> Anyone reasonable would presume the machines are tested and get check up type service before their big day. Certainly they don't go from storage to stage without being thoroughly tested.

> Next, anyone reasonable would presume there's a know rate of failure and given that data an appropriate number of machines are on-hand as back ups so that the integrity of the process is maintained.

These would be reasonable thoughts if this was an industrial process that was done regularly, but it’s not. It’s done once every 2 years, and the voting machines are relatively new. There is no reason to assume they’ve done all this testing and contingency planning, and frankly they don’t need to.

A small glitch that is resolved doesn’t compromise the integrity of the process. There is absolutely no logic to that idea.