Data analysis #002
Gary Carson
Gary Carson
http://www.garycarson.com
_______________________________________________________________
Watch Lists, Block Lists, Favorites - http://www.recpoker.com
Gary Carson
Sorry about that. Wrong button.
I'm still just talking about data and some of the relevant characteristics that
data might have. Today I'm going to talk about a mathematical property of a
relationship between different data -- Transitivity.
We ll get back to more discussion of the nomenclature of different types of data
shortly. But, right now we ll look at the importance of a mathematical property
that is often wrongly assumed to hold with all relations.
One mathematical property that we often implicitly assume, but which sometimes
doesn t hold when dealing with poker related data (or in gambling in general) is
transitivity. That s the property that if A > B and B > C then we know that A >
C.
Whenever we re dealing with processes based on multi-player match-ups, whether
it s a poker game or a basketball tournament, this property of transitivity does
not hold.
Transitivity is actually a property of a relation, the relation > is the one
that s either transitive or not, it s not really a property of the data.
Transitiveity is a necessary property for ranking things in order of size or
value or importance or some other concept of better or bigger. It s entirely
possible to be able to say that X is better than Y without being able to
determine which is best among X, Y, and Z
Because of the need for transitivity, it s often not possible to say who the
best poker player is, or what the best starting hand is.
A football example is probably the best known common example of an intransitive
relation. It s the well known intransitivity of football match-up results that
makes the BCS rankings preferable to a playoff system. With any kind of playoff
system you can only match up teams two at a time. But, it s entirely possible
for LSU to have a better team than UT, UT to have a better team than UCLA, but
UCLA being a better team than LSU. The skill sets that make up a winning team
are complex and there s no reason for individual match-ups to result in
meaningful rankings overall.
The same thing happens when you try to rank starting hands in poker. The
characteristics that give value to a poker hand. Tend to combine in complex
ways, since the deck is finite the hand you re dealt is not independent of other
players hands and are not independent of cards to come. In my book The Complete
Book of Hold Em Poker I give some characteristics of a starting hand in hold em
which give value to the hand: current poker hand, prospects for improvement,
position, characteristics of the game, tendency of opponents to make mistakes.
These are not simple, independent characteristics.
In general it s not possible to logically deduce how these characteristics that
give value to a hand combine or interact to give value, it s an empirical issue
dependent on exact combination of levels of the various characteristics of
inputs.
We can go ahead and use PokerStove right now to look at transitivity in attempts
to rank poker hands.
Let s just look at three hold em hands AsKd, JdTd, and 2h2c.
Which is the best hand of those three? Some people would say the 2h2c because
it s a pair and it beats both the non pair hands. But, that line of thinking
presumes that it s possible to have a showdown with 2 card hands. Of course
that doesn t happen, you never have a showdown until all the cards have been
dealt, a hold em hand at the showdown is the best 5 cards out of 7. So,
although it s possible 2h2c is the best hand it s not because a pair beats no
pair, it s would be because it s the most likely to end up the best hand when
all the cards are out. That s where PokerStove comes in handy.
Set up your PokerStove with the three hands (be sure and specify the suits). If
you don t have a copy of PokerStove then get one, it s a free download at
http://www.pokerstove.com/.
You should get AsKd with 36% chance of winning, JdTd 34% and 2h2c with 30%.
So, of the three, if you think of better in terms of highest chance of ending up
the winner, the best hand is AsKd.
But, let s look at how they stack up if we just compare AsKd and 2h2c, without a
third hand. Rerun PokerStove after removing the JdTd hand and you should get
AsKd with 47% chance of winning, 2h2c 53%. Now what was the worst hand when all
three are compared simultaneously has become the best hand.
The point of this is that the concept of best when measured by pair-wise
match-ups does not have the mathematical property known as the transitive
property. The implication of that is that you can t create a meaningful ranking
of hands.
Next week I'm going to talk some more about transitivity. I'm still on the 1st
chapter and will be for another 2-3 weeks.
I've made some changes to Chapter 1 and posted it on my website for download.
I'm also putting some links to supplimentary material on the site.
CommentS? Questions?
On Feb 28 2006 6:30 PM, Gary Carson wrote:
>
> Gary Carson
> http://www.garycarson.com
>
>
Gary Carson
http://www.garycarson.com
_______________________________________________________________
Your Online Poker Community - http://www.recpoker.com
Dwight Abbott
Great stuff Gary. Very clear and it has helped me to understand transitive as
applied (or not applied) to poker hands. Please keep going.
Dwight
On Feb 28 2006 4:36 PM, Gary Carson wrote:
>
>
> Sorry about that. Wrong button.
>
>
> I'm still just talking about data and some of the relevant characteristics
> that
> data might have. Today I'm going to talk about a mathematical property of a
> relationship between different data -- Transitivity.
>
> We’ll get back to more discussion of the nomenclature of different types of
> data
> shortly. But, right now we’ll look at the importance of a mathematical
> property
> that is often wrongly assumed to hold with all relations.
> <strong />
> <a href="http://www.pokerstove.com/" target="_blank">www.pokerstove.com</a>.
_______________________________________________________________
The Largest Online Poker Community - http://www.recpoker.com
davidwatts
: CommentS?? Questions?
No questions. The material was well organized and easily
understood. Thanks!
cb
James L. Hankins
"Gary Carson" <garyc...@alumni.northwestern.edu> wrote in message
news:1141173411$746...@recpoker.com...
Would AA be an exception to the non-transivity rule?
bonkey
You had me until you said BCS was good.
--
Bonkey
"Jodie Foster held two pair, Bach had three of a kind. Gandhi said, 'With my
full house, I will blow your mind'" --They Might be Giants
_______________________________________________________________
* New Release: RecPoker.com v2.2 - http://www.recpoker.com
Gary Carson
On Feb 28 2006 7:18 PM, James L. Hankins wrote:
>
>
> Would AA be an exception to the non-transivity rule?
No.
It's the relation "better than" than's not transitive, not one value. It only
makes sense to even talk about transitivity when you have three values.
There are many 3-tuples of hands for which "better than" is transitive, AA KK QQ
is an example.
But all it takes to show that the "better than" relation when applied to all 169
hands isn't transitive is to find one 3-tuple that doesn't satisfy transitivity.
You can't rank starting poker hands in poker if you can't rank AK JTs and 22 if
your concept of poker hands is that X is better than Y if and only if X beats Y
in a matchup.
The concept of "better than" is transitive if you mean to rank the hands
according to their pot equity against a random hand. Or against a big pair. Or
against two random hands. Next week I'll get into that.
Those criteria will give transitive rankings, but they still aren't useful
because they aren't stable or meaningful. Each of the different criteria listed
above gives you a different ranking.
This idea of hands having value is an emperical question, one that can only be
answered in the conext of the exact situation. And, the specification of that
situaiton can require some subtle distinctions.
For example, I had a doctors appointment at the VA in Muskogee today and stopped
by the Creek Casino there for a couple of hours after. Played 1/2 NL HE. Won
$160. $140 of that win was when I called $10 preflop with a 96c and flopped two
pair against a KK held by a straightforward tight ass who couldn't lay down an
overpair.
What does that mean about the value of 96c? Nothing. It says something about
calling a raise with a couple of mid sized cards against a straightforward
tightass who can't lay down an overpair.
It's all situational and much of the situaiton has very little to do with the
hand.
But, that's next week. I'm not sure I can, but I'll see if I can come up with a
situation for next week where AA isn't the best hand to have preflop. I think I
can do that.
Gary Carson
On Feb 28 2006 7:40 PM, bonkey wrote:
>
> You had me until you said BCS was good.
I didn't say it was good.
I said it's preferable to a playoff system.
The whole idea of crowning a "best team" is just pandering to ignorance.
>
> --
>
> Bonkey
> "Jodie Foster held two pair, Bach had three of a kind. Gandhi said, 'With my
> full house, I will blow your mind'" --They Might be Giants
>
>
Gary Carson
http://www.garycarson.com
_______________________________________________________________
Posted using RecPoker.com v2.2 - http://www.recpoker.com
Gary Carson
btw, I'll try next to cut and past from Word to notepad then a seperate cut and
past to a recpoker.com message window.
Last week I thought the bold came from a cut and paste where the beginning was a
header line in bold. But none of today's stuff was bold in the word file.
So, I don't know why it's coming up bold.
Gary Carson
http://www.garycarson.com
_______________________________________________________________
The Largest Online Poker Community - http://www.recpoker.com
Omaha8_Beach
on rock, paper, scissors and it went into some detail on the game being
a nontransitive relation.
It gave several examples of modified versions of rock paper
scissors. My favorite: "cockroach, nuclear explosion, foot".
Cockroach beats nuclear explosion, nuclear explosion beats foot, foot
beats cockroach.
