Click here for Part 1.

Virginia: I am worried about this Virginia projection. It depends critically on how Top 100 recruit, freshman point guard London Perrantes plays. But on paper, everything is coming together for Virginia. They have their three top scorers (really the only guys who scored at all), back from last year. Former Top 100 recruit Mike Tobey should be ready for that sophomore year leap in production. They add depth with South Carolina transfer Anthony Gill. Malcolm Brogdon (who missed the entire season) and Darion Atkins should be back healthy. And Virginia played very good basketball in conference games last year.

Georgetown: Greg Whittington should return from an academic suspension, incoming transfer Josh Smith is a dominant offensive player, and Georgetown brings back the majority of the players from a dominant defensive team. Thus picking the Hoyas for the Top 25 makes sense even without Otto Porter. But I am very concerned how much the offense will depend on mid-year transfer Josh Smith. Smith is a natural offensive player, and seems like the perfect high-post player to make the Georgetown offense flow. But will he be able to integrate into the lineup (and get in shape) for a 3 month run? Statistically he looks like a good fit, but betting on Josh Smith has been a bad bet the last two seasons.

Iowa: Iowa isn’t going to blow you out of the water with Top 100 recruits. But the majority of the lineup that went to the NIT final will be back. The time for winning is now. Don’t forget that the team welcomes Wisconsin transfer Jarrod Uthoff.

Ohio St: The loss of DeShaun Thomas shouldn’t hurt the Buckeye defense that much. And because players like Sam Thompson and LaQuinton Ross are only starting to show signs of how good they can be, the offense might not slip that much either.

Colorado: Another very young but talented team last year. People are raving about how good Josh Scott could be in 2013-2014.

Memphis: The five Memphis seniors are great players and this might be Josh Pastner’s best team. But the addition of a dominant recruiting class should be the x-factor that gets the team deeper into the tournament this year. If Tarik Black transfers (which is speculated), Memphis might fall a little bit in these projections, but not that much. This team has depth.

Creighton: They are only here if Doug McDermott comes back.

Connecticut: The best thing about your seniors transferring out is that all your main talent has eligibility left.

Syracuse: On paper they have the pieces they need to replace their departing stars. Duke transfer Michael Gbinije should be a nice wing player to replace James Southerland. Tyler Ennis can step in at point-guard. But if Michael Carter-Williams isn’t back (what I project above), Trevor Cooney is still going to be logging major minutes. The lack of options on the perimeter is still a concern.

UCLA: The Wear twins are going to be seniors next year. It only seems like we’ve been talking about them forever because of the transfer season.  I’m assuming Kyle Anderson sticks around, because where is he going to go? Between Anderson, a healthy Jordan Adams, and a more developed Tony Parker, UCLA has enough talent to compete for a Pac-12 title again. This may seem like a low ranking given what UCLA brings back, but remember UCLA was only 47^th in the Pomeroy Rankings last year. I’m predicting they will be better on offense and defense.

Alabama: This might be my most unorthodox Top 25 pick. Anthony Grant teaches outstanding defense, so Alabama will always be competitive in that area. And Bama brings just about everyone back, so expect a more coherent offense from the start of the season. There is no excuse for losing to Mercer or Tulane this season. The model particularly expects Devonta Pollard to take a big step forward. The freshman forward often looked unprepared last year, but he was a Top 25 recruit out of high school, and he should develop into a more polished post player. Plus Alabama adds two more Top 100 forwards to the rotation, which should help as well. My biggest concern is that far too often in recent seasons, Grant’s returning upperclassman haven’t delivered. Last year it was Andrew Steele’s injuries that caused him to fall apart and miss the end of the season. In previous seasons it was off-court issues that derailed the team. But on paper, there is no reason Alabama shouldn’t take a step forward and become an elite team next year.

Pittsburgh: It really helps that Pittsburgh can develop underclassmen to replace its seniors. Tray Woodall is gone, but James Robinson played point guard very well last season, and should keep the team playing at a high level.

New Mexico: The starting five from the MWC champs could have been back if not for the early entry of Tony Snell. But even without Snell or head coach Steve Alford, the experience the other four starters have working together should ensure another strong season. New head coach Craig Neal might not have the most credentials, but all he has to do is make sure the starters don’t transfer, and New Mexico will compete for a MWC championship again.

