for a missing tail light (humorously, literally hours after I had purchased that vehicle). So I do consider myself a better than average driver.
But this idea of "better than average driver" just isn't very useful. Because presumably the majority of drivers have not been in an accident, while the average number of accidents per driver is obviously greater than zero. Thus most people are better than average drivers, in the same sense that most people have more fingers than average.
Did you read the two definitions of "average" I posted above? Under the second definition, if we go by traffic accidents, I am clearly better than average, because I have been in zero traffic accidents (as a driver). Since you seem to be going by the first definition, I'm curious how you are distinguishing drivers that all have a perfect driving record.
Consider a population of 100 drivers, none of whom have ever had a traffic accident. The average number of accidents per driver is therefore zero.
Suddenly, two of them collide. Now, 2/100 drivers have had an accident, and 98/100 drivers have still never had an accident. The average number of accidents per driver is now therefore 0.02.
Since 98/100 drivers have had 0 accidents, but the average number of accidents per driver is 0.02, then 98% of drivers have had below the average number of accidents.