100 percent accurate. And to complicate things further, stats start doing counter-intuitive things that go against the way people colloquially talk about a “difficult” schedule.
Two simplified models for the five game “meat” of two schedules:
Team A: Four games where they have a 75 percent chance of winning, one game that is very hard against a very good opponent and they have a 25 percent chance of winning.
Team B: Five games, they are all favored, all have a 60 percent chance of winning.
The probability of running the table for both teams is basically identical (a bit under 8 percent), and in fact Team B has a slightly lower chance. But most fans and observers will greatly reward Team A because of that one very good win, and comparatively dismiss Team B because they are all games that they “should have won.”
And that’s where a committee needs to plunge beyond surface “best wins/losses” discussions and get deeper on analytics when weighing teams.
In general, humans tend to do two things when looking back at events have have happened whether sports, elections, or games of chance:
1. Are over-impressed when something happens that is “against the odds,” even something that only has a 10 or 5 percent chance of happening WILL happen. And when it does, it doesn’t mean the original odds were wrong.
2. Are under-impressed when something happens that the “odds favor,” as cumulatively these things begin plummeting quickly.