For those that have ears...
The independent game model is not my own. I watched
Basketball Reference track the most recent NBA Finals, and I was impressed by how well it read the pulse of that series. It's at least as good as someone sitting on their couch claiming to read into the psyche of both teams... and possibly better (I apologize for the heresy).
It's a very simple idea: Suppose the home team has a 60.4% chance to win any game in a series. Apparently they arrived at this number looking at history(!). Then watch the graph as it tracks the ebb and flow of the series.
These odds mathematically gave a 46.7% chance of the Mavs winning the series from the beginning. This made sense, since at the time they lacked the home advantage. Then these are how the odds evolved (and would have evolved) going into each game as the series progressed:
Game One: 46.7% (A win improves it to 66.1%, a loss moves it down to 33.9%*)
Game Two: 33.9% (win: 54.0% loss: 20.8%)
Game Three: 54.0% (win: 69.3% loss: 30.7%)
Game Four: 30.7% (win: 44.6% loss: 9.5%)
Game Five: 44.6% (win: 63.5% loss: 15.7%)
Game Six: 63.5% (win: 100.0% loss: 39.6%)
I see nothing objectionable here. The Mavs began under 50% because they lacked HCA, but they rose above 50% when they stole it. The Game Three winner would've had a 69.3% chance of winning either way, since they would have lead 2-1 with two upcoming home games. Dallas barely avoided a monstrous 3-1 deficit (which would've been devastating even without whatever psychological impact you care to postulate about). The Mavs' 3-2 advantage made them a 63.5% favorite, but had Miami taken a 3-2 edge, they would've been at 84.3%. With Game Five in Dallas, the Mavs had a fairly better chance to become a fairly weaker favorite.
Many people don't like this kind of analysis. They find the premise rather dry and the calculations over their heads. But typically their only counter is to call individuals who are competent at this sort of thing "stupid" and "moronic".
*I'm not sure why a few numbers in their synopsis deviate from the combinatorical theory by a few tenths of a percent. I guess there's a rounding issue somewhere.