> it's nonetheless improbable, and it isn't wrong to disbelieve you. Is this the right assumption about the meaning of "improbable" in the context of this data

> it's nonetheless improbable, and it isn't wrong to disbelieve you. Is this the right assumption about the meaning of "improbable" in the context of this data and category of studies? One way we could interpret improbable is the way David Hume does in the "Of miracles" chapter of his 1748 treatise: that is, http://en.wikipedia.org/wiki/Almost_surely http://en.wikipedia.org/wiki/Almost_surely not the case. The other way would be that the parent post either himself, or the memory to which he is referring is simply an outlier -- that is, a lottery winner of people or memories simply not clustered around the mean of being lost to this amnesia phenomenon. In the case of the former, I would totally agree with the conclusions of your comment and the professor cited in the parent. In the case of the latter I think it merely points out that we're applying the wrong kind of skepticism over these claims. Somebody always, eventually wins the lottery. I think this is a valid distinction that can be made, and if so, doesn't hold us to discounting anecdotes as summarily false.