TODO.md

In bbolker/cricketr: Evaluation of cricket stopping rules

Paper

• write!
• AB: import existing text into BMB's outline

Code

• import scraping code
• implement general cross-validation stuff
  • kappa, binary-bias measures
  • RMSE, bias measures
  • allow different resource table predictors (optfun, raw table ...)
  • RMSE: prediction depends on whether game ended early (i.e. team 2 won); need to rescale resources accordingly
• code for extracting stopping distribution/averaging over it
• more pictures:
  • ggplot_tile capacity
  • combo capacity (do we need a restab.list or a merge-restab functionality?)

Low priority

• consider making a fake database for general distribution

bbolker/cricketr documentation built on May 11, 2019, 9:28 p.m.