If they are worried about dislike attacks why don't they blind the data with delayed effects?? Like update of a dislike appears days later but you can still see

If they are worried about dislike attacks why don't they blind the data with delayed effects?? Like update of a dislike appears days later but you can still see a ratio of likes vs dislikes for the last year or so...