From Sweet To Elite To Final Four

Sixteen teams remain in the quest for the NCAA National Basketball Championship. VandyMania's Howell Peiser provides us with a statistical analysis of the Sweet 16 teams and gives us a prediction of who will be the last dancing. Howell also compares this year's Vanderbilt team to the final 16 teams.

Last week, I discussed what I thought were the most vital statistics the NCAA tournament teams who advanced to the Final Four possessed.  The historical averages for the last 40 Final Four teams were:
FG%; 47.7
Defensive FG%: 40.1
Rebound Margin: +5.8
Turnover Margin: +1.8
Rebound + Turnover Margin (R+T): 7.6
Scoring Margin: 13.3
Conference Strength: ACC, Big East (includes former C-USA teams now in the Big East), Big 10, Big 12, SEC, PAC-10
I mentioned the teams that looked to best fit the criteria.  With the field reduced to 16, exactly half of the remaining field possesses statistical averages in the ball park of the standards I put forth in my hypothesis.  Connecticut, Texas, Florida, Washington, Memphis, Boston College, LSU, and UCLA are still alive.  Four teams, including two number one seeds (Duke, Villanova, West Virginia, and Gonzaga) are still alive even though they do not meet the statistical expectations.  The remaining four teams (Bradley, George Mason, Wichita State, and Georgetown) fall in a gray area; they meet several, but not all of the criteria or come from a mid-major conference.  The last mid-majors to make the Final Four were Penn and Indiana State in 1979.  Indiana State fell in the category of mid-major but top 5 team, just like Gonzaga this year.
Note: The Missouri Valley Conference is on the precipice of becoming a Major conference once again.  Going back 40 to 60 years ago, this conference produced NCAA and NIT champions as often as the ACC and Big 10.  MVC members Cincinnati and Bradley used to be on par with today's Connecticut and Villanova.  Louisville, Drake, and Wichita State all advanced to the Final Four as members of this league.  Indiana State and Memphis State made it to the championship game as MVC members.
I am going to add one more statistic to the hypothesis this week.  Average steals per game is an important stat to determine how much weight to give to turnover margin.  If a team forces turnovers by picking off the ball, it gives more weight to those turnovers forced.  A steal holds more weight than any other turnover or a defensive rebound.  Steals are converted into a higher percentage of points on average.  Defenses do not have the time to respond to this type of transition; frequently the offense has a head start on its journey to the basket and almost always has numbers or an even-numbered break opportunity following a steal.
The magic numbers here are 10 and 7.5.  A team that averages 10 steals per game is in rarified air.  A team that averages at least 7.5 steal per game is one to worry about.  Combine 7.5 steals with an R+T of at least 7.5 as well, and you have a team that is 10 points better than opponents with an R+T of 0  before any other stat is involved.  Those numbers equate to five to 10 extra shots per game.
Today, I am going to break down the numbers for each of the eight games to be played Thursday and Friday and the expected four games on Saturday and Sunday.
Atlanta Region
Duke vs. LSU
FG%:  Duke .494 to .430 for a difference of 6.4%
             L S U  .472 to .406 for a difference of 6.6%
Rebounds:     Duke -2.7      LSU +8.8
Turnovers:    Duke +2.5     LSU -0.7
R+T:               Duke -0.2     LSU +8.1
Steals:            Duke 9.4        LSU 8.3         
Scoring:         Duke +13.6  LSU +10.0
SOS:               Both teams come from historical Final Four conferences
Duke has one of the toughest SOS's in the nation
                        LSU has an excellent but less impressive SOS than Duke
Prediction: LSU is primed for the upset here if they do not commit 18 or more turnovers and allow Duke to score easy baskets off steals.  The Blue Devils should find themselves out-muscled in the paint.  Glen Davis and Tyrus Thomas should score enough on put backs to provide the decisive margin. Shelden Williams and Josh McRoberts will not be able to stop the two large Tiger stars.
J.J. Redick may score 20+ points in this game, but I believe he will find the going tougher than he has in Duke's first two tournament games.  He could put up an 8 for 22 effort.  Darrel Mitchell and Garrett Temple have the quickness and the muscle to contain Redick just enough to force Greg Paulus and Sean Dockery to contribute 25 combined points just to stay in the game.
In an upset, I pick LSU to win by 4 to 8 points.
