#31
|
|||
|
|||
Re: Sushi vs Clarkmeister
I've not read the Theory of Poker. I'll try to find a copy [img]/images/graemlins/smile.gif[/img]
|
#32
|
|||
|
|||
Re: Sushi vs Clarkmeister
[ QUOTE ]
You're probably very right here; I certainly have no idea when or why it doesn't work. Do you have any good guidelines for when it might not work? [/ QUOTE ] Shillx's post above is a good start. |
#33
|
|||
|
|||
Re: Sushi vs Clarkmeister
OK here goes... I figured I'd do some good and generalize it while since I bothered.
Its written as an excel document, but rather than just uploading the document I'll hack the code to make it easier to check... and more instructive for those who bother... but didn't bother... whatever. If it passes the scrutiny maybe I'll repost. def. <ul type="square"> [A1] "RA" (weighted range we are ahead of) [B1] "RB" (weighted range we are behind to) [C1] Equity (RA/(RA+RB)) [D1] Pot (potsize, LDO) [E1] %VcWB (percent Villan check when behind, note Vils alternative is to bluff) [F1] %VbWA (percent Villan bets when ahead (*) [G1] %Hftob (percent Hero folds to bet) [H1] FRMLA (as cryptic as I could make it) [XY] Refers what should be enterd in cell X,Y i.e [A1][/list] aaaaand so... type the following; <ul type="square"> [A1] RA [B1] RB [C1] Equity [D1] Pot [E1] %VcWB [F1] %VbWA [H1] %Hftob [.I] EV[/list] now crawl down a row and enter... <ul type="square"> [D2] =(A2/(A2+B2)) [H2] =(C2*(D2*E2+(D2+1-2*(D2+1)*G2)*(1-E2)))-(1-C2)*F2*(1-G2)[/list] Now, you can enter in in fields A2-B2 and D2 the constants for your hand and fill down a few together will the FRMLA (in [H2]). We are now ready to extract knowledge... type in resonable % actions for Villans actions as well as heroes. To show that it gives resonable numbers... lets benchmark against Sushis hand using Wooks hand ranges (RA = 9, RB = 9,5, Pot = 11,5) (1) Assume V always bluffs, always valuebet and H always fold. <ul type="square"> [E2] = 0 [F2] = 1 [G2] = 1 --> EV == -6,08[/list] Well that sucked, but the numbers seems to make sense, we are loosing more than our fare share of the pot because we are also missing out on the bet the bluff adds to the pot (2) V always check when behind, Never value bets and... we never have to bother. <ul type="square"> [E2] = 1 [F2] = 0 [G2] = 0/1 --> EV(0) == 5,59 -->EV(1) == 5,59[/list] Well, if they aren't betting how can we be a calling station, note that this number is the same as our Equity suggests; 11,5*Equity == 5,59 (3) V always bluff, always value bet and we always call <ul type="square"> [E2] = 1 [F2] = 1 [G2] = 0 --> EV(0) == 5,59[/list] Bluff stopping is sweet witch should be and added +EV component So lets hit the Wookie questions... I assume that V will bet 95% of the times he is ahead. (W1) Villain always bluffs and we always call. <ul type="square"> [E2] = 0 [F2] = 0,95 [G2] = 0 --> EV == 5,59[/list] (W2) Villain never bluffs and we always call. <ul type="square"> [E2] = 1 [F2] = 0,95 [G2] = 0 --> EV == 5.11[/list] (W3) Villain never bluffs and we always fold. <ul type="square"> [E2] = 1 [F2] = 0,95 [G2] = 1 --> EV == 5.59[/list] Expert play by Hero, Villain needs bereducation in chat. The identical result compared to (W1) is a fluke consequence of Equity being close to 0.5 Aaannnnnnddddd finally, V is expert and operates on game theory level (W4) V bets at game theory frequency. <ul type="square"> [E2] ==(1-(C5)/(D5)) [F2] = 1 [G2] = [img]/images/graemlins/laugh.gif[/img] --> EV == 5.10[/list]I figure that optimum bluffing strategy would be bluff to value bet <-> 1: Pot and this is how to describe it. It gives good numbers, but I haven't done maths in many years... so any feedback would be welcome. <ul type="square"> %Hftobet FRMLA 0,1 5,10156255 0,2 5,101464096 0,3 5,101365643 0,4 5,10126719 0,5 5,101168736 0,6 5,101070283 0,7 5,10097183 0,8 5,100873376 0,9 5,100774923 1 5,10067647[/list] At least it predicts stuff and hero can relax, and have a beer because it really doesn't matter. heh, -d |
#34
|
|||
|
|||
Re: Sushi vs Clarkmeister
Ideal Clarkmeister
3 limpers, BTN raises, hero calls in the BB with 9 [img]/images/graemlins/heart.gif[/img] 8 [img]/images/graemlins/heart.gif[/img]. Flop: 7 [img]/images/graemlins/club.gif[/img] 6 [img]/images/graemlins/club.gif[/img] 2 [img]/images/graemlins/heart.gif[/img] Check to button who bets, hero calls and someone else calls Turn: 2 [img]/images/graemlins/club.gif[/img] Check to button who bets and only hero calls River: 4 [img]/images/graemlins/club.gif[/img] Hero bets [img]/images/graemlins/smile.gif[/img] |
#35
|
|||
|
|||
Re: Sushi vs Clarkmeister
Thanks Shillx [img]/images/graemlins/smile.gif[/img]
|
#36
|
|||
|
|||
Re: Sushi vs Clarkmeister
X,
I don't believe your math again. For example, if we check, villain bets or bluffs 100% of the time, and we always fold, our EV should be 0, not negative something. We don't lose what's in the pot already. That's already long gone, but we hope to win it back. If we check and fold, we don't win anything, but we don't lose any more either. Sushi got the right answers in his other thread on this hand, so check that out if you still aren't understanding. |
|
|