The Pi-Rate Ratings

February 8, 2020

PiRate Ratings College Basketball For February 8, 2020

Filed under: College Basketball — Tags: , , , , , , , , , — piratings @ 6:04 am

Saturday’s Games

 

Home

Visitor

Spread

Abilene Christian

Lamar

6.0

Air Force

San Diego St.

-14.6

Akron

Eastern Michigan

13.4

American

Army

6.1

Appalachian St.

Texas St.

-1.0

Arizona

UCLA

16.3

Arizona St.

USC

3.9

Arkansas-Pine Bluff

Alabama A&M

0.3

Auburn

LSU

4.4

Baylor

Oklahoma St.

13.8

Belmont

Austin Peay

7.8

Bethune-Cookman

North Carolina A&T

2.8

Binghamton

Maine

2.6

Boston University

Holy Cross

17.4

Bowling Green

Toledo

1.7

Brown

Harvard

-4.6

Bryant

Central Connecticut St.

17.3

BYU

San Francisco

11.6

Cal St. Fullerton

UC Davis

1.5

Cal St. Northridge

UC Riverside

1.0

California Baptist

Grand Canyon

7.3

Campbell

Presbyterian

7.2

Charleston

Elon

14.3

Charleston Southern

Hampton

5.0

Charlotte

UTSA

5.4

Chicago St.

UT Rio Grande Valley

-15.6

Citadel

VMI

-0.1

Coastal Carolina

UT Arlington

-2.4

Colgate

Navy

10.4

Colorado

Stanford

6.8

Columbia

Pennsylvania

-4.6

Coppin St.

North Carolina Central

-1.7

Cornell

Princeton

-3.2

Creighton

St. John’s

8.6

Dayton

Saint Louis

15.8

Delaware

James Madison

9.4

Denver

Purdue Fort Wayne

0.4

Drexel

Towson

-0.7

Duquesne

St. Bonaventure

5.4

Eastern Illinois

Morehead St.

7.0

Eastern Washington

Montana St.

6.7

Florida A&M

Howard

11.8

Florida Atlantic

Florida Intl.

1.2

Florida Gulf Coast

North Florida

-6.2

Florida St.

Miami (Fla.)

13.7

Fordham

Richmond

-10.3

George Washington

Rhode Island

-8.9

Georgetown

DePaul

6.0

Georgia

Alabama

-0.7

Grambling

Southern

2.4

Hartford

Vermont

-10.8

High Point

Radford

-10.3

Hofstra

Northeastern

4.2

Houston Baptist

Northwestern St.

-1.8

Idaho

Montana

-5.8

Idaho St.

Sacramento St.

-1.3

Illinois Chicago

Green Bay

1.7

Illinois St.

Indiana St.

-5.3

Incarnate Word

Southeast Louisiana

-0.2

Indiana

Purdue

1.3

Iowa

Nebraska

14.0

Iowa St.

Kansas St.

4.8

IUPUI

Milwaukee

-3.3

Jackson St.

Alcorn St.

1.2

Kennesaw St.

Lipscomb

-10.2

La Salle

Saint Joseph’s

6.2

Lafayette

Lehigh

11.2

Little Rock

Arkansas St.

8.3

Long Island

Merrimack

2.1

Longwood

Winthrop

-10.4

Louisiana

Georgia Southern

-2.1

Louisville

Virginia

11.0

Loyola (MD)

Bucknell

1.1

Marshall

Louisiana Tech

-3.6

McNeese

Central Arkansas

6.7

Memphis

South Florida

10.2

Mercer

East Tennessee St.

-9.3

Michigan

Michigan St.

-3.7

Middle Tennessee

Rice

-0.3

Mississippi St.

Vanderbilt

14.3

Mississippi Valley St.

Alabama St.

-8.4

Missouri

Arkansas

-4.1

Missouri Kansas City

New Mexico St.

-6.3

Morgan St.

MD Eastern Shore

12.3

Nevada

San Jose St.

18.5

New Hampshire

Stony Brook

-5.3

New Mexico

Wyoming

12.4

Nicholls St.

Sam Houston St.

1.9

NJIT

Jacksonville

1.1

Norfolk St.

Delaware St.