Tennessee: The preseason loss of Jeronne Maymon derailed Tennessee’s plans. But with Maymon, Jarnell Stokes, Trae Golden, and Jordan McRae, Tennessee has an impressive lineup coming back.

Kansas: A very good comparison for the 2014 Jayhawks might be the 2013 Tar Heels. I fully expect Bill Self to make Perry Ellis into a high volume scorer next season, just as I thought James McAdoo would become a star last year. And there are some impact recruits like Wayne Selden. But this is still a complete rebuild on offense. Bill Self’s teams always play outstanding defense, and he’ll need that to stay in the Top 25. Kansas could desperately use Andrew Wiggins to commit to the school if they want to win a 10^th straight Big 12 title, but it helps that the rest of the conference should be down next year.

Missing

Wichita St.: Last year Gregg Marshall returned only 35% of his minutes so with 54% of his minutes returning this year, winning should be a piece of cake. Seriously, if he can build another dominant team with unheralded recruits, he might be the best coach on the planet.

Indiana: The Hoosiers have a lot of good pieces, but the model doesn’t like how they fit together. It concludes based on last year’s substitution patterns that Yogi Ferrell, Jeremy Hollowell, Remy Abell, and Will Sheehey will see plenty of action. And they just aren’t the most effective offensive players. Ideally the younger players will see a lot of playing time early, and some of them will become stars, but it should be a transition year.

Miami: The Hurricanes don’t have enough incoming talent to replace what they lose.

Providence: The model questions whether Ed Cooley is a good enough defensive coach. Personally, based on what Providence did late in the year, I think they probably will have a solid defensive team next season. But getting elite talent to commit to defense isn’t easy. It certainly didn’t happen in January this year. And if Ricardo Ledo goes pro, they aren’t even in the conversation.

VCU: Troy Daniels (and his 124 made three pointers) is gone, and this isn’t a team loaded with elite talent. But the system should keep them in the Top 25 discussion again next year.

St. Louis: Very close.

Gonzaga: Sam Dower was a solid backup forward, but it is really asking too much for him to replace Kelly Olynyk and Elias Harris. Losing two high volume shooters is going to lower the efficiency of all the returning players.

Oregon: They lose a ton of value with their seniors. They could have really used that Aaron Gordon commitment.

Mississippi: Marshall Henderson is back, but three key starters are gone.

UNLV: Mike Moser may be transferring to Washington, but even if Moser comes back, they lost a lot of production this off-season.

St. John’s: D’Angelo Harrison says he wants to return to the team, and they will be better than people think.

Maryland: Even without Alex Len they will be an upper division ACC team.

Villanova: Lots of solid pieces.

Coming in the next three months:

1)      Updated rankings when early entrant decisions become final.

2)      Where the lineup-based model went right and wrong in year one, and how it can be improved.

3)      Conference projections for the top conferences. See the shockingly low projections for teams like Temple and Oregon.

Inconsistency can be a good thing. Illinois is a borderline NCAA tournament team this year. But thanks to the peaks of their inconsistency, the Fighting Illini won the Maui invitational, won at Butler, and beat the top ranked Hoosiers. Sure, the Illini were blown out by 20 at home by Northwestern, but the season wouldn’t have nearly been as fun if the Northwestern and Indiana victories were flipped.

On the flip side, if your team is a national title contender, inconsistency is a bad thing. If you want to win six games in the NCAA tournament, you cannot afford to have lapses every few games. In the next table, I attempt to determine which teams are most likely to have those lapses.

I take the teams in the Top 16 of the Pomeroy Rankings and figure out how often they look beatable on the basketball court. Determining a cut-off for beatable is somewhat hard, but I decided to count any game where the team’s game-specific Pythagorean Rating would be less than 0.9000. (Pythagorean Rating adjusts for opponent and venue as on kenpom.com.) 

This cut-off is quite different from counting losses. A team doesn’t have to lose to have a bad game. Miami’s two-point win at Clemson and one-point win at Boston College were bad games even though the Hurricanes ultimately prevailed. Those games showed Miami was beatable, even when at full strength. Similarly, Michigan St.’s three-point home win against Louisiana-Lafayette was one of the worst performances of the season by an elite team, even if the Spartans were victorious.