Ken
Gary Carson
On Feb 28 2006 8:18 PM, Omaha8_Beach wrote:
> Good stuff, Gary. Recently I was reading the Wikipedia article
> on rock, paper, scissors and it went into some detail on the game being
> a nontransitive relation.
>
> It gave several examples of modified versions of rock paper
> scissors. My favorite: "cockroach, nuclear explosion, foot".
> Cockroach beats nuclear explosion, nuclear explosion beats foot, foot
> beats cockroach.
The podiatrist told me today that I should stop wearing my pointy toed cowboy
boots (I'm diabetic and have potential for foot problems). So, foot doesn't
even beat cockroach any more if they get in the corner.
> > > http://www.garycarson.com/
> > >
> > >
> > Gary Carson
> > http://www.garycarson.com/
> >
> >
> >
> >
> > _______________________________________________________________
> > Your Online Poker Community - /
Gary Carson
http://www.garycarson.com
_______________________________________________________________
The Largest Online Poker Community - http://www.recpoker.com
eleaticus
"Gary Carson" <garyc...@alumni.northwestern.edu> wrote in message
news:1141173411$746...@recpoker.com...
\> A football example is probably the best known common example of an
intransitive
> relation. It s the well known intransitivity of football match-up results
that
> makes the BCS rankings preferable to a playoff system. With any kind of
playoff
> system you can only match up teams two at a time.
Gary, I appreciate your efforts and figure(d) to learn something from them,
but that is just complete nonsense.
The BCS - by intention - is a two-team playoff.
Also, by your own admission, the second ranked team might be the only team
of the top (3/4/5/6/7/8/9/10/n) that would be likely to lose to the first
ranked team. Or vice versa.
The craphead BCS might keep AA out of the hand. What a rule that is: pocket
aces must fold because some dunderheads decided JT and 22 were better hands.
eleaticus
ee-lee-AT-i-cus
Michael George Lerner
> CommentS?? Questions?
I've taken a lot of math courses, so no questions yet.
Comments: this is good work. I'm looking forward to the
later stuff. You deserve a lot of credit and thanks for
putting this together for free.
Thanks,
-michael
eleaticus
"James L. Hankins" <jhan...@cox.net[no spam]> wrote in message
news:yD6Nf.43998$Dh.1276@dukeread04...
>
> Would AA be an exception to the non-transivity rule?
Transitivity requires at least three 'hands'. A>B>C
eleaticus
ee-lee-AT-i-cus
>
>
eleaticus
>
>
>
> On Feb 28 2006 7:40 PM, bonkey wrote:
>
> >
> > You had me until you said BCS was good.
>
> I didn't say it was good.
>
> I said it's preferable to a playoff system.
>
> The whole idea of crowning a "best team" is just pandering to ignorance.
You reallly haven't heard about the BCS championship game?
eleaticus
Gary Carson
start to finish since I was at LSU. Should I take out the reference to BCS?
On Feb 28 2006 8:52 PM, eleaticus wrote:
> "Gary Carson" wrote in message
> news:1141178589$746...@recpoker.com...
> >
> >
> >
> > On Feb 28 2006 7:40 PM, bonkey wrote:
> >
> > >
> > > You had me until you said BCS was good.
> >
> > I didn't say it was good.
> >
> > I said it's preferable to a playoff system.
> >
> > The whole idea of crowning a "best team" is just pandering to ignorance.
>
> You reallly haven't heard about the BCS championship game?
>
> eleaticus
>
> >
> >
> > >
> > > --
> > >
> > > Bonkey
> > > "Jodie Foster held two pair, Bach had three of a kind. Gandhi said,
> 'With my
> > > full house, I will blow your mind'" --They Might be Giants
> > >
> > >
> > Gary Carson
> > http://www.garycarson.com/
> >
> >
> >
> > _______________________________________________________________
> > Posted using RecPoker.com v2.2 - /
Peg Smith
>Next week I'm going to talk some more about transitivity.
My eyes didn't glaze over, so you know it's still understandable.
Again, very well written.
Peg
Gary Carson
On Feb 28 2006 10:12 PM, Peg Smith wrote:
> Gary Carson wrote:
>
> >Next week I'm going to talk some more about transitivity.
>
> My eyes didn't glaze over, so you know it's still understandable.
> Again, very well written.
>
> Peg
Gary Carson
http://www.garycarson.com
_______________________________________________________________
New Feature: Mark All As Read! - http://www.recpoker.com
Chris in Texas
On Feb 28 2006 10:21 PM, Gary Carson wrote:
> I knew some transitives when I lived in SF.