West Virginia vs. Texas
FG%: West Virginia  .445 to .448 for a difference of -0.3%
            Texas  .466 to .381 for a difference of 8.5%
Rebounds:     West Virginia -9.4   Texas +10.3
Turnovers:    West Virginia +7.2  Texas +0.6
R+T:               West Virginia -2.2   Texas +10.9
Steals:                        West Virginia  8.2    Texas  8.0
Scoring:         West Virginia  7.6    Texas  15.9
SOS:               Both teams come from historical Final Four conferences
                        West Virginia clearly has a tougher SOS than Texas
Prediction: I smell a blowout here.  West Virginia will have to shoot lights out from behind the arc and get to the foul line enough to score about 10 more points there.  I just don't see it happening.  Texas is clearly better by leaps and bounds and should dominate this game.  LaMarcus Aldridge, P. J. Tucker, and Brad Buckman are just too big and strong for the Mountaineers.  The Longhorn frontcourt trio is quick enough to contain Kevin Pittsnogle, Mike Gansey, and Frank Young.  The backcourt isn't as big of a mismatch, but the ‘Horns have the advantage here as well.  I think turnovers will be close to even, while Texas controls the boards by a large number, maybe as much as 15.
This all adds up to a huge win for the Longhorns—the type that allows Coach Rick Barnes to give his starters a little rest prior to Saturday's Region Final.
I pick Texas to win by 15 to 22 points.
Region Final:  Texas over LSU by 7 to 10 points.
Oakland Region
Memphis vs. Bradley
FG%:  Memphis       .453 to .382 for a difference of 7.1%
             Bradley         .442 to .400 for a difference of 4.2%
Rebounds:     Memphis  +6.7         Bradley +3.2
Turnovers:    Memphis  +2.3         Bradley +1.7
R+T:               Memphis  +9.0         Bradley +4.9
Steals:            Memphis  +9.8         Bradley  +9.1          
Scoring:         Memphis +15.1        Bradley  +7.1
SOS:               Neither team comes from historical Final Four conferences
Memphis qualifies based on their top 5 ranking
Bradley actually has a much better SOS than Memphis
Prediction: This game may be interesting, at least for a while.  Bradley has the talent to challenge for a Final Four berth if the ball bounces just right for them.  The Braves have a credible inside game with big Patrick O'Bryant and Marcellus Sommerville. 
The problem is the Braves may be quick enough and strong enough to contend with most NCAA-caliber teams, but they don't have the overall quickness and other abilities to do so with a Memphis squad if the Tigers plays up to normal standards. 
Rodney Carney and Darius Washington are more athletic than the Braves can handle.  Eventually, the Tigers will assume control of this game and win going away.  Don't overlook the Memphis defense.  When you hold opponents to 38.2% shooting, you are one of the best.  This Tiger team can win the national title if they bring their "A" game four more times.  If not, they may only advance one more round.
I pick Memphis to win by 12 to 16 points.
Gonzaga vs. UCLA
FG%:  Gonzaga       .478 to .433 for a difference of 4.5%
             UCLA             .483 to .418 for a difference of 6.5%
Rebounds:     Gonzaga  +3.6         UCLA  +4.9
Turnovers:    Gonzaga  +0.5         UCLA  -0.2
R+T:               Gonzaga  +4.1         UCLA +4.7
Steals:            Gonzaga  +7.2         UCLA  +5.8  
Scoring:         Gonzaga +7.6          UCLA  +10.0
SOS:               UCLA comes from a historical Final Four conference
                        Gonzaga qualifies based on their top 5 rating
                        UCLA has a somewhat stronger SOS
Prediction:  Neither team has a commanding advantage over the other team, so this should be a close ball game.  Adam Morrison has the flu, and as a diabetic, there can always be complications with the immune system.  It is unsure how much he will be able to contribute; some players actually play better with a slight fever, while some players have difficulty putting on their shoes when they are sick.
Even if Morrison is near full strength, I expect him to struggle against the Bruin defense.  Jordan Farmar and Arron Afflalo get so much publicity about their offensive skills, their excellent defensive skills often get overlooked.  The Uclan backcourt can slow the national scoring leader and keep him under 20 points.
Where Gonzaga has the advantage is in the low post.  J. P. Batista can score 20 points and pull down 10 rebounds against any other post player.  He can do it so efficiently, that opponents, commentators, and fans don't realize he has hit five quick baskets in five to eight minutes of action.  Bruin center Ryan Hollins will have his hands full.
Where I think UCLA will win this game is on the offensive end.  The Bruin guards can penetrate the Gonzaga defense and either score, dish to open teammates, or draw fouls.  Gonzaga doesn't have a defensive answer for Farmar and Afflalo.
I pick UCLA to win by 8 to 12 points.
Region Final: Memphis edges UCLA by 2 to 5 points.
