From the article:
The real problem is that Vermeil's chart only accounts for one of the three factors that go into making the right decision. One missing part is how much time is left in the game, which to me is one of the most important factors. When the Raiders attempted their two-point play, with over 31 minutes left in the game, they were going to get at least six more possessions. So why the sense of urgency to tie the game when there is an entire half to be played?
The third element that needs to be factored into the decision is the probability of being successful on the two-point play. The odds of converting two-point plays are not high -- 137 of 300 since 2006, or a 45.7 percent success rate -- so why pass up an additional point when there is a significant amount of time left in the game? Also, teams normally go into a game with one two-point play -- when they use that specific play early, they are left to improvise if they need another one later.
Therefore, a chart should have the degree of probability of making a two-point play and the amount of time left in the game. When those two are factored together, only then can a correct decision be made.
That is exactly how I felt when they went for early. I think you always save the 2 point conversion, when you are down, for later in the game because you don't know what else will happen in between.