# A Scientific Look At March Madness

Okay folks, tell me the truth. How many of you reading this message right now are not at work today and won't go tomorrow either? How many of you anxiously await the start of that theme music CBS Sports will blast? It now joins the themes to Wimbledon, The Olympics, and horse racing's My Old Kentucky Home, Maryland My Maryland, and New York New York (I wish the Belmont Stakes would bring back The Sidewalks of New York.

If you still have time to read something before filling out your bracket or you just want to look at one more primer telling you who should wind up in the Final Four, then read on.  I am going to break down the game scientifically by looking at trends from the last 40 Final Four participants.  There are some constant statistical information that the last 10 years of Final Four participants share.  Let's look at the pertinent stats, and then find the teams that meet those standards this year.

1. Field Goal Percentage and Defensive Field Goal Percentage

The average field goal percentage amongst the last 40 Final Four teams is 47.7%, while the average defensive field goal percentage is 40.1%.  The average difference in field goal percentage offense and defense is 7.6%.  What this means is that a team must not only be able to consistently stop the other team, but they must also be able to shoot the ball well on offense.  Look for teams that tend to get more baskets from inside five feet than they allow.  That's where much of this discrepancy exists.

2. Rebounding Margin:  In the last 10 years, the average rebounding margin for the 40 Final Four teams is +5.8.  39 of the 40 teams out-rebounded their opponents; 32 of them were +3 or better.

3. Turnover Margin:  The average turnover margin amongst the last 40 Final Four teams is +1.8.  32 of these teams had positive turnover margins.

4. Combination of Rebound and Turnover Margin.  Much like the baseball statistic OPS (On-base percentage plus slugging), the R+T Margin is more important than just rebounding or turnover margin.  Most Major League teams give a little more weight to on-base percentage than slugging.  I give a little more weight to turnovers over rebounding.  Steals result in easier baskets for a team than rebounding, so if a team has a lower than the norm rebounding margin, a higher than average turnover margin will compensate.

The average R+T margin of the last 40 Final Four teams is 7.6.  All 40 finished on the plus side here.  36 finished +3 or better, and 33 finished +5 or better.  When a team averages +5 or better in R+T, it's like getting five extra possessions.  Averaging one point per possession, that's five extra points.  In NCAA Tournament play, that's a big plus.

5. Scoring Margin: This should be obvious.  When a team averages more points than they allow, they usually are a winning team.  What we are looking for here is a team that averages better than 10 points more than they allow.  The average scoring margin of the last 40 Final Four teams is 13.3.  32 of the last 40 Final Four teams had 10+ scoring margins, and 35 had scoring margins of 8+.

6. The last 40 Final Four teams included 10 from the ACC, nine from the Big 10, five each from the Big 12 and SEC, four from the Big East and two more from current Big East teams that were once in C-USA, three from the Pac-10, and one each from the Atlantic 10 and Mountain West.  That means the power conferences placed almost every Final Four team in the last 10 years.  The only exceptions were Massachusetts and Utah, two teams that were highly-rated all year the season they made it to the Final Four.  Gonzaga fits in that category this year, while George Washington is on the cusp.

A few statistics that appear not to have much if any bearing on picking Final Four teams are Assist to Turnover Ratio, Foul Shooting, and Three-point Shooting.  Actually a higher assist to turnover ratio appears to be one of those over-rated stats.  Teams that get a lot of offensive put backs score baskets without earning assists.  Offensive rebounding is much more important and plays a much larger roll in deciding winners.

Which teams meet these qualifications?  Off the bat, we can throw out some of the heavy hitters that appear to be paper tigers this year.  Duke is -3 on the boards this year and only +2.8 in turnover margin.  It looks like the Blue Devils will bite the dust before they can win four times.

Villanova only out-shot their opponents by 0.5% and only out-rebounded them by 1.4.  While they did hold a +4.4 turnover margin, it looks like they fail to meet the standards as well.

Tennessee was out-shot by 0.8%, out-rebounded by 3.2, and only outscored their opponents by 7.4.  It looks like the Vols will have to wait another year to advance beyond the Sweet 16.

Michigan State out-shot their opponents by 4.5%, and outscored them by 5.8.  While, they did okay on rebounding, they committed 1.2 more turnovers per game than they forced.  This doesn't look like Tom Izzo's year to repeat.

West Virginia looks like a team that could be upset early.  They were out-shot by 1%, out-rebounded by an unbelievable 9.7, and outscored opponents by just 6.1.  They own an excellent turnover margin of +7.2, but it comes about because of few offensive turnovers.  They don't get many points off opponent turnovers.

Ohio State has a low +1.7 R+T rating and only out-shot opponents by 4.5%.  That's not enough cushion for you this year Thad.

