find.bracket: Fill out a bracket based on some criteria

Description Usage Arguments Value Author(s) Examples

View source: R/find.bracket.R

Description

Fill out a bracket based on some criteria

Usage

1
2
3
4
5
6
find.bracket(bracket.empty, prob.matrix = NULL, prob.source = c("pop",
  "Pom", "538"), pool.source = c("pop", "Pom", "538"), league = c("men",
  "women"), year = 2017, num.candidates = 100, num.sims = 1000,
  criterion = c("percentile", "score", "win"), pool.size = 30,
  bonus.round = c(1, 2, 4, 8, 16, 32), bonus.seed = rep(0, 16),
  bonus.combine = c("add", "multiply"))

Arguments

bracket.empty

a length-64 character vector giving the field of 64 teams in the tournament, in order of initial overall seeding

prob.matrix

a matrix of probabilities, with rows and columns corresponding to teams, matching the output of bradley.terry(). This probabilities are used to simulate candidate brackets and outcomes on which to evaluate the candidates. If NULL, prob.source is used.

prob.source

source from which to use round probabilities to simulate candidate brackets and outcomes — "pop": ESPN's population of picks (default), "Pom": Ken Pomeroy's predictions (kenpom.com), or "538": predictions form fivethirtyeight.com. Ignored if prob.matrix is specified.

pool.source

source from which to use round probabilities to simulate entries of opponents in pool. Same options as prob.source.

league

which league: "men" (default) or "women", for pool.source.

year

year of tournament, used for prob.source and pool.source

num.candidates

number of random brackets to try, taking the best one (default is 100)

num.sims

number of simulations over which to evaluate the candidate brackets (default is 1000)

criterion

how to choose among candidate brackets: "percentile" (default, maximize expected percentile within pool), "score" (maximize expected number of points) or "win" (maximize probabilty of winning pool).

pool.size

number of brackets in your pool (excluding yours), matters only if criterion == "win" (default is 30)

bonus.round

a length-6 vector giving the number of points awarded in your pool's scoring rules for correct picks in each round (default is 2^round)

bonus.seed

a length-16 vector giving the bonus awarded for correctly picking winner based on winner's seed (default is zero)

bonus.combine

how to combine the round bonus with the seed bonus to get the number of points awarded for each correct pick: "add" (default) or multiply

Value

the length-63 character vector describing the filled bracket which performs best according to criterion among all num.candidates brackets tried, across num.sims simulations of a pool of pool.size with scoring rules specified by bonus.round, bonus.seed and bonus.combine

Author(s)

sspowers

Examples

1
2
find.bracket(bracket.empty = bracket.men.2017, prob.source = "538",
  pool.source = "pop", league = "men", year = 2017)

mRchmadness documentation built on May 30, 2017, 6:49 a.m.