Poker day to day TODO

[DolevArtzi]

• [ ] revisit and fix stat mode
• [ ] start thinking about the interactive mode --> will eventually have agents betting
  • [ ] this includes printing the cards as they're opened, not just at the end
• [ ] enumerate possibilities including all opponent hands
  • [ ] Fun challenge: think of an encoding scheme that will allow comparison on hands and cards with < etc.
  • [x] create all possible hands for player and opponent at each stage in the hand
    • I thought this was done but there is a bug, see #14
  • [x] count all possible combinations of hands between player and opponent at each stage in the hand
  • [x] clean and efficient
  • note that creating all possible hands including opponents is probably intractable unless we can come up with something clever
• [ ] probability for each 'node' (won't be a tree yet)
• [x] table.inject
• [x] probability of falling into each score class
  • [x] at the end of the round
  • [x] at each point as the round progresses
  • [x] only considering the best you can guarantee and above
    • [x] done, but not done the intuitive way: considers prob. of improving to that, given you improve.. changed. see https://github.com/DolevArtzi/probably-probability/commit/c63f8b5f4ee6970fa4a1095781110c816e65928c
• [x] refactor stuff out of eval class like comparators, sorters
  • [ ] Comparator not going to do for now, may need to revisit. Sorter isn't the most clean/encapsulated with Eval
  • [x] Sorter
  • [x] Counter - plan: function tree in Combiner class, dont forget to compose with the tuple 'melter' when prod has a length 1 comb. Done
  • then, fix connections to sorting/combiner class, finish probability by score class, while fixing len. 1 tuple bug for generating possible hands Done