What about Gonzaga?  I don't see them getting to the Elite Eights, much less the Final Four.  They out-shot opponents by just 4.2%, posted an R+T of +4.1, and outscored their opponents by a meager 6.7.  These are not the stats any of the mid-major powers of yesteryear posted when they advanced to the Final Four.  Adam Morrison will not repeat the heroics of Larry Bird.

Okay, which teams appear to have the best stats when it comes to mimicking the last 40 Final Four teams?  Defending champion North Carolina looks good.  They out-shot opponents by 6.9%, out-rebounded them by 8.9, and outscored opponents by 11.5.

Connecticut did much better than that.  They have the look of a true Final Four team and one that resembles their previous two National Champion seasons.  They out-shot their opponents by 9.5%, out-rebounded them by 10.5, and out-scored them by 15.9.  It will take a near-perfect game by a really good team to beat them.

Texas posted some strong results.  They out-shot opponents by 8.4%, out-rebounded opponents by 10.1, and outscored them by 16.  The Longhorns have the look of a Final Four team.

Florida looks like the SEC's best chance for placing a team in the Final Four.  The Gators shot 50.8% for the season compared to 40.9% allowed for a 9.9% difference.  Their R+T was +4.7, and they outscored opponents by 14.6.

Washington has not looked all that sharp as of late, but they have some fantastic numbers.  They out-shot opponents by 4.2%, posted an R+T of +10, and outscored opponents by 13.

Memphis is worthy of their ranking.  The Tigers out-shot opponents by 7%, posted an R+T of 9.8 (10 more possessions!), and outscored them by 15.

Kansas got stronger as the season progressed.  The Jayhawks' final numbers are Final Four worthy.  KU out-shot opponents by an amazing 10.5%.  They posted an R+T of 6.8, and they outscored their opposition by 14.1.

Other teams with total statistics that fall into the range of past Final Four teams, but don't have numbers as good as the examples I just gave are: Boston College, Illinois, LSU, UCLA, and Pittsburgh.  These teams should all advance to the Sweet 16, unless they have to play one another in Round Two (like Illinois and Washington).

First Round Picks (Thursday/Friday)

The Computers will probably not pick these games as well as some of the better handicappers out there.  Nevertheless, here are the first round selections as based on an average of the best 12 ratings:

Washington Region
Connecticut 78  Albany 56
Kentucky 72  UAB 70
Washington 77  Utah State 71
Illinois 60  Air Force 52
Michigan State 64  George Mason 62
North Carolina 79  Murray State 64
Wichita State 70  Seton Hall 67
Tennessee 72  Winthrop 64

Minneapolis Region
Villanova 81  Monmouth 56
Wisconsin 66  Arizona 65 (ratings say WI +.1)
Nevada 74  Montana 67
Boston College 68  Pacific 63
Oklahoma 72  UW-Milwaukee 68
Florida 78  South Alabama 66
Georgetown 67  Northern Iowa 65
Ohio State 77  Davidson 62

Atlanta Region
Duke 87  Southern 61
George Washington 73  UNC-Wilmington 70
Texas A&M 71  Syracuse 70
LSU 74  Iona 67
West Virginia 72  Southern Illinois 69 (I think SIU will upset WVU)
Iowa 70  Northwestern State 63
North Carolina State 69  California 67
Texas 79  Penn 60

Oakland Region
Memphis 80  Oral Roberts 64
Arkansas 70  Bucknell 68
Pittsburgh 73  Kent State 66
Kansas 77  Bradley 67
Indiana 74  San Diego State 72
Gonzaga 78  Xavier 75
Marquette 71  Alabama 67 (I think ‘Bama wins)
UCLA 70  Belmont 54

Round of 32 Picks (Saturday/Sunday)
Based on which team the computers picked in round one

Washington Region
Connecticut  74  Kentucky 62
Illinois 76  Washington 67
North Carolina 77  Michigan State 74
Tennessee 75  Wichita State 73

Minneapolis Region
Villanova 72  Wisconsin 65
Boston College 73  Nevada 66
Florida 75  Oklahoma 67
Ohio State 71  Georgetown 66

Atlanta Region
Duke 77  George Washington 68
LSU 72  Texas A&M 68
Iowa 73  West Virginia 68
Texas 74  North Carolina State 64

Oakland Region
Memphis 78  Arkansas 70
Kansas 67  Pittsburgh 65
Indiana 73  Gonzaga 69
UCLA 71  Marquette 65

My Own Final Four Picks

Washington Region: Connecticut over North Carolina

Minneapolis Region: Florida over Boston College

Atlanta Region: Texas over Duke

Oakland Region: Memphis over UCLA