864t, button v. big blind: following up on thursday's post

quoted text from the previous post is in grey, new stuff follows:

...

raised pot oop w/o initiative, c/r flop: 

• btn-or top 50% [22+, Ax, Kxs, suited 2-gap+, offsuit 1-gap+] v. 
• bb-cc middle 14.3% [55-99, 50% Axs, broadways JTo-AJs, 98s-J9s]
• bb checks flop
• btn bets all hands
• bb raises all hands - EV = -13.6bb
• btn continuing ranges:
• [>=1p, >=bdfd, >=gs] = 79.6%
• [>=tp, fd, oesd, 1pr+gs, 1pr+bdfd] = 46.4%
• [>=tp, fd, oesd] = 35.7%

• if we define a more complete decision tree as follows:

• btn cbets [1pr+, fd, overs+bdfd, gs+, and var2% of all other hands]...
• bb c/r [tp+, fd, pr+gutter, oesd and var1% of all other hands]...
• btn shoves [tp+, fd, pr+gutter or better], folds otherwise...
• bb calls [tp+, fd, oesd], folds otherwise.

• we we can then vary var2 horizontally (btn cbet air %) and var1 vertically (bb c/r air %) to get an idea of how the ev of bb's c/r (in bold) fares v. different btn cbet tendencies shown in the following table:
  
  
  
  

  	0	25	50	75	100
  0	1.13	1.23	1.32	1.39	1.46
  25	0.09	0.41	0.68	0.92	1.13
  50	-0.95	-0.42	0.04	0.44	0.79
  75	-1.99	-1.25	-0.61	-0.04	0.46
  100	-3.03	-2.08	-1.25	-0.52	0.12

  
  
  
  
  

the not-so-surprising conclusion is that the more air btn cbets the more air bb can get away w/ raising. that said, even if btn is cbetting 100%, bb does best by only raising their legit value/semibluff hands. i think this is an interesting result, as i know i often feel compelled to c/r w/ weakish hands on this flop v. those compulsive blind thieves who cbet like it's their job.

these numbers, and my interpretation, might be off somewhat. i've asked some questions of the software developer on 2p2 that will hopefully clear things up...