The Mathematics of Ranking Sports Teams - Insights and Methods

 
The mathematics of
ranking sports teams
Who’s #1?
 
Jonathon Peterson
Purdue University
The Ranking Problem
 
Why is ranking of sports teams important?
College football – BCS
College basketball – NCAA tournament
Win $1 billion!!!
http://www.quickenloansbracket.com/
 
What is so hard about ranking teams?
Strength of schedule matters.
Non
-transitive property
http://www.myteamisbetterthanyourteam.com
Ivy League Football - 2009
 
What is the best team?
 
Is Dartmouth better than Yale?
 
Ranking Methods
 
Statistical Methods
Gather as much data as possible
Cook up a good predicting function
Examples
Jeff Sagarin
RPI
Problems
ad-hoc techniques
Dependent on parameters
 
Ranking Methods
 
Mathematical methods
Ranking based on a mathematical model
Minimize ad-hoc choices
Based on simple principles
Examples
Colley matrix
Massey’s method
Generalized point-difference ranking
Colley Matrix Ranking
http://www.colleyrankings.com
Team i Data:
Schedule Data:
 
Only simple statistics needed (wins, losses, & schedule)
Doesn’t depend on margin of victory
Does include strength of schedule
Colley Matrix Method
Ranking
 
SOS Adjustment
 
 
 
Keep iterating and hope for convergence
 
Iteration – Simple Example
 
Two teams and one game (team 1 wins)
 
Iteration – Simple Example
Colley Matrix - Solution
Two equations:
Solution – Simple Example
Two teams and one game
 (team 1 wins)
 
Matrix Form Solution
 
Ivy League Football - 2009
 
What is the best team?
 
Is Dartmouth better than Yale?
Massey Rating Method
http://www.masseyratings.com
Massey – linear algebra formulation
 
Massey – Least squares
 
 
 
Ivy League Football - 2009
 
What is the best team?
 
Is Dartmouth better than Yale?
 
 
 
Colley – Massey comparison
Another Ranking Method
“A Natural Generalization of the Win-Loss Rating System.”
Charles Redmond, Mercyhurst College
Mathematics Magazine, April 2003.
 
Compare teams through strings of
comparisons
 
Yale vs. Columbia
 
Columbia is 14 better than Brown
Brown is 14 better than Yale
So… Columbia is 28 better than Yale
 
Columbia is 20 worse than Harvard
Harvard is 4 better than Yale
So… Columbia is 16 worse than Yale
 
Average of two comparisons:
Columbia is 6 better than Yale
Average Dominance
Average margin of victory
 
Add self-comparisons
Second Generation Dominance
 
Avg. 2
nd
 Generation Dominance
Connection to Linear Algebra
Adjacency Matrix
Dominance Vector
Limiting Dominance
 
Limiting Dominance
 
Ivy League Football - 2009
 
What is the best team?
 
Is Dartmouth better than Yale?
Conclusion
 
Linear Algebra can be useful!
Matrices can make things easier.
 
Complex Rankings, with simple methods.
 
Methods aren’t perfect.
What ranking is “best”?
Slide Note
Embed
Share

Exploring the importance and challenges of ranking sports teams, this content delves into various ranking methods, including statistical and mathematical approaches like the Colley Matrix, highlighting the complexities and iterative processes involved in determining team rankings.

  • Sports
  • Ranking
  • Mathematics
  • Methods
  • Challenges

Uploaded on Sep 13, 2024 | 1 Views


Download Presentation

Please find below an Image/Link to download the presentation.

The content on the website is provided AS IS for your information and personal use only. It may not be sold, licensed, or shared on other websites without obtaining consent from the author. Download presentation by click this link. If you encounter any issues during the download, it is possible that the publisher has removed the file from their server.

E N D

Presentation Transcript


  1. The mathematics of ranking sports teams Who s #1? Jonathon Peterson Purdue University

  2. The Ranking Problem Why is ranking of sports teams important? College football BCS College basketball NCAA tournament Win $1 billion!!! http://www.quickenloansbracket.com/ What is so hard about ranking teams? Strength of schedule matters. Non-transitive property http://www.myteamisbetterthanyourteam.com

  3. Ivy League Football - 2009 What is the best team? Is Dartmouth better than Yale?

  4. Ranking Methods Statistical Methods Gather as much data as possible Cook up a good predicting function Examples Jeff Sagarin RPI Problems ad-hoc techniques Dependent on parameters

  5. Ranking Methods Mathematical methods Ranking based on a mathematical model Minimize ad-hoc choices Based on simple principles Examples Colley matrix Massey s method Generalized point-difference ranking

  6. Colley Matrix Ranking http://www.colleyrankings.com = # # wins n , w i = losses n Team i Data: , l i = = + # games n n n , , , tot i w i l i games # = G ( ) i j Schedule Data: ij Only simple statistics needed (wins, losses, & schedule) Doesn t depend on margin of victory Does include strength of schedule

  7. Colley Matrix Method Ranking SOS Adjustment = 2 n n + 1 n j , , + w i l i eff w G n r , = w i r , i ij j i + 2 n , tot i + eff w 1 n n n j , = i ' r , , = + w i l i eff w G ' ' n r i + 2 n , i ij j 2 , tot i Keep iterating and hope for convergence

  8. Iteration Simple Example Two teams and one game (team 1 wins) 1 , + i tot n + eff w n 2 n n = i r , , = + w i l i eff w n r i 2 , i j ,

  9. Iteration Simple Example Iteration r1 r2 0 0.500000 0.500000 1 0.666667 0.333333 2 0.611111 0.388889 3 0.629630 0.370370 4 0.623457 0.376543 5 0.625514 0.374486 6 0.624829 0.375171 7 0.625057 0.374943 8 0.624981 0.375019 9 0.625006 0.374994 10 0.624998 0.375002