On the flip side, by using the 0.9000 cutoff, a few losses will not be counted as bad performances. Michigan lost by three at Ohio St. and in overtime at Wisconsin, but neither of those count as bad games. Florida lost by one point at Arizona and Georgetown lost in OT to Indiana, but those do not count as bad games either.

One thing I noticed is that a lot of games against teams ranked 300+ in the rankings looked like bad performances. Even if you beat a team ranked 300+ by 20 points, that doesn’t look like a dominant game because everyone beats those teams by 20 points. But I don’t think a 20-point win should ever count against a team, so I’m throwing those games out. The table only includes games against teams ranked 1-300 in the Pomeroy Rankings.

Surprisingly, this metric paints a different picture than the Pomeroy rankings. While Indiana and Louisville have had disappointing moments this season, the truth is that both teams rarely look beatable. Indiana and Louisville have only played poorly about 15 percent of the time this year. Conversely, Kansas has looked beatable 43 percent of the time. And while Michigan St. has played better in-conference, the Spartans were simply pedestrian throughout the non-conference schedule. 

Next, I wanted to look for the common-denominator in these bad performances.  This can be different from an overall weakness. For example, Florida’s overall weakness is getting to the free throw line. But that does not explain what happens when the Gators play poorly. When the Gators play poorly the problem is that the defense isn’t playing at an elite level. In the five games listed above, the Gator’s opponents have an eFG% of 52 percent compared to 41 percent during the rest of the season. 

When Teams Play Poorly 

Indiana: The Hoosiers struggle most in slow-paced games where their opponents avoid turnovers. Presumably, the Hoosiers do not want to see a team like Notre Dame running the burn offense in their bracket. The Hoosiers have been nearly identical on offense in their good and bad games. But forcing turnovers is key to their defensive success, and when they don’t force turnovers, the Hoosiers are beatable.

Louisville: The 0.9000 cutoff is a little generous for Louisville. Both the Syracuse and Georgetown losses do not count as bad games. If we raised the cutoff to 0.9100, Louisville would actually have two more bad games on their ledger. Still, I think Louisville’s bad performances are illuminating. In the team’s four worst performances on the season (the loss to Villanova, a home squeaker against Kentucky, a home squeaker against Illinois St., and the loss to Notre Dame), Russ Smith has struggled to make good decisions with the basketball. Louisville needs Russ Smith to play well to win. Louisville also plays poorly when a lot of fouls are called in the game and they cannot force clean turnovers.

Michigan: If Russ Smith is an indicator of whether Louisville will win or lose, Tim Hardaway Jr. is equally important for Michigan. Hardaway had an ORtg of 19 in the team’s loss at Michigan St., an ORtg of 68 at Bradley, and an ORtg of 73 vs Penn St. The other issue for the Wolverines is offensive rebounds. Michigan is not a great offensive rebounding team, but in the Wolverines worst offensive games, they’ve seen their offensive rebounding percentage plummet from 34% to 26%. Finally, Michigan’s biggest strength is avoiding fouls, but against a team like Wisconsin, that doesn’t need to get to the free throw line to score, that asset was less valuable. Mid-majors like Creighton or South Dakota St. could be a nightmare for Michigan, because both teams can score without needing to get to the free throw line, and both teams do a solid job on the defensive glass.

Florida: When the Gators play poorly, it is usually about shooting. They shoot worse and their opponents shoot better. That’s a little hard to predict, but one thing to keep in mind, Florida does seem to do better when the pace is in the 50s. Missouri and Arkansas had success against Florida by upping the tempo. But this isn’t all about turnovers. Florida played poorly against Kansas St. in a game where they didn’t turn the ball over much at all. I think a team like Memphis that can push the ball up and down and attack before the Florida defense gets set, might be the true kryptonite for the Gators.

Duke: Duke’s five worst performances have come since Ryan Kelly went down. But before we get too excited about his return, remember that not all players return from injury and dominate. Mike Moser is still getting back into shape for UNLV, and we cannot guarantee that Ryan Kelly will return Duke to early season form. For Duke, throwing off the pace also seems like a big deal. They struggled in a very slow game at Boston College, and in very high-paced games at Maryland and Miami FL. With Kelly out and Seth Curry playing through an injury, I don’t think Duke wants to play high possession games quite like in past seasons.

