Tuesday, September 16, 2008

Picking on Brother Ed, part CVIII

Okay, I'll admit it, I don't really like Ed Miller. The reason is that he's just a bad neighbor. An unfriendly member of the poker blog community.

To me that just makes him a jerk.

He has a policy of simply not recognizing that he's a member of a community. He never links to other blogs and never gives any sign of recognition when other's link to him. That makes him a bad neighbor, makes him a jerk.

The only thing I can figure is that he learned his neighborly skills from his ultimate mentor, Mason Malmuth. He certianly didn't learn it from his wife, who is actually a pretty good internet neighbor.

Ed has partially parted ways with Mason, primarily because Mason couldn't understand why Ed couldn't (Mason doesn't understand that maybe Ed just didn't want to) control his wife. But for some reason known only to Ed it is still important to Ed that he please Mason.

Part of pleasing Mason means to not recognize other internet sites. It's kind of sad.

But none of that is why I'm making this post. I'm making this post to point out a recent example of Ed missing the point.

Ed gets a question from a loyal reader

Recently a reader of my website, John, asked about the $1-$2 blind game at his local cardroom. Here’s what he had to say about his game:

The players there are very loose and pretty terrible, and they call everything pre-flop.

For example, I’ll be under the gun with A-K, raise to $20 preflop, and I’ll get four callers consistently!

That means the pot is already at $100 when the flop comes, and I usually start the hand with only $300.

A-K is a good hand and is meant to flop top pair, best kicker (1 out of 3 times), but as we all know, flopping one pair is only a ‘good’ hand and not a ‘great’ hand. So when I do flop top pair it puts me in a sticky situation. The pot is already too big!

I have gotten to the point where I think I may do one of the following things:

1. Don’t play as deep, so when I do hit my flop I don’t have to worry about reverse implied odds. (Loose play collusion is a killer there with people hitting two pair with junk.)


2. Simply limp in with hands like A-K, K-Q, A-Q, and so forth, in an effort to keep the pot small if I do hit it.

Ed responds
While John has identified an interesting issue, I think he’s overestimating the frequency of bad outcomes and underestimating the frequency of good ones.

I actually agree with that. AK (even if it's not suited) is actually going to hit a good flop about half the time. It will flop top pair about a third of the time, but other flops put it in pretty good shape (JTx as an example).

But reading Ed's next paragraph leaves me unsure that he and I agree on the above as much as it might seem superficially.

He says,
First off, if you raise to $20 with A-K and a $300 stack and four people call (and it isn’t a fluke), you’re in a fantastic situation. It’s such a good situation that you could probably push all-in every time it’s checked to you on the flop and still turn a profit. If you do something smarter than that (and it’s not hard to get smarter than that) then you’ll be solidly in the black.

No, Ed, you aren't in a fantastic situation in early position with AcKh and a flop of Kd 9s8s with a $100 pot and four people behind you. In fact that's a terrible situation.

Sure, you probably have the best hand. But you aren't real likely to make any money with it. You might, however, lose a hefty slice.

Ed seems to think that the frequency various outcomes is what matters. He says,
Before I continue, I want to disagree with something John said: “Loose play collusion is a killer there with people hitting two pair with junk.” That’s an example of overestimating the frequency of bad outcomes. It’s not the mathematical reality. Even four opponents are a significant underdog to flop two pair among them. For instance, an unsuited connector such as 8-7 has a 4.8 percent chance to flop two pair or better. A pocket pair obviously has a better chance to flop a set, but on average I’d guess one of your four opponents will flop two pair or better no more than 25 percent of the time.

When in early position against a large field you're not going to be able to maximize your win in hands you're ahead in, and you're not going to be able to minimize your lose in hands you're behind in.

The solution is to never make the pot big to begin with when you're out of position. If you hit a perfect pot then fine. If you don't then just let it go.

In a really lose game save those preflop raises for late position.

Labels: ,

Thursday, September 11, 2008

Feel v. Math

Feel, or instinct, is just a technique used to estimate the parameters used in math, it's part of what math is.

Pushing Small Edges

One really large difference between nolimit poker and limit poker is the need to push small edges. In limit poker it's usually important to push every edge you can. Not so much in nolimit poker.

