ProPublica:

From his basement in upstate New York, Herbert MacDonell launched modern bloodstain-pattern analysis, persuading judge after judge of its reliability. Then he trained hundreds of others.

But what if they’re getting it wrong?

In 2009, a watershed report commissioned by the National Academy of Sciences cast doubt on the whole discipline, finding that “the uncertainties associated with bloodstain pattern analysis are enormous,” and that experts’ opinions were generally “more subjective than scientific.”

It’s kind of terrifying to think this “science” made just be a load of bollocks. Even worse, how many innocent people were convicted? How many guilty people went free?