This will be by region
3) Michigan St
4) St Louis
None of the top seeds look promising as upsets or are playing promising upset prospects. Valpo is a little underrated as a 14 is the best option.
5-12 I'm not sure what the ESPN population is doing with the shower of picks for Oregon as a probable upset over Oklahoma St. Oregon isn't very good, and not so much so that they're near equals to OKSU
6-11 I'm also not sure what they're doing taking Memphis so heavily (87%). I'd take either of the 11 seeds over them, especially St Marys. It's possible this will shift after the play-in game is determined, but I doubt by that much. There's a huge value pick in taking the upset here.
7-10 And again, I'm not sure what the UC over Creighton pick is about. UC has a good defence, and Creighton.. doesn't know what defence is. But they score in bunches and are in the top tier of teams. Pretty good value just sticking with Creighton here.
8-9. This is a funny one. I don't think it matters who wins, so take your pick. Missouri is higher in my ratings, but not by that much. I'd weight toward CSU because Missouri is so bad on the road and because the public seems to favor Missouri to a very high extent.
Louisville is playing in Lexington. While that could be a semi-hostile crowd, I don't think that's sufficient. They're a good road team. Neither of their probable opponents are.
Duke- Creighton would represent one of the best second round matchups (much like UNC-Creighton did last year, not because it was close, but because it was high-scoring). I'd take Duke, mostly because I don't think Creighton can guard Duke but Duke could guard Creighton.
Michigan St-St Marys/MTSU/Memphis, I might take the 11s again here, but betting against Izzo isn't usually very smart. The game is also in Auburn Hills...
St Louis- OKSU. I'd take OKSU here. Typically the 4-5 is decided by who can score more. Neither can, and both can defend, but OKSU plays a little faster.
Louisville is clearly better than either St Louis or OKSU. Oklahoma St does seem like a more promising upset if forced to choose.
Duke is more of a tossup with Michigan St, as it's the first really good interior team they'll play. Duke though looks more promising to advance this far.
Duke. Partly this is because Louisville is the most common team picked to win it all. Partly it's because Duke is really pretty good this year. I don't think Louisville's defence is that much better than Michigan State's to match up here. I also don't think Louisville's offence is that good (it's worse than UConn's was when they won with the Kemba Walker year). I don't trust them to score enough to keep up with Duke and they're not an interior style team that could pose problems (Maryland's got size, if we insist on the Maryland template for beating Duke).
5) VCU, if forced to pick anyone.
Again, no promising upset potential in the top seeds. Northwestern St at 14 is fast paced and has beaten a 3 seed before several years ago, but they're playing Florida. Who is probably one of the best teams in the tournament. Whereas before they played Iowa, who was a high seeded Big 10 tournament winner and was pretty weak for a 3 (more like Marquette this year).
5-12 Akron is pretty good, but VCU is much better. The best advantage Akron has is the game is in Auburn Hills. The other is that it's a value pick as VCU is really popular for a 5 seed.
6-11 If Minnesota weren't so unpredictable and poor on the road, they'd make an excellent upset pick here. They're still very likely (not least because they'll crush UCLA on the boards). UCLA isn't that good, but is respectable on the road. The game is in Texas, which does put somewhat less stress on the travel involved for a west coast team. I lean toward UCLA here, but not by much.
7-10 San Diego State vs Oklahoma. I'd take Oklahoma, mostly because this game is in Philadelphia and all the way across the country for SDSU and because SDSU seems favored. Also MWC teams tend to under-perform, badly.
8-9. Take Villanova. North Carolina is heavily favored. I think UNC is better, but not 69% likely to win better.
Kansas is much better than either UNC or Villanova and playing in Kansas City...
Georgetown is vulnerable as a 2 seed (they should have been a 4). But neither team is that promising as an upset candidate. This might be one of the ugliest games of the second round also. None of the 3 are good offensive teams, and two are top 15 defences.
Florida really got screwed on its location (four potential games in Texas), but there aren't many southern cities hosting. Neither team looks probable here to hurt them here.
Michigan-VCU, a very interesting game. It's in Auburn Hills, which helps Michigan, but VCU is very dangerous. I lean toward Michigan but not by much.
Kansas may have trouble with either team here, VCU causes more problems I suspect (highest TO margin in the country, and Kansas is the only top 10 team with a negative margin). Michigan certainly has the talent to beat them too.
Florida, no contest over Georgetown. Florida isn't getting very much love on the public picking them at all for later rounds. Just keep taking them.
Florida. Kansas has potentially a harder path and Florida is better anyway.
2) Ohio St
3) New Mexico
4) Kansas St
This is the screwiest bracket by far, with some heavy over/under seeds from RPI effects.