Risk control is much, much more important in no-limit than it is in limit.

Tuesday, September 09, 2008

Counting your outs

Before the river (and the showdown) hand evaluation isn't based on the poker hand rankings of one pair, two pair, three of a kind, straight, flush, etc. Hand evaluations and rankings are based on the distribution of possible outcomes.

An out is a card that will improve your hand to a winning hand on the river based on the traditional poker hand rankings. A hand with a lot of outs before the river is often a very good hand even though it hasn't yet achieved anything in terms of the traditional hand rankings.

For example, a no-pair hand on the flop that has a flush draw, a straight draw, and two overcards is a much better hand than top pair on the flop even though one pair beats no pair. That's because that hand with all the draws is the probable winner at a showdown on the river, after two more cards have come. It's a hand with a lot of outs.

The breakeven point between a pair and a draw is 14 outs on the flop. A hand that has 14 outs on the flop is a slight favorite over a hand with one pair and no redraws on the flop.

That would mean hands like:

bottom pair and a flush draw (14 outs)
a flush draw with two overcards (15 outs)
a flush draw with a straight draw (15 outs)
a straight draw and two overcards on a rainbow flop (14 outs)

are pretty good hands to have on the flop.

Note that when I counted a straight draw I specified a rainbow flop. That's because a hand like 8c7c on a flop of 2d 5d 6s doesn't have 8 clean outs to make a straight. Two of the cards that will make it a straight will put three diamonds on the board, possibly killing the straight by giving someone else a flush. You can't really count those 8 outs as an improvement to a winning hand. I count the straight draw on a two suited flop as 6 outs, 7 outs if you have a backdoor flush draw (such as a 6c instead of 6s in the above example).

Counting outs for straight draws can sometimes be problematic. Not only might your straight draw give someone a flush, it might give someone a higher straight. So it's best to be very conservative when counting outs on a straight draw. That's the case in both limit and nolimit forms of hold'em, but it's especially true in nolimit becazuse of the potential cost of making your hand and losing anyway.

Dawn, of I had Outs, recently described a hand she plays which serves as an example of what I'm talking about.
On the button I get 9c8c, I raise to fifteen. Now, I haven’t raised in hours, no one has seen me do anything but listen to my ipod and drink Grey Goose with pineapple. And yet, six people call me.
The flop is a lovely TcJcAs I bet out $40, I get two callers. Qh comes on the turn. I bet $50 and the guy in the one seat raises all-in. I have like $92 left.

Although she's right that a flush draw with a straight draw is a god flop for her, it's not as good a flop as she thinks it is. She effectively has something like 11 outs, not enough to make her a favorite against even bottom pair. A queen will make her a straight, but it's a very dangerous straight. Anybody with a single King will make a bigger straight with that card (which, of course, is the card the guy in seat one has).

She should have at least thought about checking that flop, keeping the pot small so that should a queen fall it probably won't cost her all her chips to draw for that flush. I'm not saying she should have checked, I'm saying should have given it some serious thought.

Labels: , ,

Friday, September 05, 2008

Protecting your bankroll with short buyins

Some writers, Ed Milller in particular, advocates buying in short as a way to protect your bankroll. It's complete nonsense.

It might work to protect your bankroll if you're a losing player.

--- I have an errand to run. I'll have to finish this post later. Go ahead and share your thoughts on the subject while I'm gone.

A commenter took issue with my suggestion that short buyins might help protect your bankroll if you're a losing player. He said, "most deep stacks don't know how to play against an aggressive SS".

Okay, that might be true, but that doesn't mean that you can protect your bankroll by using short buys even if you're a winning player otherwise. It's possible that a loser with deep stacks can convert himself into a winner by playing short stacks if he has opponents who don't know how to play against a short stack.

So even if playing a short stack might improve your lot, that doesn't mean you won't improve even more by just learning to play better and keeping a big stack.

If you're a winning player with a big stack then you'll give up a huge edge by reducing your stack size. Reducing your stack size is reducing the size of the game (for you). In no limit games your win is usually going to be expressable as a percentage of your stack size. If you cut your stack size by 90% your hurting your win in a big way even if cutting your stack will double your percentage win.