16.8

North Alabama

Liberty

-10.4

North Carolina

Duke

-12.7

North Dakota

South Dakota

-0.5

North Dakota St.

Oral Roberts

4.4

Northern Iowa

Drake

11.4

Northern Kentucky

Detroit

12.4

Ohio

Miami (O)

4.6

Oklahoma

West Virginia

-3.9

Old Dominion

UTEP

3.4

Ole Miss

Florida

-3.1

Oregon St.

Oregon

-3.5

Pacific

Pepperdine

4.0

Penn St.

Minnesota

6.2

Pittsburgh

Georgia Tech

2.6

Prairie View A&M

Texas Southern

6.5

Sacred Heart

Robert Morris

4.6

Saint Mary’s

Gonzaga

-4.9

San Diego

Portland

7.0

SIU Edwardsville

Eastern Kentucky

-2.1

South Carolina

Texas A&M

8.5

South Dakota St.

Omaha

8.8

Southeast Missouri

Tennessee Tech

3.7

Southern Illinois

Missouri St.

3.3

Southern Utah

Portland St.

6.1

St. Francis (NY)

Mount St. Mary’s

0.2

Stephen F. Austin

New Orleans

16.3

Syracuse

Wake Forest

7.7

TCU

Kansas

-10.3

Temple

SMU

1.5

Tennessee

Kentucky

-0.7

Tennessee St.

Murray St.

-3.7

Texas

Texas Tech

-1.9

Tulane

East Carolina

6.1

UAB

North Texas

-4.2

UCSB

UC-Irvine

-0.8

UL Monroe

Georgia St.

-7.7

UMass Lowell

UMBC

3.9

UNC Wilmington

William & Mary

-5.4

UNLV

Fresno St.

4.7

USC Upstate

Gardner-Webb

-0.8

UT Martin

Jacksonville St.

-0.8

Utah

California

7.5

Utah St.

Boise St.

7.5

Utah Valley

Seattle

0.1

Villanova

Seton Hall

2.1

Virginia Tech

Boston College

8.6

Wagner

St. Francis (PA)

-5.9

Weber St.

Northern Arizona

-0.1

Western Carolina

Furman

-3.3

Western Kentucky

Southern Miss.

13.4

Western Michigan

Ball St.

-6.0

Wofford

Chattanooga

5.5

Wright St.

Oakland

10.7

Xavier

Providence

4.5

Yale

Dartmouth

15.1

Youngstown St.

Cleveland St.

7.3

 

Saturday’s Best TV Games

 

Time (EST)

Network

Home

Visitor

12:00 PM

ESPN2

TCU

Kansas

12:00 PM

ACCN

Florida St.

Miami (Fla.)

12:00 PM

12:00 PM

ESPN

Fox

Auburn

Michigan

LSU

Michigan St.

1:00 PM

CBS

Tennessee

Kentucky

2:00 PM

CBSSN

Dayton

Saint Louis

2:00 PM

ESPNU

Oklahoma

West Virginia

2:00 PM

ESPN

Indiana

Purdue

2:30 PM

Fox

Villanova

Seton Hall

3:00 PM

NBCSN

Duquesne

Saint Bonaventure

4:00 PM

ESPN

Louisville

Virginia

4:00 PM

BTN

Penn St.

Minnesota

4:00 PM

ESPN2

Texas

Texas Tech

4:00 PM

ESPN3

Northern Iowa

Drake

4:30 PM

ESPN+

Mercer

East Tennessee St.

6:00 PM

ESPN

North Carolina

Duke

6:00 PM

CBSSN

Creighton

St. John’s

6:00 PM

Pac12

Colorado

Stanford

6:00 PM

ESPN+

Belmont

Austin Peay

6:00 PM

ESPN+

Brown

Harvard

7:00 PM

ESPN+

Hartford

Vermont

8:00 PM

FS1

Xavier

Providence

8:30 PM

ESPN+

Tennessee St.

Murray St.

10:00 PM

ESPN

Saint Mary’s

Gonzaga

10:00 PM

ESPN2

Arizona

UCLA

10:00 PM

ESPNU

UCSB

UC-Irvine

 

 

 

 

 

 

4 Comments

  1. Hi There!

    First off – I LOVE your website, thank you so much for what you do.