  10. Colley Matrix - Solution Two equations: 2 + eff w n n 1 n j , , = + w i l i , = eff w i G n r r , i ij j i + 2 n , tot i G n n 1 j ij , , = + + w n i l ) 2 i Combine both : r r i j + + + 2 ( 2 2 n n , , , tot i tot i tot i = + A Matrix Form : r b r

  11. Solution Simple Example + + i tot i tot n n ( 2 2 , , G n n 1 j ij , , = + w i l ) 2 i r r i j + + 2 n , tot i Two teams and one game (team 1 wins) / 1 2 0 / 1 3 / 1 = = A b / 1 6 3 0 Matrix Form Solution / 1 1 / 1 3 2 r / 1 1 = = I A ( ) r b / 1 3 1 6 r 2 / 5 8 625 . r 1 = = / 3 8 375 . r 2

  12. Ivy League Football - 2009 Team Colley Rating Penn .792 Harvard .625 Columbia .583 Princeton .583 Brown .542 Dartmouth .375 Cornell .250 Yale .250 What is the best team? Is Dartmouth better than Yale?

  13. Massey Rating Method http://www.masseyratings.com Ratings should predict score differential ??= rating of the ?-th team If team ? plays team ?, want net point difference to be ?? ?? ?????? ?????= 14 ????????? ??????= 14 ????????? ????????= 10 12 equations with 8 variables - unique solution?

  14. Massey linear algebra formulation # teams = n, # total games = m m x n matrix ? Vector ? = (?1,?2, ,??) Rating vector ? = (?1,?2, ,??) In k-th game team team ? beats team ?. ???= 1, ???= 1, and ???= 0 if ? ?,? ??= margin of victory Massey equation: ? ? = ? No unique solution instead try to minimize ? ? ?

  15. Massey Least squares Want to minimize ? ? ? Try ? = (???) 1?? ? ??? ??? is not invertible Add condition that ? 1 = 0 New least squares problem

  16. Ivy League Football - 2009 Team Massey Rating Penn 25.25 Harvard 10.75 Columbia 0 Princeton -3 Brown -3.75 Yale -7 Cornell -11 Dartmouth -11.25 What is the best team? Is Dartmouth better than Yale?

  17. Colley Massey comparison Team Colley Rating Team Massey Rating Penn .792 Penn 25.25 Harvard .625 Harvard 10.75 Columbia .583 Columbia 0 Princeton .583 Princeton -3 Brown .542 Brown -3.75 Dartmouth .375 Yale -7 Cornell .250 Cornell -11 Yale .250 Dartmouth -11.25

  18. Another Ranking Method A Natural Generalization of the Win-Loss Rating System. Charles Redmond, Mercyhurst College Mathematics Magazine, April 2003. Compare teams through strings of comparisons Yale vs. Columbia Columbia is 14 better than Brown Brown is 14 better than Yale So Columbia is 28 better than Yale Columbia is 20 worse than Harvard Harvard is 4 better than Yale So Columbia is 16 worse than Yale Average of two comparisons: Columbia is 6 better than Yale

  19. Average Dominance Average margin of victory Add self-comparisons Team Average Dominance Team Average Dominance A 3.5 A 2.33 B 4 B 2.67 C -5 C -3.33 D -2.5 D -1.67

  20. Second Generation Dominance 2nd Gen. Dominance Team Dominance A 2.33 3.44 Avg. 2nd Generation Dominance B 2.67 3.22 + + + + 0 5 12 5 2 0 19 12 33 = = . 3 44 C -3.33 -4.11 9 9 D -1.67 -2.56

  21. Connection to Linear Algebra Adjacency Matrix Dominance Vector 1 1 0 1 7 1 1 1 0 8 = = M S 0 1 1 1 10 1 0 1 1 5 1 = 1st dominance avg. gen. S 3 M 1 1 1 = + = + 2nd M dominance avg. gen. 3 ( 9 ) S S S S 3 3 3 k M 1 - n 1 = th n dominance avg. gen. S 3 3 = k 0

  22. Limiting Dominance k M 1 - n 1 = k Does the limit exist? lim S 3 3 n 0 / 1 / 1 2 / 1 2 0 2 / 1 2 0 / 1 2 / 1 2 = = = = , , , v v v v 1 2 1 1 / 1 / 1 / 1 2 2 0 2 / 1 / 1 2 0 2 / 1 2 , 1 = , 3 / 1 = , 3 / 1 = = / 1 3 1 2 3 4 = + + + 0 17 13 3 S v v v v 1 2 3 4 k k k k M 1 1 1 = + + 17 13 3 S v v v 2 3 4 3 3 3 3

  23. Limiting Dominance = 3 3 0 k k k k k M 1 17 1 13 1 3 1 = k = k = k = + + S v v v 2 3 4 3 3 3 3 3 3 0 0 0 17 1 13 1 1 = + + v v v 2 3 4 + 3 1 / 1 3 3 1 / 1 3 1 / 1 3 17 13 3 = + + v v v 2 3 4 2 2 4 31 / 8 . 3 875 29 / 8 . 3 625 = = 37 / 8 . 4 625 23 / 8 . 2 875

  24. Ivy League Football - 2009 Team Dominance Rating Penn 24.34 Harvard 10.06 Columbia -0.09 Brown -2.84 Princeton -2.91 Yale -7.13 Dartmouth -10.56 Cornell -10.88 What is the best team? Is Dartmouth better than Yale?

  25. Conclusion Linear Algebra can be useful! Matrices can make things easier. Complex Rankings, with simple methods. Methods aren t perfect. What ranking is best ?

Related


More Related Content

giItT1WQy@!-/#giItT1WQy@!-/#giItT1WQy@!-/#