Beating the Party 10+1, Part 2

[Iceyburnz]

I will definately try this out in my next 10+1 as my SNG results lately havent been to my liking.

[dana33]

[ QUOTE ]

The question I wanted to ask was: what happens if you choose a top % of the list, and feed it back into itself? what happens if you keep repeating that process? Does it settle into a stable list, or does it cycle around as hands which do well against the current top move up, and then hands which do well against those move up, etc.

One way to try this is to pick the top% that you want to feed back into the calculation. So let's say you start with the "vs random hand" rankings, then you pick the top half of those, and re-rank according to how well each hand matches up to that select list. Pick the top half again and repeat.


[/ QUOTE ]

Dammit, eastbay, you keep beating me to the punch! [img]/images/graemlins/smile.gif[/img] (But I guess I can't complain, since it was your post which got me thinking about this stuff in the first place.)

I was wondering about the exact same question (but I have been too busy lately to work on my poker code and answer it for myself). I.e., do things converge if you keep re-ranking the hands against the top 50% in the prior iteration? And if they do, what does the final ranking even mean? Is this question just of interest to math geeks, or are there any practical implications?

In a heads-up all-in-or-fold match, I think the right way to think about hand rankings is this:

From the SB, since you are first to act and have no knowledge of what the BB has, you should choose to push with the top X% of hands as ranked against a random hand. From the BB, if your opponent has pushed, you should choose to call with the top Y% of hands as ranked against the hands he would push with. (X and Y will vary depending on the size of the blinds relative to the stacks and your opponent's specific strategy.)

For example, if you know that your opponent pushes with the top 50% of hands (vs a random hand), then you should choose your calling hands according to how they rank against these. If you know he's too tight and only pushes with pairs, then you should rank your calling hands against pairs. And so on.

Does this make sense?

[eastbay]

[ QUOTE ]

In a heads-up all-in-or-fold match, I think the right way to think about hand rankings is this:

From the SB, since you are first to act and have no knowledge of what the BB has, you should choose to push with the top X% of hands as ranked against a random hand. From the BB, if your opponent has pushed, you should choose to call with the top Y% of hands as ranked against the hands he would push with. (X and Y will vary depending on the size of the blinds relative to the stacks and your opponent's specific strategy.)

For example, if you know that your opponent pushes with the top 50% of hands (vs a random hand), then you should choose your calling hands according to how they rank against these. If you know he's too tight and only pushes with pairs, then you should rank your calling hands against pairs. And so on.

Does this make sense?

[/ QUOTE ]

I've thought along the same lines, but I'm not convinced it makes sense.

If your opponent folds, he folds, and it doesn't matter what hand you held.

The only times it matters what cards you selected was when he calls. And when he does call, you want to be holding something which does well against his calling hands. His calling hands aren't random. Therefore, pushing against random hands doesn't make sense, I don't think.

eastbay

[dana33]

[ QUOTE ]
If your opponent folds, he folds, and it doesn't matter what hand you held.

The only times it matters what cards you selected was when he calls. And when he does call, you want to be holding something which does well against his calling hands. His calling hands aren't random. Therefore, pushing against random hands doesn't make sense, I don't think.


[/ QUOTE ]

Hmmm... I guess I was confining myself to "level 1" thinking. It seems you could get into an endless cycle of measures and counter-measures here. That is, you will rank your push hands against the hands he will call with. But he knows that, so he will rank his calling hands against the hands you will push with. But you know that he knows that, so you will re-rank your push hands against the hands he will call with based on his previous assessment of the hands you will push with. And so on.

Based on this, I'm all the more interested to see if there is any convergence of this process to a stable or semi-stable set of hand rankings.

[Canine K9]

Wow, this is a very good thread, thank you.

[Pitcher]

Hi Ben Franklin and AleoMagus,

Great summary. This will work at 10+1 and higher levels(certainly 30+3, that play almost exactly like 10+1). It is a good starting point for any level, from what I can see.