Washington, D.C. Region
Connecticut vs. Washington
FG%:  UConn .472 to .378 for a difference of 9.4%
             Washington .472 to .433 for a difference of 3.9%
Rebounds:     UConn  +10.1           Washington  +6.3
Turnovers:    UConn  -0.4              Washington  +3.0
R+T:               UConn  +9.7                         Washington  +9.3
Steals:            UConn  +6.6                         Washington  +7.9   
Scoring:         UConn  +15.4           Washington  +12.7
SOS:               Both teams come from historical Final Four conferences
                        UConn has a considerably stronger SOS
Prediction: For my money, this is the most exciting match up in the Sweet 16 round.  The winner of this game will be a huge favorite to make it to the Final Four on Saturday.  Both teams have the talent and statistics to cut down the nets in Indianapolis.
Connecticut has not played up to their standards since the end of the regular season victory over Villanova.  They looked lackluster in wins over South Florida and Louisville in the Big East Tournament, and they looked more like the number six seed in wins over Albany and Kentucky to advance to this point.  If Connecticut doesn't play the way they are capable of playing, Washington will send them back to Storrs.
The Pac-10 Huskies are more than just all-American guard Brandon Roy.  Washington has excellent sixth and seventh men in forward Jamaal Williams and guard Ryan Appleby.
The Big East Huskies have the top mix of talent in the Big Dance.  Forward Rudy Gay is the Roberto Clemente of college basketball.  He isn't the best at any one facet of the game, but he is among the top 10 in all of them.  Forward Josh Boone compliments Gay well.  With Rashad Anderson coming off the bench, UConn probably has the best sixth man left in the tournament.
So, the 64,000 dollar question is, "Will UConn return to their regular season form?"  If the answer is "yes," then we know who the national champion will be.  If the answer is "no," then warm up the plane for a return trip to The Constitution State.  It's anybody's guess, but I sense that UConn lays another egg and gets beat.
I pick Washington to pull off the upset by 3 to 5 points.
George Mason vs. Wichita State
FG%:  GMU              .484 to .386 for a difference of 9.8%
             Wichita St.    .445 to .401 for a difference of 4.4%
Rebounds:     GMU  +2.9    Wichita St.  +5.5
Turnovers:    GMU  +0.7    Wichita St.  -0.6
R+T:               GMU  +3.6    Wichita St.  +4.9
Steals:            GMU  +7.2    Wichita St.  +5.3     
Scoring:         GMU  +10.4 Wichita St.  +7.8
SOS:               Neither team comes from historical Final Four conferences
                        Neither team qualifies based on ranking
                        Wichita State has a slight SOS advantage
Prediction:  So, you missed the Bracket Buster a month ago, hmm?  Well, here's your chance to see it again.  This time, the Colonials get a tiny bit of home team advantage.  GMU won at Wichita in February's Bracket Buster, and that is more than likely why they earned an at-large ticket to the dance.
When you look at the statistical data, GMU looks like a team capable of getting to the Elite Eight.  They are clearly a better team than the Kent State team that made its way to the Elite Eight a few years back and made a household name of Coach Stan Heath.
GMU's defense is interesting to watch.  Coach Jim Larranaga is one of the foremost advocates of the scramble defense.  The Colonials can trap out of their half-court man-to-man as well as any other team.  It frequently takes opponents out of their normal offense and can disrupt most disciplined teams.
Wichita State has had a chance to face GMU's scramble defense, so the Shockers should be better prepared to attack it and get the ball inside to Paul Miller, P.J. Couisnard, and Kyle Wilson.  Guard Matt Braeur has the passing skills to render most traps useless.
The only thing keeping me from taking Wichita State in the rematch is the fact that GMU will be playing less than 30 miles from home.  Give GMU about three points for home team advantage.
I pick George Mason to win in a squeaker reminiscent of their previous battle with Wichita State.
Region Final: Washington over George Mason by 6 to 10 points.
Note: If Connecticut returns to form, they waltz to the Final Four with two double-digit wins.  I just have a gut feeling they will continue to play sub-standard ball. 
Minneapolis Region
Villanova vs. Boston College
FG%:   Nova .426 to .420 for a difference of 0.6%
             B C     .485 to .429 for a difference of 5.6%
Rebounds:     Nova  +1.5    B C  +3.8
Turnovers:    Nova  +4.4    B C  +0.2
R+T:               Nova  +5.9    B C  +4.0
Steals:            Nova  +7.7    B C  +6.2      
Scoring:         Nova  +11.4 B C  +8.3
SOS:               Both teams come from historical Final Four conferences
                        Villanova has a much stronger SOS
Prediction:  I thought Villanova would be the first number one seed to be sent packing, but the Wildcats survived a tough game with Arizona.  The Wildcats may survive for yet another round Friday night, because I don't think Boston College has the horses to match ‘Nova man for man.
Three of the four Villanova guards (Allan Ray, Randy Foye, and Mike Nardi) can knock down the long-range jumper and penetrate to the hoop.  All four (add Kyle Lowry to the group) are deadly at the foul line. 
