ESSAY: Malmuth Bluffing Errors
QuadNines
frequently can take advantage of opponents who bluff too little or too
much. When you exploit, though, you usually can be exploited.
In his essay "A Note on Bluffing" (POKER ESSAYS: VOLUME III, pp.
88-90), Malmuth attempted to analyze the question of "how often should
you bluff?" He got it wrong.
Malmuth asked:
Suppose there are two players, lets call
them A and B. They play exactly the same,
except that on the end Player A will bluff
50 percent of the time and Player B will
never bluff. Who do you think plays worse?
If you've read Malmuth's essay, ignore his answer and think through
the problem for yourself. Malmuth got it wrong.
Also assume you are heads-up on the end and both A and B play well,
except for their bluffing frequencies.
---------------------
Consider Player B, who never bluffs. According to Malmuth:
Since he never bluffs, you should never
call. Your expectation on the end is also
zero because you never give up a bet that
you shouldn't.
Malmuth got it wrong. He seriously misunderstands expectation, since
any reasonable river strategy here has a solidly positive expectation.
Malmuth also seriously misunderstands strategy, for you sometimes
should call players who never bluff.
Can you develop a strategy that has zero expectation on the end?
Here's one. If you act first, fold. If you act last, fold.
That's an absurd strategy, of course, since (among many other flaws)
it has you folding the nuts.[1] The problem is, any zero-expectation
river strategy here is going to be ridiculous.
Why did Malmuth settle for zero expectation when any decent river
strategy here has positive expectation? You'll have to ask him.
What is a better strategy when, say, someone bets into you on the end?
Generally, you want to value raise with your strongest hands, call
with your second-tier hands, bluff raise with some of your third-tier
hands, and fold your remaining hands.
With many of your calls, you'd prefer that your opponent had not bet.
In cost/benefit terms, it usually costs you some fraction of a big bet
for the privilege of seeing a showdown. But it's worth the price.
Your likelihood of winning is high enough that your fair share of the
pot (i.e., your "pot equity") exceeds the cost of seeing the showdown.
How does Player B hurt himself by never bluffing? There are at least
two causes.
First, he allows you to win showdowns with some hands you normally
would fold to a bet/raise on the end. This costs him more than he'd
lose by occasionally bluffing and sometimes losing an extra big bet.
Second, you can exploit his mistakes. By never bluffing, Player B is
more readable, because all his bets/raises are for value. You'll call
fewer bets/raises than usual, so his value bets/raises are less
profitable, and his expectation is less.
---------------------
Now consider Player A, who bluffs half the time. According to
Malmuth:
Since he bluffs half the time you should
call him every time. Thus, half the time
you will lose an additional bet, but this
will be canceled out by the other half of
the time when you will win an additional
bet. Your expectation on the end will be
zero.
Malmuth got it wrong. Again, he showed a poor understanding of
expectation and poker strategy. Any decent river strategy here will
have positive expectation and will include all betting actions--not
just calls.
What is a reasonable strategy when, say, Player A bets into you on the
end? While Malmuth will call every time, I generally will value raise
with my strongest hands, call with my second-tier hands, bluff raise
with some of my third-tier hands, and fold my remaining hands.[2]
I'll call with more hands than I would if my opponent never bluffed,
but I won't call every time. As an extreme example, I'll rarely call
when my hand has the worst possible fit with the final board. I'm so
unlikely to win anything that, unless the pot is humongous, it isn't
profitable to toss in another big bet to see the showdown.
How does Player A hurt himself by bluffing half the time? There are
at least two causes.
First, Player A bluffs so often that many of his bluffing hands have
higher expectations if he checks them and takes his chances at
showdown. Risking an extra big bet in hopes of winning an uncontested
pot just isn't worth it with these hands; it diminishes his
expectation.
Second, you can exploit his mistakes. As I noted above, you should
call more often. You'll lose a little extra when Player A is value
betting/raising, but you'll win back all that and more by picking off
some of his excessive bluffs.
---------------------
Do you agree with Malmuth's interesting conclusion that "both A and B
play equally bad?" Do you agree with Malmuth that both players have
zero expectation on the end? If you answered "yes" to either
question, you might want to review some basic poker concepts.
---------------------
[1] Malmuth noted:
I'm aware that you may occasionally hold a
strong hand that you will play--perhaps
with a raise--no matter who bets. But for
practical purposes, these hands have no
impact on the result of this question.
Malmuth got it wrong. These hands have a huge impact on the results
of this question . . . and on your bankroll. When you fold the nuts
on the end, for example, it's both an expected value (EV) disaster and
a results-oriented disaster (i.e., a "Malmuthian disaster" or
"Sklanskian catastrophe").
[2] I assume Player A bluffs on the end at the 50 percent rate only
once per hand. If he still bluffs half the time when his bluff is
raised, can you concoct a crushing counterplan?
William Coleman
including Mason, think he is. In particular, Mason isn't even in the same
league as his coauthor David Sklansky, and many posters on RGP, including Russ,
Steve, Lee, Doyle, and Abdul Camel Jockey.
