A problem with rare things is it's hard to get good sense of what to do about them. We know the "no O2" method is very rarely needed, simply from observed stat

A problem with rare things is it's hard to get good sense of what to do about them. We know the "no O2" method is very rarely needed, simply from observed statistics. We know that O2 masks come down more regularly than that. Such stories are not hard to find. Here's a list: http://www.avherald.com/h?search_term=Cabin+pressure&opt=256&dosearch=1&search.x=-862&search.y=-106 http://www.avherald.com/h?search_term=Cabin+pressure&opt=256... . Looks like a few dozen per year. If there were a switch like you suggest, what is the probability that it will engage by accident, and prevent the masks from otherwise deploying as they should?