Here is something else to consider. The problem that claimed the Challenger was a design flaw. It was bound to take a shuttle sooner or later. That flaw has been corrected and should no longer be a threat. Perhaps were can caulk it up as a very costly part of R&D.

The Columbia accident was more like an accident waiting to happen. I think NASA was unintentionally playing the odds on this one.

Anyway, my point is that with each accident, a flaw is discovered and corrected so that it doesn’t happen again. NASA decreases the odds of another accident. Over time the shuttle become safer.

Look at it this way. We had an accident some five or so years after the program started. Then another one 17 years later. By this pattern the next accident shouldn’t happen for about another 30 years.

We look planes we can see a similar pattern. At first we had more accidents, but as more planes were built and the bugs worked out, we have fewer accidents.