Syracuse: Syracuse’s weaknesses are the typical ones with a zone defense. If a team has a forward who can dominate from the high post (see Davante Gardner), that team can carve up Syracuse. And as is typical with the zone, when Syracuse doesn’t grab defensive rebounds, the Orange do not play well. Syracuse struggled with defensive rebounding in narrow wins over Cincinnati and Detroit and losses to Pittsburgh, Connecticut, Villanova, and Temple. A strong offensive rebounding team like Colorado St. or Minnesota is presumably Syracuse’s worst nightmare. This year, without depth on the perimeter, Brandon Triche is also a concern. Triche isn’t a perfect predictor of Syracuse’s poor play (he did play well against Temple and Villanova), but when Triche is struggling, Syracuse also typically struggles. Finally, Syracuse struggles in slow-paced games this year. Because they are not a great half-court team, Syracuse needs transition opportunities to play well.

Pittsburgh: When Pittsburgh plays poorly, it is usually about the offense not working. Their adjusted offense is 123.1 in their good games and 104.9 in their bad games. Teams that have perimeter depth and can keep Tray Woodall, James Robinson, and Lamar Patterson out of the lane are Pittsburgh’s biggest problem. Guard-oriented teams from Rutgers to Cincinnati to Marquette to Michigan have been able to do that.

Arizona: The Wildcats might win the Pac-12, but seven of Arizona’s eight bad performances have come in conference play, so I am worried that this team is not peaking. Arizona is a team that really relies on getting to the free throw line, and teams that avoid fouls can be Arizona’s kryptonite. In the tournament, Arizona does not want to see a team that can defend well without fouling. Whether it would be a team with size (like Maryland) or just good fundamentals (like Wisconsin), there are a lot of teams that fit this profile.

Gonzaga: Gonzaga hasn’t been challenged much lately, so it is hard to find a stat that really does the job, but I’m going to focus on defensive rebounding. Since Gonzaga does not have a great eFG% defense, they cannot afford to give up second chance points. Teams like San Diego and Santa Clara hung with the Zags in conference play by crashing the glass. In most games Gonzaga grabs 72% of the defensive boards, but that percentage has dipped to 66% in their worst games. Gonzaga’s defensive rebounding also nearly cost them the game against Washington St. Gonzaga’s defense also has trouble containing great point-guards, but there are not enough of those in the WCC to really prove the point.

Ohio St.: Ohio St. has been Jekyll and Hyde defensively this year. In the 16 good games, the adjusted defensive efficiency has been 83.3. But in the team’s 9 worst performances on the year, the adjusted defense has only been 97.9. It seems like the Buckeyes have struggled more with perimeter-oriented-teams. Northwestern, Wisconsin, Indiana, and Illinois have all made the Buckeye defense look porous this year by shooting lights out. But you have to blame the Ohio St. offense a little too. When the team falls behind, because the offense isn’t good enough to score points in bunches, they have to take more chances defensively. And those gambles have led to some ugly final margins. Really, I wish someone could explain why Ohio St.’s defense is so bad in some games, because I really believe defense is Ohio St.’s strength.

Georgetown: Georgetown’s worst games have come at a slow pace against mediocre teams. See Liberty, Towson, and South Florida. They also usually involved a horrible offensive performance by the team, but that has been changing. Only one of the Hoyas bad performances has come in February, the narrow win at Rutgers.

Miami: Miami should probably be a little higher on this list, because some of the worst performances (like the Florida Gulf Coast game) came with personnel missing. Still, I think it is revealing that four of Miami’s worst performances have come in February. Miami may only have one ACC loss, but that doesn’t mean they have been dominant every time out. Much like Georgetown, Miami has struggled against some bad teams (Boston College, Wake Forest).

Kansas: Kansas has looked like an average team remarkably often this season. When the offense isn’t working, turnovers are almost always to blame. Kansas has to feed Ben McLemore and Jeff Withey to score consistently, and teams that can take the ball away on entry passes are potentially a big problem for the Jayhawks. A team like Robert Morris could be a nightmare first round opponent, given the Colonials ability to create turnovers. This is especially true given that Kansas has struggled against mediocre teams this year from Chattanooga to TCU.

Oklahoma St.: Oklahoma St. is not a great shooting team, particularly from the perimeter. So they need offensive rebounds and free throws to compete. Baylor and Virginia Tech avoided fouls and beat the Cowboys. Kansas St. crashed the defensive glass and did the same thing.

