The right comparison is to what would have been lost in corrections but for the preparations.

I mean, if you have x losses in preparation to achieve y losses in actual corrections, when you would otherwise have z losses in corrections for z > x+y, that's a win.