>
who were better than the ones in Austin who were better than the ones in Tunica
who were better than the ones in SF?
See, I got it.
Good stuff Gary. I understand where you're going with the BCS example, but
kinda agree with the others it should be beefed up or dropped. The trouble with
the BCS example, IMHO, is football, like many things, is all about Mano-a-mano
competition, not 3-way. That's why we want a playoff system. Think Oakland
Raiders 1981-ish, first wildcard team to win the SuperBowl. On paper, they were
the 8th seed??? (memory fails me), but they played the best in January. Not
many can recall who the best team on paper was going into the playoffs that
year.
One thing I'd like to see about your AK/JT/22 example is include the JTs vs 22
to drive the point home. Many people (incl me recently) are under the
assumption any pocker pair is a slight favorite over any two overcards.
Looking forward to next instalment,
Chris
_______________________________________________________________
Block Lists, Favorites, and more - http://www.recpoker.com
bonkey
On Feb 28 2006 8:05 PM, Gary Carson wrote:
>
> btw, I'll try next to cut and past from Word to notepad then a seperate cut
> and
> past to a recpoker.com message window.
>
> Last week I thought the bold came from a cut and paste where the beginning was
> a
> header line in bold. But none of today's stuff was bold in the word file.
>
> So, I don't know why it's coming up bold.
>
>
> Gary Carson
> http://www.garycarson.com
>
>
You can turn off the bold. In the recpoker window just highlight your text and
hit Ctrl b. bold not bold bold not bold bold italics (ctrl i) not bold italics
--
Bonkey
"Jodie Foster held two pair, Bach had three of a kind. Gandhi said, 'With my
full house, I will blow your mind'" --They Might be Giants
_______________________________________________________________
leande...@aol.com
account for 2h2c > AsKd by 53% with stovepipe. But even with the JdTd
removed from the deck I still get 53.32%. and more strange JdTd > 2h2c
by 52.25% with the AsKd removed from deck?
Michael Sullivan
> No. I bet on horses, keno, and poker. I havn't watched a football game
> from start to finish since I was at LSU. Should I take out the reference
> to BCS?
Let's just say it's not self evident to anyone who doesn't already
understand much of what you'll be teaching that the BCS (or something
similar which doesn't take polls into account) would actually be
preferable to a playoff system in determining the best team.
So yeah, I'd take it out.
Michael
eleaticus
"Michael Sullivan" <use-re...@spambegone.null> wrote in message
news:1hbippw.11eeg8qa5hdylN%use-re...@spambegone.null...
Both of you are espousing idiocy.
For your position to be correct a couple of nonsensical ideas would have to
be true:
The major one is that the season to date, with much less inter-conference
play than would be true after 2-3 more (playoff) games, must somehow have
produced the right team. Yet, the added games required to reach the
championship game would increase the data sample greatly.
(And, for purposes of best computation, increased the relevant and revealing
data relationships tremendously. The multiplicative principle in this
regard is similar to adding another team to say, a 4-picks 3-way-parlay
ticket. With just four picks you have just four 3-ways. Add a fifth and you
have ten. Add a sixth and you have 20.)
If you don't understand that, it may be why - with pure number crunching and
no data entry or fudging based on knowledge of the teams (immediate or long
term) - I went 2-0, 4-0, 3-0, 3-0, 0-0, 2-1, 2-2 on the NFL this past
season and you couldn't in your wildest dreams. (I'da done it on the NBA
but I can't collect the data necessary.)
Your Bat Crap System could have produced JT versus 22 and left AA out in the
cold. If the AA were included in a four-hand playoff, it would be the most
likely to win out, but your moronic BCS > playoff idiocy says it shouldn't
play.
Implicit in your nonsense is an idea you yourself contradicted, Gary: that
the BCS could pick the best team. Well, a repeated test-by-fire is the way
to go, not a test by moronic BullCrapSystem formulae.
eleaticus
ee-lee-AT-i-cuss
Aardvark
> But, that's next week. I'm not sure I can, but I'll see if I can come up with a
> situation for next week where AA isn't the best hand to have preflop. I think I
> can do that.
It isn't 3 hands, but in the following situation AA is no longer the
leader as far as EV:
Ad As (EV: 0.200)
Ac Ah (EV: 0.205)
8h 7h (EV: 0.237)
Kd Ks (EV: 0.191)
Qd Qs (EV: 0.167)
BTW, good post.
-Tom.