In terms of what I have learned from books, I have learned most of what I know
about holdem from Doyle Brunson's "Super System" and Sklansky and Malmuth's
"Holdem Poker for Advanced Players". For this reason, I put Mason on a
pedestal, along with David and Doyle. Over the years, I have become aware of
more and more errors and misconceptions in the joint writings of Sklansky and
Malmuth, and in the books Mason has written on his own. I have found no such
errors in books written by David alone, nor in the writings of Doyle Brunson.
There are many other "world class" players such as Mike Caro, T. J. Cloutier,
Phil Hellmuth, and Annie Duke who have made the mistake of publishing their
ideas about poker theory. They may have the poker playing public convinced that
they are experts, but the real experts, who hang out on RGP, know that these
pretenders are basically clueless when it comes to the specifics of correct
poker theory.
I finally lost all respect for Mason when he caved into pressure from a few
censorious complainers, and barred me from 2+2. While Matty Sklansky was
moderator of 2+2 forums, Matty let me post whatever I wanted to, and ignored all
complaints. As soon as Matty quit, Mason barred me. At this point I consider
Mason to be a joke, as do many knowledgeable members of RGP. Why do you think
Mason, and other "experts" like Mike Caro, hardly ever post here? Because they
are laughed off the newsgroup every time they show their face. You are really
not going to shock or surprise anybody on this newsgroup by pointing out major
errors by Mason Malmuth. I have just glanced at your post, so I really cannot
comment on how valid your particular criticisms are.
--
William Coleman (ramashiva)
homepage: www.home.earthlink.net/~ramashiva
windows messenger: ramas...@hotmail.com
IESOUS CHRISTOS THEOU YIOS SOTER (corrupted version)
IESOUS CHRISTOS THEOS YIOS SOTERES (true version)
Sell all your possessions; give the money to the poor; and come, follow me.
-- Jesus Christ
"QuadNines" <Quad...@hotmail.com> wrote in message
news:99fd5b1.02111...@posting.google.com...
Douglas Knoll
> In his essay "A Note on Bluffing" (POKER ESSAYS: VOLUME III, pp.
> 88-90), Malmuth attempted to analyze the question of "how often should
> you bluff?" He got it wrong.
>
> Malmuth asked:
>
> Suppose there are two players, lets call
> them A and B. They play exactly the same,
> except that on the end Player A will bluff
> 50 percent of the time and Player B will
> never bluff. Who do you think plays worse?
>
> If you've read Malmuth's essay, ignore his answer and think through
> the problem for yourself. Malmuth got it wrong.
I don't have the essay in front of me, but I remember thinking that he
forgot to add the condition that your holding can only beat a bluff. It
seemed to me that this is what he was talking about, though.
Yes, I realize that the following paragraph seems to contradict my
interpretation:
I'm aware that you may occasionally hold a
strong hand that you will play--perhaps
with a raise--no matter who bets. But for
practical purposes, these hands have no
impact on the result of this question.
But it still seems the most reasonable, given this:
Since he bluffs half the time you should
call him every time. Thus, half the time
you will lose an additional bet, but this
will be canceled out by the other half of
the time when you will win an additional
bet. Your expectation on the end will be
zero.
> Malmuth got it wrong. He seriously misunderstands expectation, since
> any reasonable river strategy here has a solidly positive expectation.
> Malmuth also seriously misunderstands strategy, for you sometimes
> should call players who never bluff.
I got the idea that the main thrust of the essay was to explain why
bluffing was essential to limit holdem strategy. I think most players
would realize that the analysis dealt with a simplified and contrived
poker-like situation, and was not intended to directly yield a usable
river strategy for real games. It seems a bit strong to state that
Malmouth "seriously misunderstands" expectation and strategy.
By the way, thanks for writing that essay. You bring up a lot of important
things to consider when acting on the river in limit holdem. I just wanted
to suggest that the principle of charity might be getting short shrift.
_________________________________________________________________
Posted using RecPoker.com - http://www.recpoker.com
Matu
familiar to me, but
> Suppose there are two players, lets call
> them A and B. They play exactly the same,
> except that on the end Player A will bluff
> 50 percent of the time and Player B will
> never bluff. Who do you think plays worse?
by stating that both participants are playin exatly the same, except the
bluffing thing, that sort of says that all he cares is the bluffing aspect
and the rest is fuck all, because whatever Player A does B follows...
..and the question was "Who do you think plays worse?", if there are
better ways of doing whatever, it still goes in the general category of
fuck all, as said in a wee bit more nicely, meaningless.
_________________________________________________________________
Bruno
> I finally lost all respect for Mason when he caved into pressure from a few
> censorious complainers, and barred me from 2+2.
Wondered how long it would take him to hijack this thread about his 2+2
ban. Heh.
