Actually, it's very possible. The kinds of people who do absentee ballots might be military people - in which case they might vote 70-30 for Sullivan. Or they could be people from the remote areas of the reservations, in which case they might vote 70-30 for Begich. I'd say there's a good chance Begich might still win.
Don Young, Republican of Alaska, is the longest serving member of congress. He has to run a statewide election every two years. One would think Sullivan would get the same votes from the villages that Young got.
Just a follow up, Sullivan won by about 8800 votes. Final is 119,000 vs 111,000, suggesting that the absentee votes fell almost exactly in the same proportion as the regular votes. Which is what “should have” happened statistically speaking. It was 110,000 to 102,000 at the time of this article, so the remaining votes split almost exactly 50-50 (9000 each). Its just incredibly unlikely that the absentee vote would break hard (70%) one way or the other unless it was from a very focused locale.