casualmente il libro che ho ripreso a leggere in questo momento recita:
Postdictive Error Bias
People use postdictive errors when they use information in their
testing that would only be available after the fact. This kind of error
is very common in system testing. It is easy to make. For example,
in some software, unless you are careful, you can use today’s data
in your testing-which is always a postdictive error. For example,
imagine the value of being able to use today’s close to predict what
prices will do today. That’s a postdictive error.
Sometimes these errors are quite subtle. For instance, since the
highest prices in your data are nearly always followed by lower
prices, it’s quite possible to sneak high prices,into a trading rule
SO that the rule works great-but only postdictively.When you are testing data, if your results seem too good to be:
true, they probably are. You probably got those results through
postdictive errors.
coincidenze del destino 
C