A $100 stack with a growth expectation of 20% is more profitable than a $20 stack with a growth expectation of 150%

----- I'm being called for supper, I"ll have to finish this later.

Labels: ,

Attempt to define specialization

I take a look at the discussion forums at 2+2 every once in a while but I'm not a regular reader. For some reason I've never actually looked at their list of forums on PL/NL hold'em.

Here's a list of forums.

High Stakes NL
Medium Stakes NL
Small Stakes NL
Micro Stakes NL
Medium-High Full Ring
Small Full Ring
Micro Full Ring
Heads Up NL

I think that's just amazing. They don't distinctions between pot limit and no limit (which are actually different games) but do distinguish based on blind size. I'm just floored by that.

High Stakes
If you don't at least play 5-10nl or higher then don't give advice in strategy threads. Without a familiarity with the games and players in question your advice is not relevant to the discussion. If you are a lower stakes poster you probably are also not thinking on a high enough level for your advice to add anything but clutter.

Medium Stakes
It is generally assumed that players in the Mid Stakes forum will have a solid understanding of the fundamentals of NL Holdem. Questions related to math (odds), fundamentals, and basic ABC play belong in Small Stakes.

Small Stakes
The SSNL forum is centered around discussion of shorthanded no-limit hold'em cash games from the $0.50/$1 blind level to $1/$2. For full ring games or other blind levels, use the forums listed below.

This is all just nonsense. They are attempting to make fine distinctions between situations (blind sizes) where no actual distinctions exist. But the one place that has real distinctions -- the difference between pot limit and no limit -- they completely ignore.

I wonder if there's anyone at all at 2+2 who has any concept of what poker is actually about?

I sometimes play in a cardroom in Newkirk, OK that spreads no limit games with two different blind structures. A 1/3 blind and a 1/2 blind.

But the one with the smaller blind also has a $5 minimum bet -- it's played as a "red chip game". That makes the 1/2 game much, much bigger than the 1/3 game. It also changes player habits. There's a lot more preflop raising in the 1/3 game than in the 1/2 game.

There's a hell of a lot more to strategic theory than the size of the blinds.

I tend to agree with the commenter that says he thinks 2+2'ers are arrogant, egotistical jackasses.

I think there's an explanation for that which is important to keep in mind whenever you follow any sort of intellectual pursuit.

The 2+2 publishing company touts itself as a provider of "correct information". I'm not sure that phrase actually has any real meaning, but the thought behind it is very comforting to a class of people who want desperately to be able to think of themselves as intellectuals even though they don't actually have the mental capacity to engage in true intellectual pursuits.

Such people tend to be quick to support ideas that simplistic and rigid. Being correct takes on some sort of meaning with bears no resemblance at all to being accurate. Complicated situations make them very uncomfortable. So they are quick to just assume away situational complications.

Labels: ,

Thursday, September 04, 2008

Playing from the small blind

It's often tempting to simply complete the small blind and see the flop with rage. Resist that temptation.

If there's been a couple of limpers you'll be getting good odds to call from the small blind. But the good pot odds are very deceptive. You'll be out of position in every subsequent betting round and your implied odds for most hands on future betting rounds are negative.

Labels: , ,


I haven't looked at the book yet, but based on Hammer's comments on it I think I should read Arnold Snyder's book The Poker Tournament Formula

Maybe somebody will buy me a copy.

My Amazon.com Wish List

Monday, September 01, 2008

Value Bets

Random Shuffle takes Brother Ed to task again. This time for screwing up an important point about value betting.
Today's installment is entitled, "Why do so many of Ed's examples suck so bad?"

The crime scene (emphasis mine):

Someone open raises and a player in the blinds calls. The flop comes J85. The blind checks, the raiser c-bets 3/4 pot, and the blind calls. The turn is a 2. The blind checks, and the raiser checks. The river is a K. The blind checks, and the raiser checks. The blind shows A8, and the raiser wins with QJ.

This is from a post on value betting the river.

Random Shuffle then goes on to point out that in Ed's example the hero missed a more important bet on the turn. Go ahead and read the whole post.

I just wanted to take this opportunity to point out that turn bets are often very important bets.