No, he has a point about the average misrepresenting the whole. His previous year's poor play at the end of the season drops his average OPS below .800, when in two of the three yeasr, his OPS in August and Sept. was above .800, including a .900+ two month stretch. Based on the fact that, as noted before, time based splits are often not predictive, especially when, as you see with Dunn's splits, they aren't consistent to begin with, I'd bet that he'd be closer to his season OPS over the last two months than his three year average in Aug-Sept. Later on in his career, he may establish some pattern, become a better first or second half player. Right now, we have one very good second half, one solid one, and one terrible one to go on. I don't see any convincing evidence.