Seriously. (Note that I'm talking about fiction).
This is sort of a corollary to "write what you know." I often hear beginning writers opine that "but that's the way it happened." "How it happened" is often how first time novelists write or are inspired to write (at least for a first draft), which gives rise to the oft-repeated sentiment that first novels are often at least semi-quasi-autobiographical. (Or at least more so than subsequent novels).
But, globally speaking, "how it happened" is not a legitimate criterion. Just because something actually happened that way does not make it realistic for a novel. This may be counter-intuitive, but someone (I want to say Dr. Johnson) said that it's easier to believe (in?) the impossible than the improbable. This is true. Most of what happens in people's lives that we think "would make a great novel" falls in the category of the improbable. I tend to agree, and therefore advise that it's usually best to shy away from "how it happened" and go with "what you know."
"What you know," of course, need not be a limitation (see post below on Research), and telling "how it happened" is kind of, ahem, disadvantageous (also possibly litigious), anyway, because, after all, what we're writing is fiction. (That which actually happened can inspire, but the author's imagination should lead, not follow).
There's this story I read somewhere (I can't remember the source, so don't vouch for it) that Lawrence Olivier was once told about how Dustin Hoffman prepared for roles. There was one, in particular, for which Hoffman supposedly deprived himself of sleep for something like three or four or five days. When Olivier met Hoffman, it is said, he told Hoffman, "My dear fellow, why don't you just try acting?"
In a similar vein, to beginning novelists, I would recommend "simply making it up." (Unless, it's a memoir).
After making sure you've done your research, of course.