Vince lepore
essay but I can tell you that if your opponent is the type that never
bluffs then you should never call nor raise. Mason is not idiot nor
are his readers. Except for maybe you. It is obvious that Mason is
correct. Even if he doesn't state it I am sure he is speaking of
making a calling decision based on your holding versus your opponents
potential holding. He is most certainly, again I'm sure, referring to
situations in which your hand can only beat a bluff. Since your
opponent never will bluff then why would you call or raise? Again, it
is obvious that Mason was using this extreme example to emphasize a
point. I have been playing poker for many moons as has Mason
(longer), but not you obviously, I have rarey if ever come accross
someone that NEVER bluffs. However, to illustrate a position, I'm
sure you would agree that one could visualize such an opponent.
Oh..never mind.
Perhaps, that's maybe, someday, although I seriously doutbt it, you
will use you talents for something other than attempting to ridicule
Mason Malmuth. If you believe you have come up with an effective
bluffing strategy then why not just put it out for others to critique.
Vince
Quad...@hotmail.com (QuadNines) wrote in message news:<99fd5b1.02111...@posting.google.com>...
QuadNines
> On Nov 12 2002 9:56AM, QuadNines wrote:
>
>> In his essay "A Note on Bluffing" (POKER ESSAYS: VOLUME III,
>> pp. 88-90), Malmuth attempted to analyze the question of "how
>> often should you bluff?" He got it wrong.
>>
>> Malmuth asked:
>>
>> Suppose there are two players, lets call
>> them A and B. They play exactly the same,
>> except that on the end Player A will bluff
>> 50 percent of the time and Player B will
>> never bluff. Who do you think plays worse?
>>
>> If you've read Malmuth's essay, ignore his answer and think
>> through the problem for yourself. Malmuth got it wrong.
>
> I don't have the essay in front of me, but I remember thinking
> that he forgot to add the condition that your holding can only
> beat a bluff. It seemed to me that this is what he was talking
> about, though.
When Malmuth makes an error, some folks blame it on his poor writing
skills. That excuse is starting to wear thin. Not only does Malmuth
need help as an author, he also needs help as a theorist.
First, Malmuth did not simply forget "to add the condition that your
holding can only beat a bluff." He explicitly stated he doesn't need
to rely on that condition. Again, Malmuth wrote:
I'm aware that you may occasionally hold a
strong hand that you will play--perhaps
with a raise--no matter who bets. But for
practical purposes, these hands have no
impact on the result of this question.
Second, Malmuth still gets it wrong, even if we do add that condition.
Consider Player A, who bluffs half the time. According to Malmuth:
Since he bluffs half the time you should
call him every time. Thus, half the time
you will lose an additional bet, but this
will be canceled out by the other half of
the time when you will win an additional
bet. Your expectation on the end will be
zero.
This still shows a poor understanding of expectation and poker
strategy, even with that added condition. Any decent river strategy
here still will have positive expectation and still will include more
betting actions than just calls.[1]
> I got the idea that the main thrust of the essay was to
> explain why bluffing was essential to limit holdem
> strategy.
That was one of Malmuth's conclusions:
This brings us to our final conclusion.
Bluffing is an important part of poker, and
to be a successful player you should be
prepared to do your share.
If that was his only point, Malmuth's essay could have been two
paragraphs long. But the first half of the essay dealt with whether
Player A or Player B played worse. And Malmuth got it wrong:
This brings us to an interesting conclusion:
Someone who never bluffs plays as badly as
someone who bluffs half the time.
I'm not criticizing Malmuth's unoriginal conclusion that "bluffing is
an important part of poker." I'm criticizing his erroneous conclusion
that "both A and B play equally bad."
> I think most players would realize that the analysis dealt
> with a simplified and contrived poker-like situation, and
> was not intended to directly yield a usable river strategy
> for real games.
As noted above, Malmuth didn't realize that. And, even if we add your
condition, Malmuth still gets it wrong.
> It seems a bit strong to state that Malmouth "seriously
> misunderstands" expectation and strategy.
I think many people give Malmuth too much credit. He clearly isn't
the sharpest tool in the 2+2 shed. On the subject of bluffing, I
believe Malmuth seriously misunderstands expectation and strategy--at
a pretty basic level. Furthermore, bluffing is just one of the
concepts he gets wrong.
> By the way, thanks for writing that essay. You bring up a
> lot of important things to consider when acting on the
> river in limit holdem.
You're welcome, and thanks.
> I just wanted to suggest that the principle of charity
> might be getting short shrift.
Malmuth gets plenty of charity over on 2+2. If more of us questioned
some of his conclusions, we might gain a better understanding of
poker. One way science advances is by people questioning hypotheses
and theories.
------------------
[1] Let's assume: (1) Player A is first to act, (2) your hand loses
to all Player A's checking and value-betting hands, (3) your hand
beats all Player A's bluffing hands, (4) Player A bluffs far more
often than he value bets, and (5) Player A always folds when his
bluffs are raised. For this contrived scenario, Malmuth's strategy
(i.e., always call) might be reasonable, but he still gets it wrong
with his expectation claim.
JonCooke
applies to "How do I play my second tier hands".