New Mexico is the best 3 seed upset candidate in the field, but they're playing Harvard, which isn't that impressive or very good. The play-in game against Kansas St doesn't impress, and the game is in Kansas City. Iona is really good for a 15 and plays really fast but actually picking a 15 is usually dumb. So no upsets at the top end.
5-12 This is a curious one. Wisconsin seems to like playing the Mississippi style teams (they do well against Indiana for example). Both teams are underseeded. I'd say this is the second most likely 5-12 upset but I don't think I'd bite on it.
6-11 Could be the year Belmont actually plays up to their potential as a spoiler. They finally got the seeding to show for it, a year or two late (they're not as good as they were the last two years). Arizona has played poorly against the P-12 (mostly fast teams) and down the stretch. Probably one of the best first round games.
7-10 Iowa State is a gunner team, which could be a problem for Notre Dame, but they're also terrible on the road. Much like the Colorado St game, it probably doesn't matter. It's also a lot closer to ND (Dayton). Iowa St could be a decent value pick here, if they're not terrible on the road.
8-9 Pitt is a top 10 team. Wichita is not. The worst of these last year had two top 20 teams in one 8-9 game (Memphis and St Louis). I'll take Pitt here.
Pitt is a heavy upset value pick here. Gonzaga is heavily favored. They are pretty good, but they're the weakest of the 1 seeds by a bit over Kansas and Pitt is way better than the other 8 seeds.
Kansas St playing basically at home against a slow moving team with a poor road record, I'll take them.
I'd take either Arizona or Belmont over New Mexico. Especially Arizona.
Ohio State playing in Dayton against either of ND/Iowa St, I'll take them. Iowa St would be the more dangerous of the two (if they're hot, they're going to score a lot).
Pitt over Kansas St. Gonzaga over them too if hedging. Kansas St is more like a 7 seed here than a 4.
I might take Arizona over Ohio State. Games are in LA. But OSU is pretty good.
Arizona, if not taking Belmont in the first round. Gonzaga if not taking Pitt in the second, Ohio State otherwise. As I said, it's a screwy region.
None. This bracket is the weakest by far in the field. Even Marquette (4th best team) is not good enough to merit wild card potential.
Davidson is very likely over Marquette and worth looking at taking. None of the others are likely upsets.
5-12 Probably the most likely 5-12 upset. It's in San Jose so UNLV's road record is a factor. UNLV is probably the most overseeded team in the field. Take Cal. Could be the ugliest first round game (neither team can score).
6-11 Butler is probably the second most overseeded team. Bucknell isn't much worse than them and there's a heavy Butler push on ESPN. Probably take Bucknell on a value pick, but Butler is more likely.
7-10 Illinois is probably the 3rd most overseeded team. Having watched them at various points in the season, it can be painful for them to score unless Brandon Paul decides to have a national TV style game. I'd probably take Colorado here, largely because they're not well liked as a pick.
8-9 NC St is way better than Temple.
Miami over Colorado/Illinois is okay here. Miami is a fairly weak 2 seed though (not quite as weak as Georgetown). They're also old. And old teams don't do well.
Davidson over Butler (or Butler over Marquette). Davidson, for a 14 seed, rates almost as well as Butler does as a 6.
Syracuse over Cal, but I'd consider Cal as an upset here. Cuse isn't that great outside of NY.
Butler over Miami, or Miami over Davidson
Indiana vs Florida, Indiana, and this looks like the "title" game to me.
Duke vs Gonzaga/Ohio St/Pitt. Duke
Indiana vs Duke, Indiana. Taking Duke or Florida as the champ is a decent value pick, Duke is more likely in my view (the with and without Kelly stats are helpful) but Florida is less picked.
Arguments against Duke: they can't rebound and got blown out by Miami.
Argument against Florida, they struggle in close games (most of their games have been considerable blowouts).They also play slow, which can be a trouble spot (as when they get behind in a close game).
Louisville isn't a terrible pick, but if they're likely the most common to pick as winning, find someone else to pick is good advice to try to win (in case someone else nails the early rounds, it's best to beat them by having more available points later, where the rounds are more important anyway). I lean toward Indiana because I trust their offense more and they won't play anyone who slows it down (ala Wisconsin or Minnesota). Miami and Florida play slow, but Miami is suspect defensively.
I don't trust Gonzaga because the coach (Few) isn't very good. They haven't done as well with the often higher expectations since Monson left. They do have size this time however, which they didn't sometimes in years past. They're hard to evaluate as well with fewer meaningful games. One has to rely on the 3 peat over St Mary's a lot to presume they've got a lot of quality wins, when they've otherwise just got a close win at Oklahoma St, over Baylor, and pair of blow outs over Kansas St and Oklahoma.