I'm hopeful that Michigan will improve its shooting before the Michigan State games, but the team has to adjust to how well-scouted they are compared to non-con games.
One key advantage Michigan has from looking at the Barttorvik.com numbers is in turnovers at both ends. MSU is really bad at forcing them while Michigan is good at not committing them. Michigan is actually slightly above average at forcing them and MSU still commits an average amount at 18.7%.
To the other point, MSU's OR% is the same as the OR% allowed by Michigan at ~27%, but MSU's OR% and Michigan's DR% both rank 18th nationally with ther 12% discrepancy, so that will be key.
That site doesn't have Foul or FT rate numbers, but I know Michigan still ranks high in not committing fouls but is only average at drawing them, while MSU appears to still draw and commit a lot of fouls.
But to win, Michigan will have to shoot well and/or win the turnover margin significantly. Winston has had some of his worst games against Simpson, so I wouldn't be surprised if that happens, again.
This week is big for Michigan, though, with Ohio State and at Iowa.