Originally Posted by
star_world
Based on what sort of failure rate? Where is your threshold?
If 1 in 10 fails, then yes I'd probably agree with you. If 1 in 10,000 fails, then what?
I think the failure rate is a lot closer to 1 in 10 than it is to 1 in 10,000.
If it were 1 in 10,000 we probably would have only seen 1 report on FT, if that. We've seen a lot more corroborating data points on this one. And that does not count the people who are not monitoring this possibility, not on FT, and/or simply not monitoring the systems and checks. So you have to multiply what FT posts and multiply it by some unknown factor to increase it.
Originally Posted by
star_world
I think the underlying issue here is your interpretation of how frequently these things happen vs. mine.
Quite possible.
Originally Posted by
star_world
The world is full of examples of people being told to follow a certain set of instructions or a complex process when the underlying process isn't infallible. It's rare to see caveats placed around it to the extent that you believe should be.
Right, but it's the degree and impact of the failure that's key. Once the failure rate increases, that's when the caveats and checks come into play.
Unfortunately, what we've seen since 3/3 is an unreasonable failure rate, IMO, of this and other processes, with consequences that have a large impact to the affected customers.
Any time an item happens enough that customers can corroborate it and know the issue happens as a known failure, it's probably happening too frequently.