A common criticism of Sarkisian's Husky teams is that they couldn't win on the road. I did a little research to test that theory, and here's what I've come up with.
I looked at the differential in winning percentage for all Pac-10 teams, conference games only, between 2005 and 2009. I chose those years even though it includes Sarkisian's first year because it was easy to get the data from College Football Data Warehouse. I ignored Colorado and Utah, since we have less data for them. Focusing on conference games only avoids the problems of severe mismatches and home-only series. I.e., we don't know how Portland State would perform at home against a Pac-12 team.
First, the home team won Pac-10 matchups 56 percent of the time. This means that, on average, Pac-10 teams had a differential of 56-44 = 12 percent between home and away. However, there was a large range: the smallest differential (WSU) was -0.5 percent, i.e., they were slightly better on the road. (It pays to leave Pullman.); and the largest differential (UCLA) was about 36 percent. Here's a summary.
This is a wider range than I would have thought. Partly that's probably because I looked at only a few years. Note that the differential for Washington is right in the middle of the Pac. Willingham was equally bad at home and on the road.
How did Sark's teams compare to this distribution? I've re-created the figure including Sark's teams, looking again at only Pac-10 games.
As you can see, the win percentage differential during Sark's tenure was way out at the extreme. This seems to offer some basic evidence supporting the conventional wisdom: over the last five years, the Dawgs have had a road problem. Or I suppose you could say the Dawgs have had a huge home-field advantage. Take your pick.