open % from CO and button

[efficacy]

How do you get these numbers? If I go to the position stats tab in PT and look at raise first in %, those numbers are way lower than they should be for Button and CO.

[img]/images/graemlins/confused.gif[/img]

[Oink]

[ QUOTE ]
How do you get these numbers? If I go to the position stats tab in PT and look at raise first in %, those numbers are way lower than they should be for Button and CO.

[img]/images/graemlins/confused.gif[/img]

[/ QUOTE ]

Thats because the "raise first in" stat measures how often it was folded to you and you raised.

Its not a measure for how often you raise if its folded to you

I found the percentages using pokerstove choosing my range to see how many of the hands it is.

[yourface]

how to check using PT

under the general tab,
filter for only hands from the CO or only hands from the button
filter for 'chance to steal and raised'. record this number 'a'
filter for 'chance to steal and folded'. record this number 'b'

open raising % = a/(a+b) *100%


but yah using stove is easier so long as you can be honest with yourself about your hand range

[sweetjazz]

My numbers are 25.7% in the CO and 48.4% on the button. I'm not sure my numbers are right, but I suspect if your open on the button % minus your open on the CO% is < 15%, then you probably understand fully the importance of position.

Also, it's really important to understand Heisenberg's point. If you add marginal hands, your opponent often in theory can't beat your move. If he decides not to pay off your big hands any more than in the past, he'll often have to fold to your weaker hands or find himself playing passively and giving you free cards and allowing you to make thin value bets. If he tries to adjust by punishing your increased looseness, he ends up putting in more bets when you have a strong hand. It's tough to figure exactly what hand range has this positive theoretical effect, and even tougher to play a good enough game to consistently take advantage of it. At the same time, how your opponent adjusts will dictate what you do. If he starts spewing more (particularly early in hands), you tighten up and punish his overaggressive play; if he goes back to playing more solidly and less unpredictably, you attack him (particularly when he's out of position) with a wider range.

[rzk]

[ QUOTE ]
Adding very barely -EV hands to your distribution CAN increase your winrate if your opponents adjust perfectly.

adding super super marginal hands forces your oppponets to call down a lot more and re-raise more for value.

After this strategic adjustment these hands become borderline, but the other strong hands in your distributions increase their winrate.

Therefore the total winrate is higher.

[/ QUOTE ]

that's not true. if your opponent always adjusts perfectly to your strategy you cannot increase your total winrate by adding a hand with EV<=0. oink and i had a long theoretical discusion about this in a recent thread. he agreed with me that if your opponent begins to play more optimally against your strategy this results in a decrease of your winrate. if you still disagree i can elaborate more on this.

EDIT: this btw doesn't disprove the shania approach. but my claim is that for it to work against a particular opponent it's necessary that he overadjust.

[Heisenb3rg]

there is a chance im wrong, I havent looked at it mathematically... just makes sense to me intuitivly.

please give me mathematical proof.

[sethypooh21]

I think we're over complicating this slightly - they point of adding these hands is to make your range +EV by making your hand harder to read, which will induce your opponent to make more FTOP mistakes. This is only really useful against opponents who will become more passive in the face of this uncertainty. Tannenbaum's new book describes this as FUD ("fear, uncertainty, doubt")

Against villains who are going to go all spazzy when uncertain, we only need to showdown one "loose" raise, and then play aggressive, but value-based poker, as their spazziness is going to force us to release too many hands or enter into a ton of high variance spots with little or no EV if we continue to raise widely.

[RudeboyOi]

am i the only one...

whos CO range is wider than their button range?

something like 42 from the CO 38 from the button

typically i respect button raises less than CO raises
im under the assumption my opponents think this same way

therefore my button range is slightly tighter than my CO

if the button is prone to coldcalling
i tighten up my CO raising range

UTG im fairly tight and straightfoward

depending on the texture of the game
my HJ range is either more like my UTG range
or slightly tighter than my button range

[rzk]

[ QUOTE ]
there is a chance im wrong, I havent looked at it mathematically... just makes sense to me intuitivly.

please give me mathematical proof.

[/ QUOTE ]

ok, the formal proof is rather long. but it's based on a simple intuitive observation that you cannot get a better EV playing against a perfectly adjusting opponent than playing against an unobservant opponent.

ok then, suppose your current range is, for example, range1={AJ+, 88+}. let's call your postflop strategy with this range X1. your perfectly adjusting opponent plays the optimal counterstrategy, which we'll call A1. we will assume that you cannot come up with a better strategy than X1 against A1 (otherwise you would certainly adjust). let's call your EV with this range and this strategy EVbest(range1 vs A1). suppose AT is a -EV hand against A1, which means even when you play it the best possible way against A1 your EVbest(AT vs A1) is less than 0.

i will now assume that there exists a strategy X2 for range2={AJ+,88+,AT} such that when your opponent plays optimal counterstrategy A2 your EVbest(range2 vs A2) is greater than EVbest(range1 vs A1)+EVbest(AT vs A1). i will now show that this leads to a contradiction.

since A2 is an optimal counterstrategy against X2, X2 vs A2 has a lower EV for us than X2 vs A1. then we have:

EV(X2 vs A1)>EV(X2 vs A2)=EVbest(range2 vs A2)>EVbest(range1 vs A1)+EVbest(AT vs A1)

now, strategy X2 consists of two substrategies:

substrategy X2' : how to play range1={AJ+, 88+} postflop,
substrategy X2" : how to play AT postflop.

then we have:

EV(X2' vs A1)+EV(X2" vs A1) > EVbest(range1 vs A1)+ EVbest(AT vs A1)

but clearly,

EV(X2" vs A1)<EVbest(AT vs A1) by definition of EVbest.

therefore,

EV(X2' vs A1) > EVbest(range1 vs A1)=EV(X1 vs A1).

but this contradicts the statement that X1 was the best strategy you could come up with against A1, so there we go.

cliffnotes: if it were possible for you to increase your winrate by adding a -EV hand to your range, you would do even better if you just change the way you play your original range.

[sweetjazz]

[ QUOTE ]
am i the only one...

whos CO range is wider than their button range?

something like 42 from the CO 38 from the button

[/ QUOTE ]

I'm too lazy to elaborate, but that's really bad. It doesn't matter that people know that you should be stealing more OTB than in the CO; half decent players are going to exploit you (even if they don't realize your odd preflop distribution) because you're more likely to run into a big hand in the CO and you're also more likely to end up OOP.