Let's see, there were 113 flights and 2 fatal accidents. The proven accident rate so far is therefore 2 failures in 113 flights, or an average of 1 accident per 56.5 flights.

Before the Challenger explosion (on the 25th flight), NASA was publically claiming that the rate would be on the order of 1 accident per 2000 flights. That estimate had no basis other than wishful thinking. On what do they base the newer estimate, not so wishful thinking?

NASA has some very good and dedicated people. They will work very hard to find the cause of the Columbia accident and to implement repairs. The remaining Shuttles will fly again, hopefully someday soon. Still, the Shuttle will continue to be a dangerous vehicle due to its very complexity.

Not too long after the Challenger explosion, I remember reading a column about how the accident was a "failure of low tech." His point was that it wasn't the high tech systems that failed but a rubber O-ring. While the cause for the Columbia accident is far from being conclusively proven, let's suppose that the early belief that the insulation from the ET caused the accident. If so, then once again, it'll be a failure of low tech that caused the destruction of one of the most complex vehicles ever built. How many other "low tech" accidents waiting to happen reside in the Shuttle design?

Based on the Shuttle's track record so far, to reach the NASA estimate of 1 accident per 240 flights, they'll have to fly some 607 flights in a row befor having another accident. I base that on (240*3 - 113) flights. That sounds somewhat unlikely.