The reason for my post is to second AleoMagus' thoughts on Axs. This is a great hand in low limit ring games because you frequently get 2+ callers. In NL, this hand is only playable in two circumstances.
The first circumstance is very early on a very loose passive table. There are not too many of those, but if you find one, this hand has some value.
The second is late where the real value of the hand is the A. Being suited is just a little icing on the cake.

Pitcher

[Pitcher]

Hi Phil Van Sexton,

I like your post on bubble play
The point about stealing the blinds cheaply should be put in highlights. Many players defending their blinds do not differentiate between raise sizes, other than all in. If they play back at you, it is frequently alot easier to back off of a 2x raise than a 3x raise. (of course, that depends on stack size....)
I want to add a bit to this concerning defending the small blind. If you are going to defend your SB, re-raise to at least 3x or at least double the raiser, whichever is larger. The reasoning behind this is to completely discourage the BB from coming into the hand. If this bet takes it down you get their raise and the BB (a substantial pot this late in the tourney!). The value in this case is in this specific hand....although it may help you steal the BB's on future hands. At this point in the tourney, you are mainly playing for this hand because there might not be a future hand to worry about. I would not do this with less than KQ, or A-10 or better (and 88-AA)

Pitcher

Pitcher

[blackaces13]

[ QUOTE ]
Most people would always bet the pot, but that's probably a lot of $. Just like pre-flop, you want to bet the minimum that will get them to fold. Most players will call any bet if they hit the flop, and fold to any if they didnt.



[/ QUOTE ]

Man I think this is an EXCELLENT point that you never hear anyone mention. Why bet the size of the pot when you're pretty much trying to steal? A typical player during bubble plaly won't call a 1/2 pot sized bet but fold to a pot sized bet.

There are a lot of times where I'll have like AT or something on the button with 4 players left and the BB that called my PF raise will check to me when the board is 3 ragged undercards. Here I'll be tempted to bet the pot, but I usually stop myself and remind myself that the reason why I'm betting is to get my opponent to fold so I only need to bet the MINIMUM amount that will get this guy to lay it down, anything more and I have made a mistake.

I think pot sized bets are good when you have a strong hand and want to protect or get a loose call. But when you're trying to pick up a heads up raised pot on the flop with 4 or 5 people left and big blinds its usually not necessary, or wise.

What do others think about this? I think the whole "bet the pot" idea is pretty pervasive and I'd imagine a lot of people disagree with this approach.

[willie838]

tried this out yesterday, 3 tourneys, 6th, 4th, and second were my results. don't really blame the guide for them though.

the 6th place was bad luck, i was gettin lowstacked from sittin tight until the end, hole cards are 99 so i go all in from early, BB calls with AA.

4th place was an idiotic play by me which i deserved to lose on. free play from BB, flop top pair 8 with jack kicker. lose to a set.

2nd place, eh i shoulda took down first but i took a tough one headsup. hole cards AK. guy raised into me, i go all in. he calls Q7. i'm ahead on the flop kings vs queens, but he hits runner 7s to bust me with a boat.



i'm gonna keep tryin to use this because it did seem pretty effective, but i'll mix in some of my own play (hopefully smarter than my 4th place finish f up). seems pretty good so far.

[jakoye]

I'm trying to understand what exactly your betting pattern is. Are you saying that you make a pot-sized bet when you have the goods and a 1/2 pot-sized bet or 3xBB bet when you don't? Wouldn't that make your hand fairly readable to your opponents?

The advantage to always making a pot-sized bet is that it will be large enough that if someone raises you or even if they just call, you are likely beaten. Also, raising the same amount each time disguises your hand: your opponents will never know when you're bluffing and when your really have it.

But I am intrigued by your point about how you should only bet the "minimum" amount needed to get your opponents to fold. I'm just not sure that's wise because of the danger of giving away too much information about your hand. Of course, you could randomize your bets between two levels, I suppose, but that seems like a bit of work. Poker's already tough enough! [img]/images/graemlins/smile.gif[/img]