M - Penn State prediction based on QB play

Back in the Lloyd Carr era I had a theory that the main thing that determined Michigan's chance of beating their opponent was who had the better QB.
Chad Henne's teams generally won even if they didn't seem to be they better team because he played better than the quarterback for the opposition.
When John Navarre was better than Craig Krenzel  -see 2003, M won. When Craig Krenzel was better than Navarre, M lost - see 2001 & 2002.
Now M won some games when the QB was not so good - Iowa and ND in 1997, Purdue in the snow in 1995 and lost some games where they certainly had the better QB - see most every game they lost with Tom Brady as QB. But in general it holds for the Lloyd Carr era at least.

The Harbaugh teams seem to follow this as well.
His first game - the 3 interceptions from Jake Ruddock were most of the difference. Although the MSU game ended in that ridiculous punt fiasco - if you switch QBs in that game M wins handily.
The 2016 Wolverines were dominant for many reasons - but when Speight played poorly at Iowa and was dealing with his lingering injury issues they lost 3 out of 4.
In 2017 - they have won despite poor QB play - but generally their defense has made the opposition QB look even worse. Last week vs. Indiana may be the exception.

I don't think O'Korn is a better QB than McSorley . There is a possibility that Michigan's defense can bring McSorley closer to his level.

But not close enough.

Michigan 13
Penn State 20