MCS100/Stats50 Final Project - Extended Abstract
Antonino Abundes & Alec Powell
December 2, 2014
The Effect of Rookie Age and College Experience on Major League Baseball
Career Performance (1985-2010)
Introduction
Does going to college have an impact on a
Stats 50/MCS 100
L. Pekelis
Handout #9
November 13, 2011
Homework Set #3
Due Thursday, Nov 20, 2011.
1. Probability of winning a game of length n
Let
X=
3 , with probability 1/3
0 , with probability 2/3
Y 1.
Let Pn = P r cfw_ n Xi > n Yi be the probabili
STATS 50
Solutions to Homework 2
Problem 1 Duel. Two players start at a distance of 1 unit apart. They walk toward each other, with x
denoting the distance between them. Suppose player I has an accuracy function P1 (x) = 1 x and II has an
accuracy functio
Stats 50/MCS 100
L. Pekelis
Handout #7
October 14, 2011
Homework Set #2
Due Tuesday, Oct 21, 2011.
1. Duel
Two players start at a distance of 1 unit apart. They walk toward each other, with
x denoting the distance between them. Suppose player I has an acc
L. Pekelis
Stats 50/MCS 100
Lecture Supplement #5
October 7, 2014
Lecture # 5 Supplement
Last Time.
Game Theory, pt. 2
Today.
Fundamental Theorem of Game Theory
Duels
Value of Field Position
1
Fundamental Theorem of Game Theory Every 2-person 0-sum ga
L. Pekelis
Stats 50/MCS 100
Lecture Supplement #6
October 9, 2014
Lecture # 6 Supplement
Last Time.
Fundamental Theorem of Game Theory
Duels
Value of eld position
Today.
Go for it! (VoFP)
Point Spreads
1
Go for it!
2
3
References
[1] David Romer. Do
L. Pekelis
Stats 50/MCS 100
Lecture Supplement #7
October 14, 2014
Lecture # 7 Supplement
Last Time.
Go for it! (VoFP)
Point Spreads
Today.
HW 2 posted (due next Tuesday, Oct 21)
Midterm next Thursday, Oct 23
Composition Rules
Midterm
We will be havi
L. Pekelis
Stats 50/MCS 100
Lecture Supplement #1
September 23, 2014
James-Stein Estimator for Batting Averages
1
2
Formulas.
We assumed each players observed BA was Normally distributed around a player specic
true mean, which we called , with a player sp
L. Pekelis
Stats 50/MCS 100
Lecture Supplement #4
October 2, 2014
Lecture # 4 Supplement
Last Time.
Runner - Tackler redux
More Overtime
Game Theory, pt. 1
Today.
HW1 Online
Game Theory, pt. 2
1
Game Theory and Major League Sports.
Authors Kenneth Ko
L. Pekelis
Stats 50/MCS 100
Lecture Supplement #3
September 30, 2014
Lecture # 3 Supplement
Last Time.
Who uses math in sport?
JS & p-Thm applied
Runner - Tackler Problem
Overtime Rules
Today.
Housekeeping: Questionnaire results, Guest speakers, Home
L. Pekelis
Stats 50/MCS 100
Lecture Supplement #11
October 30, 2014
Lecture # 11 Supplement
Last Time.
NERV
All Games Equally Exciting
Today.
Darkest before dawn
Physics of Baseball
1
Darkest before dawn
How low does the probability of winning sink be
L. Pekelis
Stats 50/MCS 100
Lecture Supplement #9
October 21, 2014
Lecture # 9 Supplement
Last Time.
Composition Rules
Combining Expert Opinion
Today.
Recurrence
How to evaluate a baseball player / OERA
Announcements.
Midterm Thurs, Oct 23
1-page Fi
L. Pekelis
Stats 50/MCS 100
Lecture Supplement #15 & 16
November 20, 2014
Lecture # 15 & 16 Supplement
Last Time.
Longer Games = Better Player?
Monte Carlo Simulation
Today.
