Well, to be fair, I have already offered an explanation:
To put it more concretely, say you watched a basketball team for four days and they did this:
L 99-105 vs. Team A (19-26)
L 96-106 vs. Team B (21-24)
W 99-95 vs. Team C (21-26)
If that was all you ever had the opportunity to see that team do, you probably wouldn't think too much of them. Maybe "wow, that 33 has freakish athleticism but he can't throw the ball in the ocean" or "looks like 23 was pissed for the whole game, he must have an attitude problem" or "what's up with the hair on 91, and why isn't he doing anything other than rebounding?" But you'd actually be watching the best basketball team in human history, the 1996 Bulls.
The tournament is largely how we judge a team's failure or success, but that doesn't mean that you can learn more about a team's body of work over those one to six tournament games than you learned in the 35 games that preceded it. If you were an NBA exec, or a college coach, or an AD, or a professional gambler, you'd not get very far if you took that approach. (However, if you were a sportswriter, you'd just be doing your job.)
It's not that there's no merit in looking at a single game and wondering why it's telling you something different than the games that came before. But it's only a very small piece of the puzzle.