Boston College has the inside power to counter the Villanova attack.  Forwards Craig Smith and Jared Dudley are both multi-skilled big men who can dominate a game inside and step outside and hit the jumper.  The Eagles can win by successfully isolating the tandem inside and letting them score against smaller defenders.
Where Villanova has a huge advantage is in overall speed and quickness.  The smaller Wildcats may be quick enough to keep the ball away from the low post and force enough turnovers to enjoy a large advantage in fast break points and points off turnovers.  I think ‘Nova survives for one more round.
I pick Villanova to win by 5 to 8 points.
Florida vs. Georgetown
FG%:   Florida                     .508 to .405 for a difference of 10.3%
             Georgetown            .471 to .420 for a difference of 5.1%
Rebounds:     Florida  +3.5             Georgetown  +3.9
Turnovers:    Florida  +1.6             Georgetown  +0.6
R+T:               Florida  +5.1             Georgetown  +4.5
Steals:            Florida  +8.0             Georgetown  +6.0  
Scoring:         Florida  +15.0          Georgetown  +7.4
SOS:               Both teams come from historical Final Four conferences
                        Georgetown has a considerably stronger SOS
Prediction:  Florida has looked as good as anybody in their current seven game winning streak.  The Gators are good on both sides of the ball.
Florida has the look of a Final Four team and possible National Title contender, but they have an Achilles heel.  Point guard Taurean Green has lapses in nearly every ball game.  He almost cost his team the Southeastern Conference Tournament Championship with some unwise shots in the final minute.  Put South Carolina's Tre Kelley with the other four Gator starters, and Florida would be right up there with Connecticut and Texas.
Georgetown doesn't have the talent of two decades ago, but the Hoyas are not too far away from being that good once again.  The problem in this game is that Florida loves to play against teams that run "The Princeton" offense.  Ask Vanderbilt how effective Florida's pressure defense was at stopping their version of the Princeton offense.
Look for Florida to make at least one big run in this game, where they score eight or more points, while Georgetown fails to get a basket.  That will be the difference.
I pick Florida to win by 8 to 14 points.
Region Final: Florida over Villanova by 6 to 10 points
How The Computers See The Sweet 16 and Elite 8
The computers really fall short during the NCAA Tournament.  They have the teams closely bunched together creating unlikely predictive outcomes.  Nearly every game is expected to be decided by two or three points.
Therefore, I have done something a little unorthodox this week.  Rather than throw out the ratings that seem to be out on a tangent when compared to the rest, I am going to include them at equal weighting. 
Sweet 16 Round
Duke 77  LSU 70
Texas 74  West Virginia 69
Memphis 80  Bradley 69
UCLA 70  Gonzaga 62
Connecticut 84  Washington 74
George Mason 69  Wichita State 68
Villanova 72  Boston College 66
Florida 65  Georgetown 60
Elite 8 Round
Duke 77  Texas 73
Memphis 73  UCLA 68
Connecticut 79  George Mason 66
Villanova 74  Florida 71
How did Vanderbilt compare with these 16 teams statistics-wise?
Field Goal %: Vandy shot .460 and allowed .443 for a difference of 1.7%
Rebounding: Vandy was out-rebounded by 1.2 per game
Turnovers: Vandy enjoyed a +0.9 turnover margin
R+T: Vandy recorded a disappointing -0.3
Steals: Vandy averaged 6.1 steals per game
Scoring Margin: Vandy outscored opponents by 3.1 points per game
SOS: Vandy's schedule was among the 20 most difficult
What the Commodores must do statistically to get back to the Big Dance in 2007:
Field Goal %: Vandy's offensive percentage was adequate this season, and it would not need to improve by more than one percent.  Defensively, the Commodores surrendered too many easy shots.  They must lower their .443 allowed to at least .429 next year.
Rebounding: This may be an even bigger liability next year unless Alan Metcalfe and Ross Neltner emerge as monsters on the boards.  Vandy needs to out-rebound their opponents by at least one half board per game.
Turnover Margin:  Vandy needs to force two to three more turnovers per game next year.  With even more depth on the perimeter next year, there is reason to believe the occasional full court pressure we saw at the end of this year will be used more frequently next season.
R+T: Plain and simple, this must improve to the positive side and top at least 5.0. 
Steals: This is how Vandy can improve its R+T.  The Commodores haven't enjoyed much success in this department in the 21st century.  Next year, the Gold Men should have the talent to top eight steals per game and get more fast break points off these forced turnovers.
Scoring Margin:  The Commodores had great difficulty this year in putting opponents away after gaining the lead.  This must be rectified next season.  Vandy needs to outscore opponents by at least five points per game and preferably eight points per game. Top Stories