Permutation Testing
The Importance of Visualization
1
Permutation Testing:
An
L. Pekelis
Stats 50/MCS 100
Lecture Supplement #2
September 25, 2014
Lecture # 2 Supplement
Last Time.
James-Stein Estimator
Bill James p-Thm of Sports
Course Outline
Today.
(optional) Sections info
Who uses math in sports?
JS & p-Thm applied
Runne
L. Pekelis
Stats 50/MCS 100
Lecture Supplement #14
November 11, 2014
Lecture # 14 Supplement
Last Time.
Luck vs Skill
Simpsons Paradox
Today.
Longer Games = Better Player?
Monte Carlo Simulation
1
Longer Games:
Note there is a handout on this topic on
L. Pekelis
Stats 50/MCS 100
Lecture Supplement #10
October 28, 2014
Lecture # 10 Supplement
Last Time.
Recurrence
OERA
Today.
NERV
Meta Theorems of Sports
Final Projects
1
Meta Theorems of Sports
Dene: p(t) = P rcfw_Team A wins given everything we kn
L. Pekelis
Stats 50/MCS 100
Lecture Supplement #13
November 6, 2014
Lecture # 13 Supplement
Last Time.
Record Breaking
Luck vs Skill
Today.
Luck vs Skill
Simpsons Paradox
1
to be completed
2
Stats 50/MCS 100
L. Pekelis
Handout #6
October 2, 2014
Homework Set #1
Due Friday, Oct 10 2014 by midnight.
Please hand in your solutions either in class, or put them in the box marked Stats 50 /
MCS 100 on the homework island across from Leos Oce (Sequoi
STATS 50
Solutions to Homework 1
Problem 1 NFLs new OT rule. In what ways do you think our estimate of W = P rob(receiving team wins)
under the new NFL overtime rule may not reect reality? In particular, how might the kicking team change
their strategy on
The Rise in Ineld Hits
Parker Phillips
Harry Simon
December 10, 2014
Abstract
For the project, we looked at ineld hits in major league baseball. Our rst question was
whether or not ineld hits have been on the rise since 2005. We found through analyzation
Rating Soccer Defenders
Jason van der Merwe
[email protected]
Jack Craddock
[email protected]
Bridge Eimon
[email protected]
MCS100 Stanford University
December 4, 2014
Abstract
The project is motivated by a deciency in the status quo of soccer st
Sean Duggan and Thomas Stephens
Professor Pekelis
MCS 100/Stats 50
December 9th, 2014
Modeling El Clsico:
Real Madrids Passing Trends and Corresponding Insights
As the name suggests, soccer aficionados and casual fans alike hold El Clsico
(translation: Th
Project Title: Overtime Rules in Soccer and their Effect on Winning Percentages
Group Members: Elliot Chanen, Lenny Bronner, Daniel Ramos
Introduction:
We will examine the overtime rules of soccer to evaluate the randomness and fairness of
different rules
Okereke 1
Michelle Okereke
Mathematics of Sports
Leonid Pekelis
11 December 2014
Extended Abstract: First Server Advantage in Tennis
Unlike football, where each team starts one half kicking and the other half receiving,
there is no guarantee in tennis tha
Nicky Sullivan
5-Page Extended Abstract
Who Wins in the MLB Playoffs?
Comparing Pythagorean record to actual record to see which is a better predictor of success in
the playoffs
Every year when the MLB playoffs roll around, articles upon articles are publ
Daniel Allen
12/2/14
MCS 100
Leonid Pekelis
The Single to Right Field: Why Left-handed batters may be undervalued one Mike Trout Unit
Abstract
This paper presents a scrappy but (Id argue) coherent analysis of how a specic play in baseball
is unaccounted f
Pairwise Comparison Models: A Two-Tiered Approach to Predicting Wins and
Losses for NBA Games
Tony Liu
Introduction
The broad aim of this project is to use the Bradley Terry pairwise comparison model as
the basis for finding strong predictive models for N
Walker Szurek
Stats 50: The Mathematics of Sports
Project Abstract
2 December 2014
Assessing NBA General Manager Performance
The purpose of this project is to determine a methodology for assessing general manager
performance in the National Basketball Ass