Stipulated: Michigan has no case for the NCAA tournament unless it wins at Minnesota on Saturday. The rest of this post assumes a Michigan win to get to 9-9 in the Big Ten, because if it does not it is pointless. Let this not be a jinx.
Why is Penn State in brackets ahead of Michigan? Look: they're currently in Lunardi's Bracketology and several others in the Bracket Matrix. 33 brackets have PSU in; just 11 have Michigan in. Hypothetical Michigan is 9-9 in conference and 19-12 overall pending the Big Ten tourney. Hypothetical Penn State (@ Iowa, Illinois) is probably 22-11 and 10-8 in conference. Right now they're 9-7.
Penn State has an advantage in conference, then. However, out of conference:
|Games Against Kenpom Top 100||Duke(5), UCLA (9), Duke(5), UConn(3), Maryland(58), Northeastern(96)||Temple (52), Rhode Island(66)|
|Wins against same||Duke, UCLA, Northeastern||None|
Penn State's best nonconference win is against #104 Georgia Tech, which is headed for a one-win season in the ACC. Hell, Michigan's win over #113 Oakland is about as impressive as that. You can wipe away the Temple/Maryland losses as equal and call Michigan's win over #96 Northeastern equal to PSU's loss to #66 Rhode Island—very generous—and then you've got two wins in four games against top-ten opposition by Michigan versus Penn State's absolutely nothing. Surely that's a more impressive record even if Penn State is 1.5 games ahead in the conference standings, especially in a year when Michigan is missing the surest conference win of all: home against Indiana.
I make this point because I've heard a lot about Michigan's lack of a killer road win this season, which is something that makes little sense to me especially in relation to conference opponents. Penn State and Michigan play virtually the same schedule in conference, but you get the impression that a team that goes out and loses all its home games and wins all its road games is in much better shape than a team that does the opposite. This is clearly bats, as both teams have the exact same SOS, but the hoo-hah about road wins is the most ominous thing we hear about Michigan, and if they get bumped off the bubble it will obviously be the #1 thing cited. No matter that Michigan actually played a nonconference schedule, unlike PSU.
I don't know where Michigan stands in relation to teams in the Big East or Pac 10 or SEC, as there are just too many variables. But if 9-9 Michigan gets left out of the field and 10-8 Penn State gets in, there will be a fit around these parts. Road wins are a stupid metric. And even if they're a good metric it's because it showcases the team's ability to play at a very high level and do a very difficult thing, which Michigan has already demonstrated by beating UCLA on a neutral court and Duke at home.
Update: forgot about the UConn game.
Update II: some protest in the comments. Let's look at it like this: in the hypothetical situation where M beats Minnesota and PSU splits its last two, Michigan and Penn State will have identical records across the 16 conference games they shared. Consider the conference games the teams didn't share as nonconference games and this is what you're looking at:
|Games Against Kenpom Top 100||N-Duke(5), N-UCLA (9), Duke(5), UConn(3), OSU (46), @ Maryland(58), @ Northwestern(68) Northeastern(96)||@ MSU(14), Temple (52), N-Rhode Island(66)|
|Wins against same||Duke, N-UCLA, @ Northwestern, Northeastern||@ MSU|
Then you are basically holding Penn State's one win at MSU against 2 losses to meh teams equivalent to Michigan going 4-4 against the schedule you see above including two wins over top ten foes. Does not compute.
Again, if Michigan loses to the Gophers this is moot, and if Penn State wins its last two it's also moot. I just don't understand how you can look at the above resumes and give Penn State the edge.