Michigan St: While most of the Spartans mediocre performances came in the non-conference schedule, the Spartans have struggled to put away Penn St. and Nebraska in conference play too. The story for Michigan St. always seems to be about turnovers, good or bad. When the Spartans force them, they win. When the Spartans give the ball away, they look bad.

Wisconsin: Wisconsin has seven games where they were completely dominant. In the blowout wins over Ohio St., Northwestern, Illinois, Cornell, California, and Southeastern Louisiana, and the big time win at Indiana the Badgers had a Pythagorean winning percentage above 0.9900. But over the rest of the schedule the Badgers have been remarkably pedestrian. As the numbers listed above show, Wisconsin has looked beatable on 14 occasions. It usually comes down to shooting. In Wisconsin’s 11 best games of the year, they have an eFG% of 54%. In the 14 games where they were beatable, the Badgers eFG% has only been 44%. Even if Wisconsin has snuck back into the Top 10 of Sagarin’s Predictor and Pomeroy’s rankings, with that kind of offensive inconsistency, they do not look like a Final Four team.

Click Here for Part 1.

Northwestern has been able to compensate for Drew Crawford’s lost offense, but they have not been able to compensate for the athleticism he brought to the defense. He was one of the only players with the quickness and size to frustrate opposing wing players.

Providence has had a nightmare of a season. First Kris Dunn was injured. Then Vincent Council was out. Plus, Bryce Cotton has missed games occasionally. And yet with one or more of those players out at the start of the season, Ed Cooley got his players to buy in and play team basketball. They knew they had to play great defense to have a chance, and the players responded. But now that everyone is at full-strength, the defensive commitment just isn’t there. Pretty much from the moment that Kris Dunn debuted, Providence has stopped playing defense. And the team has gone 2-6 since then.

North Dakota St. was one of the true surprises of the early season. They were extremely dominant and looked like a real challenger to Nate Wolter’s South Dakota St. team for the Summit League title. But the injury to leading scorer Taylor Braun has completely derailed NDSU’s offense.

This next group features three players whose absence was not a huge problem. Georgetown sat at 0-2 in the Big East and had just been blown out at home by Pittsburgh when they found out Greg Whittington was academically ineligible. Since then the Hoyas have gone 4-1 including an upset of Louisville over the weekend. Georgetown’s improvement without Greg Whittington is a little bit of a surprise, but likely has two causes. First, with Whittington academically ineligible, the Hoyas have stopped rolling out one-guard lineups. The offense has simply moved more crisply with two guards on the floor at once.  I appreciated John Thompson III’s creativity with four forwards on the floor, but there is a reason most teams don’t do that. Second, un-related to the suspension of Whittington, head coach John Thompson III has stopped trying to run the offense through center Michael Hopkins. Hopkins was the team’s least efficient offensive player, and Hopkins playing time has plummeted in recent games.

Memphis had its best offensive day of the season (against Tennessee) when Tarik Black was out, but I wouldn’t necessarily be convinced that the team is better off without him. That seems like a bit of a fluke.

North Carolina hasn’t really missed Leslie McDonald and I don’t think anyone is surprised. North Carolina has a plethora of two-guard/wing players. What the Tar Heels are missing is a second legitimate post-player to compliment James Michael McAdoo. PJ Hairston and Reggie Bullock were well-positioned to replace McDonald’s playing time.

I took a little liberty and grouped the North Carolina game in the “Moser-injured” category because he clearly wasn’t 100%. The good news for UNLV fans is that they are a better team with Moser in the lineup. The bad news is that even with the team at full-strength, they have only been a borderline Top 25 team. UNLV has Top 10 talent, and Anthony Bennett might be the best freshmen in the country, but they haven’t quite been able to put it all together.

Virginia’s splits aren’t really about Evans so much as they are about a team that has improved throughout the season. The Cavaliers started out playing mediocre basketball, began to play better defense once they added a PG with the quickness of Evans, and even with Darion Atkins going down they have continued to improve. That last split probably isn’t sustainable, but Virginia is the rare team that can say they are playing much better now than in November.

Although you might not guess it from their 8-6 record, the Washington Huskies have been at their best at full-strength.

