SB D. Sklansky: 15K
BB: 5K
Blinds 100/200 ante 50 with 8 players
Pot is 700. Action folds to DS in SB who opens to 1200. That's 100 more than a pot size raise. BB moves allin. I would have moved allin as the SB instead of betting 1200, but what do I know.
Pot is 6200. 3800 to call to DS. DS estimates opponent range of 55+, AK, AQ. That's 55 + 16 + 16 hands respectively for a total of 87 hands. I think most would have guessed a wider range. But probably DS is trying to keep things simple for pedagogical reasons.
- Against AK and AQ, an estimate of winning 17 hands is reasonable.
- Against 88+ = 42 hands, you win about 20% or 8 of those (not 7 as given by DS)
- Against 55 to 66 = 12 hands, you win about 80% or 10 of these.
- Against 77 = 1 hand, you "win" 0.5 of those.
Calculation is 17 + 8 + 10 + 0.5 = 35.5. DS gets get 34.5, because he made an estimate 20% of 42 hands as 7, when clearly the estimate should be 8.
Now calculating your chances of winning is 35.5/87 = 40.8%. Pokerstove gives 40.7%. The 40.8% is a very lucky result. DS calculated 34.5/87 = 39.7%.
DS called the allin, since he needs 3800/(3800 + 6200) = 38%.
Opponent called with KQ. DS remarks that the calc would now be 37/93 and says it would now be over 40%, but this calc is actually 39.8%. An amended calc gives 44/103 = 42.7%. PokerStove gives 41.2%
The 103 comes from adding 16 to 87. DS appears to have added 6 as if KQ were a pair. Off by 10.
The 44 comes from .5*(16 + 16 + 16) + 1.5 = 25.5 instead of 0.5*(16 + 16) + 1 = 17. We are continuing to follow DS's apparent example of adding about 0.5 for each set of 16 hands. I don't do it this way, but this follows the apparent pattern. Pairs are assumed to be small favorites against over cards, however 22 vs JTo is about 48.4% for instance.
So the calc would look like: 25.5 + 8 + 10 + 0.5 = 44 for the numerator. This gets us to 44/103 as desired.