Gary Carson
On Mar 1 2006 1:34 PM, eleaticus wrote:
> "Michael Sullivan" wrote in message
> news:1hbippw.11eeg8qa5hdylN%use-re...@spambegone.null...
> > Gary Carson wrote:
> >
> > > No. I bet on horses, keno, and poker. I havn't watched a football game
> > > from start to finish since I was at LSU. Should I take out the
> reference
> > > to BCS?
> >
> > Let's just say it's not self evident to anyone who doesn't already
> > understand much of what you'll be teaching that the BCS (or something
> > similar which doesn't take polls into account) would actually be
> > preferable to a playoff system in determining the best team.
>
> Both of you are espousing idiocy.
>
> For your position to be correct a couple of nonsensical ideas would have to
> be true:
My position is curently that since I have no clue what I'm talking about when I
reference BCS that I should just avoid the subject?
Are you saying that's a stupid positon?
Gary Carson
http://www.garycarson.com
Gary Carson
But how do I know whether it's bold or not? It doesn't show as bold in the
window I'm posting from, it isn't bold in the original document, and all if it
didn't end up bold in the posted message.
On Mar 1 2006 9:05 AM, bonkey wrote:
>
>
> On Feb 28 2006 8:05 PM, Gary Carson wrote:
>
> >
> > btw, I'll try next to cut and past from Word to notepad then a seperate cut
> > and
> > past to a recpoker.com message window.
> >
> > Last week I thought the bold came from a cut and paste where the beginning
> > was
> > a
> > header line in bold. But none of today's stuff was bold in the word file.
> >
> > So, I don't know why it's coming up bold.
> >
> >
> > Gary Carson
> > http://www.garycarson.com
> >
> >
>
> You can turn off the bold. In the recpoker window just highlight your text and
> hit Ctrl b. bold not bold bold not bold bold italics (ctrl i) not bold italics
>
> --
>
> Bonkey
> "Jodie Foster held two pair, Bach had three of a kind. Gandhi said, 'With my
> full house, I will blow your mind'" --They Might be Giants
>
>
Gary Carson
http://www.garycarson.com
_______________________________________________________________
Your Online Poker Community - http://www.recpoker.com
Gary Carson
Split pots get averaged in PokerStove and removing the AK from the deck means
fewer playtheboard split pots.
Gary Carson
http://www.garycarson.com
_______________________________________________________________
The Largest Online Poker Community - http://www.recpoker.com
eleaticus
>
>
>
> On Mar 1 2006 1:34 PM, eleaticus wrote:
>
> > "Michael Sullivan" wrote in message
> > news:1hbippw.11eeg8qa5hdylN%use-re...@spambegone.null...
> > > Gary Carson wrote:
> > >
> > > > No. I bet on horses, keno, and poker. I havn't watched a football
game
> > > > from start to finish since I was at LSU. Should I take out the
> > reference
> > > > to BCS?
> > >
> > > Let's just say it's not self evident to anyone who doesn't already
> > > understand much of what you'll be teaching that the BCS (or something
> > > similar which doesn't take polls into account) would actually be
> > > preferable to a playoff system in determining the best team.
> >
> > Both of you are espousing idiocy.
> >
> > For your position to be correct a couple of nonsensical ideas would have
to
> > be true:
>
> My position is curently that since I have no clue what I'm talking about
when I
> reference BCS that I should just avoid the subject?
>
> Are you saying that's a stupid positon?
LOL. That's a pretty good response but saying you would remove the
reference and that you didn't know the BCS' purpose is to pick the two-only
playoff teams doesn't quite do it.
I finally infer that you thought the BCS was a kind of rating based on a
body of other ratings, and hence a rational 'number'.
Gary, the BCS system started silly and then got more and more absurd.
And is based on at least two fewer critical tests than a playoff would
allow.
Innyhoo, that leaves michael sullivan's rather putrid suggestion that HE
understands how your transitivity approach to the BCS was right.
He understands? rofflmfao!
And the ignnoramous doesn't know the BCS is based on polls.
And such ridiculous ideas such as the 'computers' shouldn't give team A
fewer 'points' for beating B with a last second field goal than team C
should for beating B with a 31 point edge each of the four quarters.
The components of the BCS are increasingly ridiculous as more and more
politics enters its fudging and those components are not honestly evaluated.
eleaticus
ee-lee-AT-i-cus
Gary Carson
On Mar 1 2006 8:43 PM, eleaticus wrote:
> LOL. That's a pretty good response but saying you would remove the
> reference and that you didn't know the BCS' purpose is to pick the two-only
> playoff teams doesn't quite do it.
