Quote:
Originally Posted by pastorjoeboggs
Since 1995, teams that go into the Division Series with a record 5 games better than their opponent are 36-30. While it *is* fairly close, the evidence suggests that the team with the higher record wins about 55% of the time, which is not statistically insignificant.
Strangely enough, when you increase it to a 10 game record gap, the winning percentage drops to 53% (16-14).
Still, the evidence seems to suggest that when a team is significantly better than their opponent going into a series, there is a reasonable expectation that they will win more often than not. That has not been my experience in this game.
My point is that I now have two years of evidence in PT that upsets are far more frequent than in real life.
|
You're kind of answering all your own points here.
First off, you mention you have two years of results. Two years means two playoff series. That's a ridiculously small sample size - not enough to gauge anything on.
As far as the stats you posted, 55% and 53% win probabilities are hardly percentages that say the favored team should overwhelmingly win the playoff series. 55% and 53% win rates mean 45% and 47% upset rates, which seems like a pretty darn high upset probability to me. And you're using a sample size of just 2 playoff series, so 55/45 and 53/47 are essentially a coin flip.
And on top of everything I just said, you yourself are acknowledging the real life division series favorite vs. underdog record as being "fairly close"...... and then when the win gap gets even wider the favorite vs. underdog record gets even closer which goes to show you just how random baseball is with these small sample sizes