    I wanted to ask about how the College Basketball spreads are calculated on the Pi-Rate website and how to interpret them for regular season games, I have dug around but cannot located an explanation if there is one.

    Also – thank you for all the hard work during each March Madness – I love your R+T // SOS // TS Margin calculations and how they determine outcomes.

    Thank you!

    Comment by CP — February 8, 2020 @ 7:47 am

    • Thank you for your patronage, CP. While we have published the process for football ratings, we have never really parsed our basketball rating process, mostly because it is a boring reverse-engineered algorithmic calculation.

      Our basketball ratings underwent a large metamorphosis about a dozen years ago, after we became devotees of Dean Oliver and his great book “Basketball on Paper.” He broke the game down into the “Four Factors,” and he showed data based on a large sampling that showed how teams win basketball games. We figured that there had to be a way to predict the outcomes of college basketball games by applying adjustments for strength of schedule and finding the best linear regression line to fit the data.

      There is a small tweaking of the algorithm every year, and we have to wait until about 1,750 games have been played by the 353 Division I teams before the annual constants settle down and remain consistent.

      It is a topic for discussion whether the Four Factors work so well because they best prescribe how a team should try to win, or whether so many teams have become analytics first mathematics geek squads that fit their style to the numbers. Regardless, basketball today is played in a way that makes the outcome predictable with more accuracy than ever before.

      Our ratings may not be usable for betting against the spread, because our ratings are more like the spread itself. We have been told by a popular line originator that we have stumbled upon a similar system to how they set spreads for games.

      Thus, our hopes with our basketball ratings is to have a very low mean square error in our predictions, and we consistently have the top MSE among the ratings at the Prediction Tracker. We won’t claim superiority here, because we cannot put out spreads for every D1 game beginning in November. Our first spreads debut around the end of December or first of January.

      As for our R+T ratings, we will be the first to admit that the formula needs to be overhauled and re-formulated to use rebounding and turnover rate rather than margin. The data is relevant, but the application must be tweaked, and that won’t come this year, because reverse engineering past performances is a very long process. However, there is a definite reason for the “Hustle Stats” being more important in the NCAA Tournament than they are in the regular season, just like Sabermetrics for the Major League Playoffs take on a different view than they do in the regular season, because in the playoffs, the best teams have pitchers that will not surrender walks and give up three-run homers, and the #4 and #5 starters won’t appear except in relief.

      Comment by piratings — February 9, 2020 @ 7:34 am

      • Thank you for the detailed response! Your NCAA Tournament numbers are fantastic and over the past two tournaments, I have tracked the W/L rate of 1H // Game Spread // ML < -200 (using R+T and SOS + the Pi-Rates suggestions as a guide, while omitting those "too close to call games" between R+T and SOS) and have calculated a combined winning percentage of 61.29% across 2018-2019 tournaments.

        In 2019 the edge was immense, probably due to the face that there were a record number of Negative R+T teams that made it to the dance. The largest edge was between the Round of 64 (includes play-ins) // 32 // Sweet 16, a staggering 71.88% win rate against the spread and < -200 ML. Some glaring examples were AZ St. vs St. Johns in the play-ins and of course Baylor v. Syracuse, both of which fit the aforementioned criteria make-up.

        All this to say – I truly appreciate your work and how amazingly close the games can be called and predicted; I am very much looking forward to the dance this year and your predictions. If you Pi-Rates have a tip jar, please consider opening the lid.

        Comment by CP — February 12, 2020 @ 5:18 pm

      • I hope that we can continue to be as accurate. We have been back-testing a rate version of R+T as opposed to a count version, and we will reveal this experiment in March along with the standard R+T. In theory, using a rate version should be more accurate if we can find the right algorithm and constant. Thanks for the idea about the tip jar. Many years ago, we charged a nominal fee for football picks against the spread, which helped provide pizza money for the Captain’s nephews to gather the information, but once they graduated college and started their own professional lives, they stopped helping out.

        Comment by piratings — February 13, 2020 @ 7:24 am


RSS feed for comments on this post.

Sorry, the comment form is closed at this time.

Blog at WordPress.com.

%d bloggers like this: