Problem 3: Attournamento 0) As a preliminary helper result, show by induction that for events E1, E2,..., Em, P(E1 or E2 or .. or EM) < P(Em). (2) m=1 Consider a tournament between N teams, each team playing each of the other teams. 1) How many games will be played in total, in the tournament? We can represent the results of this tournament by a directed graph: node i represents team i, and an edge exists i → j if team i beat team j. 2) Show (by example) there is a tournament that might occur, where every team is beaten by some team. The above problem suggests that it may be impossible to declare an absolute winner, as everyone may be beaten by somebody. We could relax this slightly in the following way: let's call team i a k-winner if there is a group of k-many teams that were each beaten by team i. Other teams may have beaten team i, but there is at least a group of size k that was roundly beaten by i. 3) If the results of each game are decided by fair coin flip, what is the probability that a given team i is a k-winner? 4) Using result 3.0, bound the probability that there exists a k-winner in a tournament of size N? Write it nicely as you can but don't beat yourself up too much with it.

Calculus: Early Transcendentals
8th Edition
ISBN:9781285741550
Author:James Stewart
Publisher:James Stewart
Chapter1: Functions And Models
Section: Chapter Questions
Problem 1RCC: (a) What is a function? What are its domain and range? (b) What is the graph of a function? (c) How...
icon
Related questions
Question

please provide complete handwritten solution for Q6

 
5) For N = 100, what is the smallest k that 3.4 indicates the probability of having k-winners is less than 1? Code
or Mathematica to evaluate your answer in 3.4 is fine.
6) For N = 100 and k as in 3.5, argue that there exist possible tournaments with no k-winners.
Bonus: Using the Chernoff bound to bound the relevant probabilities, show that for a > 1/2 the probability of there
being any aN-winners goes to 0 as N → ∞. Conclude therefore that there exist tournaments without aN-winners,
for all sufficiently large N.
Transcribed Image Text:5) For N = 100, what is the smallest k that 3.4 indicates the probability of having k-winners is less than 1? Code or Mathematica to evaluate your answer in 3.4 is fine. 6) For N = 100 and k as in 3.5, argue that there exist possible tournaments with no k-winners. Bonus: Using the Chernoff bound to bound the relevant probabilities, show that for a > 1/2 the probability of there being any aN-winners goes to 0 as N → ∞. Conclude therefore that there exist tournaments without aN-winners, for all sufficiently large N.
Problem 3: Attournamento
0) As a preliminary helper result, show by induction that for events E1, E2,..., Em,
P(E1 or E2 or .. or EM) < P(Em).
(2)
m=1
Consider a tournament between N teams, each team playing each of the other teams.
1) How many games will be played in total, in the tournament?
We can represent the results of this tournament by a directed graph: node i represents team i, and an edge exists
i → j if team i beat team j.
2) Show (by example) there is a tournament that might occur, where every team is beaten by some team.
The above problem suggests that it may be impossible to declare an absolute winner, as everyone may be beaten by
somebody.
We could relax this slightly in the following way: let's call team i a k-winner if there is a group of k-many teams
that were each beaten by team i. Other teams may have beaten team i, but there is at least a group of size k that
was roundly beaten by i.
3) If the results of each game are decided by fair coin flip, what is the probability that a given team i is a k-winner?
4) Using result 3.0, bound the probability that there exists a k-winner in a tournament of size N? Write it nicely
as you can but don't beat yourself up too much with it.
Transcribed Image Text:Problem 3: Attournamento 0) As a preliminary helper result, show by induction that for events E1, E2,..., Em, P(E1 or E2 or .. or EM) < P(Em). (2) m=1 Consider a tournament between N teams, each team playing each of the other teams. 1) How many games will be played in total, in the tournament? We can represent the results of this tournament by a directed graph: node i represents team i, and an edge exists i → j if team i beat team j. 2) Show (by example) there is a tournament that might occur, where every team is beaten by some team. The above problem suggests that it may be impossible to declare an absolute winner, as everyone may be beaten by somebody. We could relax this slightly in the following way: let's call team i a k-winner if there is a group of k-many teams that were each beaten by team i. Other teams may have beaten team i, but there is at least a group of size k that was roundly beaten by i. 3) If the results of each game are decided by fair coin flip, what is the probability that a given team i is a k-winner? 4) Using result 3.0, bound the probability that there exists a k-winner in a tournament of size N? Write it nicely as you can but don't beat yourself up too much with it.
Expert Solution
trending now

Trending now

This is a popular solution!

steps

Step by step

Solved in 2 steps with 2 images

Blurred answer
Recommended textbooks for you
Calculus: Early Transcendentals
Calculus: Early Transcendentals
Calculus
ISBN:
9781285741550
Author:
James Stewart
Publisher:
Cengage Learning
Thomas' Calculus (14th Edition)
Thomas' Calculus (14th Edition)
Calculus
ISBN:
9780134438986
Author:
Joel R. Hass, Christopher E. Heil, Maurice D. Weir
Publisher:
PEARSON
Calculus: Early Transcendentals (3rd Edition)
Calculus: Early Transcendentals (3rd Edition)
Calculus
ISBN:
9780134763644
Author:
William L. Briggs, Lyle Cochran, Bernard Gillett, Eric Schulz
Publisher:
PEARSON
Calculus: Early Transcendentals
Calculus: Early Transcendentals
Calculus
ISBN:
9781319050740
Author:
Jon Rogawski, Colin Adams, Robert Franzosa
Publisher:
W. H. Freeman
Precalculus
Precalculus
Calculus
ISBN:
9780135189405
Author:
Michael Sullivan
Publisher:
PEARSON
Calculus: Early Transcendental Functions
Calculus: Early Transcendental Functions
Calculus
ISBN:
9781337552516
Author:
Ron Larson, Bruce H. Edwards
Publisher:
Cengage Learning