Durand Scott missed Miami (FL)'s loss to Florida Gulf Coast early in the season. And Reggie Johnson just missed a string of games and hasn’t quite looked 100% since returning. Miami has only had their full-lineup for 7 games, but when you include a blow-out win against Duke in that sample of seven games, you look pretty good.

I’m not sure mid-season transfer Jamal Branch has been a negative influence on St. John’s. But he just so happened to miss St. John’s best win of the season, at Cincinnati, with an eye injury. And he was around for St. John’s embarrassing home loss to UNC Asheville.

Meanwhile, I feel for Creighton’s Josh Jones having to give up basketball due to a heart condition. But as crazy as it sounds, his 109 ORtg was actually a drag on Creighton’s fine offense. The Blue Jays are in the rare position of having an embarrassment of riches in their rotation right now.

After NCAA tournament hero CJ McCollum went down, I thought Lehigh would have to cede the Patriot league title to Bucknell. But by buckling down on defense, Lehigh won its first five games against D1 competition without McCollum, and shockingly picked up the road victory at Bucknell. They actually looked like they might be the Patriot league favorites after all. But a home loss against Lafayette emphasized what was already true. Without McCollum, Lehigh’s offense isn’t nearly as good. Perhaps by emphasizing their new defensive identity, they can still win the league. But it isn’t going to be nearly as easy without their star.

Kwamain Mitchell’s ORtg is only 88.9, and he is one of only two St. Louis players with a sub-100 ORtg. Given how the team played without him, I think it is worth asking whether he is hurting the St. Louis offense.

Andrew Steele can’t be as important as the numbers here show. But his absence corresponded to the team’s December swoon. And I’m certainly willing to believe that the senior guard is an important cog to the team’s offensive and defensive chemistry. I just know that given his limited role on the team, there are other factors at play here.

Finally, Syracuse was playing like the 2^nd best team in the nation in the games before the Southerland academic suspension. Since then, the upset loss to Villanova has the Orange playing more like the 15^th best team in the nation. Still, that loss to Villanova felt like one of the annual games where an opponent got hot from three. Syracuse usually loses at least one of those. I’m not sure we have an accurate feel for the Orange without Southerland yet.

Final Notes: The following injuries are not worth a split since they were so short, but they will likely show up on NCAA scouting reports at the end of the year.

-Mike Bruesewitz missed Wisconsin's loss to Marquette.

-Justin Jackson missed Cincinnati's loss to Notre Dame. Given Jackson’s general struggles this year, I'm not sure anyone can count that as meaningful, but I'm sure it will show up on the NCAA committees report from the Big East. 

-Brady Heslip missed Baylor's loss to Charleston. I watched this game and his absence really seemed to hurt the flow of the Baylor offense.

-Xavier Thames missed San Diego St.'s loss to Wyoming.

-Dwayne Davis missed Southern Miss's loss to Arizona with the flu. It didn’t seem that important at the time, but given how well Southern Miss is playing in CUSA this year, it could matter come selection Sunday.

-Roy Devyn Marble missed Iowa's 3 point loss to Michigan St.

-PJ Hairston missed North Carolina's loss to Indiana.

-Sam Grooms missed Oklahoma's loss to SF Austin.

-Cory Remekun missed St. Louis's loss to Santa Clara.

-Terone Johnson missed Purdue’s loss to Bucknell.

-Rotnei Clarke missed Butler’s loss to La Salle.

-Milton Jennings missed Clemson's loss to Purdue.

-Nate Wolters missed South Dakota St.'s loss to Minnesota.

-Akron was missing Demetrius Treadwell and Nick Harney in their season opening losses to Coastal Carolina and Oklahoma St. The latter came in OT so the NCAA may very well care that these players were missing.

-Stanford was missing Aaron Bright for the Missouri and Minnesota losses, but oddly the team has played worse basketball since he has returned. Perhaps his 23% three point shooting has something to do with that.

-Finally, I'm not going to say much about players who missed the opening game or opening three games (Shabazz Muhammad) because with the recent games weighting on Kenpom.com those games have already been deemphasized. But there remain a number of recent injuries to keep an eye on in February. Oregon's Dominic Artis, UCLA's Travis Wear, Notre Dame's Scott Martin, Boise St.'s Jeff Elorriaga, and Texas's Jonathan Holmes have only recently joined the inactive list.