They did that with New Hampshire and they probably based it on overwhelming exit polls.
The winner can logically be called this early, but to act as if the percentages separating the top 3 are solid enough to analyze who did better or worse than expected is ridiculous.