POTD #43 Saturday Special High Roller: Wednesday Special Blog

An Online Hand vs WATnlos

When writing this Substack, I’ve been forced to create a taxonomy of the worst plays. What places higher: a play that loses a lot of EV against a perfect strategy? A bad read to exploit? A play that doesn’t lose much EV, but reveals a large hole in your game that is easy to exploit? Let’s think of a hypothetical scenario:

Player A loses 2 BBs per 100 hands heads up to the best HU NLHE bot. 90% of the decisions he makes do not lose EV. He’s randomizing and when there is a mixed strategy he picks the right option 60% of the time. Player B loses 3 BBs per 100 hands; 85% of the decisions he makes do not lose EV, but he randomizes and picks the correct mixed play 70% of the time. You have no other information. Who would you rather play heads up?

Player A does better versus the solver and is better at finding plays that don’t lose EV, but since his frequencies aren’t as sharp as Player B, I think I would rather play Player A because it’s more likely I find holes in their game to exploit. You may feel otherwise, and as always, it’s a matter of degree. What if Player C mixes well, but is a 4 BB/100 loser? Then I would probably switch from A to C. In today’s hand, I folded a pure continue that costs me around a third of a BB in equilibrium. Is it a small one-off error or the type of error that exposes a flaw in my game that can be repeatedly attacked?

HH: GGPoker

Saturday Special High Rollers $10,300
10k Starting Stack (50/100/12) (SB/BB/Ante)

It folds to Christian Rudolph on the button who raises to 200, I make it 900 in the SB with J♦️T♥️, Seth Davies folds in the BB and Christian calls.

Flop (1,984): 7♥️5♥️4♠️ I check, Christian bets 537, I fold