>
> I finally infer that you thought the BCS was a kind of rating based on a
> body of other ratings, and hence a rational 'number'.
I don't think there's anything at all rational about trying to identify the
"best" when it's pretty clear that no such thing as "best" exists.
Just because I said approach A is better than approach B doesn't mean I think
either one is very good, or that when compared to a 3rd approach any of the 3 is
the best approach to determining the best team.
Keno makes sense to me. I'll bet on Keno. But, football makes no sense at all,
I don't bet on it.
In the 7th grade I played football. We had one 7th grader who was big and fast
and could accurately throw a football and was going to win games for us.
Before our first game he was practicing with the 8th graders, a week later he
was on the 9th grade team.
So it became clear to me that it winning football games was supposed to be my
objective, but our coaches objective was to please the head coach by giving up
our chances so that the head coaches 9th grade team could win some games.
That's when I gave up on football. Fuck those assholes.
I gave up on sports in the 9th grade when I was the fastest 220 yard sprinter in
the school and the coach said I couldn't compete in any meets unless I stopped
smoking. To him it wasn't about the sport or competition, it was about
controlling a bunch of kids. Fuck all of them.
I went to the games at LSU because the student end zone seating area was just
one huge party. Half the time I didn't even know who the opposing team was.
>
> And the ignnoramous doesn't know the BCS is based on polls.
That was the part that I claimed made it superior to playoffs. Polls allow for
consideration of complex interactions of skill sets to be taken in consideratoin
in ways that just can't be done in your typical single elimination tournament
structures.
But, it's not worth arguing about it, I'll concede I don't know and don't really
care anything at all about college football.
Gary Carson
http://www.garycarson.com
_______________________________________________________________
John Forsberg
you know you have an audience. I'll pipe up once something seems unclear or if I
think I've got something meaningful to say.
_______________________________________________________________
* New Release: RecPoker.com v2.2 - http://www.recpoker.com
Gary Carson
statistics. Most of that is going to be basic except for the discussion about
different viewpoints of how to measure dispersion, in particular tail behavior.
That's still going to be basic, but will be non-tradional That's a prelude to
risk analysis which comes much later. Probably at least the next 3-4 weeks will
be pretty basic.
On Mar 1 2006 9:43 PM, John Forsberg wrote:
> This stuff is a little too basic for me right now, but I just thought I'd let
> you know you have an audience. I'll pipe up once something seems unclear or if
> I
> think I've got something meaningful to say.
>
Gary Carson
http://www.garycarson.com
_______________________________________________________________
Watch Lists, Block Lists, Favorites - http://www.recpoker.com
Dwight Abbott
I know you were saving this as an exercise for next week, but I was curious and
really didn't think you could find a situation where AA isn't the best preflop
hand. But here is one. I'm sure there are others.
cards win %win lose %lose tie %tie EV
Ac Ad 52070 18.71 224614 80.72 1572 0.56 0.188
As Qs 16658 5.99 255934 91.98 5664 2.04 0.068
Ah Qh 16658 5.99 255934 91.98 5664 2.04 0.068
Kc Kd 60 0.02 274342 98.59 3854 1.39 0.006
Ks Kh 0 0.00 274402 98.61 3854 1.39 0.005
2c 2d 29560 10.62 247568 88.97 1128 0.41 0.107
3c 3d 35478 12.75 241650 86.84 1128 0.41 0.128
4c 4d 45836 16.47 231292 83.12 1128 0.41 0.165
8s 8h 73546 26.43 203582 73.16 1128 0.41 0.265
The basic concept to making AA a non favorite preflop seems to be based in
taking away all it's outs. i.e. put Ace outs, straight outs and flush outs in
other hands. And then have one hand, in this situation 8s8h that has very few
of it's outs taken away by the other hands.
Dwight
>
> Gary Carson
> http://www.garycarson.com
>
>
_______________________________________________________________
Dwight Abbott
On Mar 1 2006 7:49 PM, Gary Carson wrote:
> Thanks. It's going to be pretty basic until I get to chapter 3 -- descriptive
> statistics. Most of that is going to be basic except for the discussion about
> different viewpoints of how to measure dispersion, in particular tail
> behavior.
> That's still going to be basic, but will be non-tradional That's a prelude to
> risk analysis which comes much later. Probably at least the next 3-4 weeks
> will
> be pretty basic.
>
I think a review of the basics is absolutely necessary even for those who have
seen the material before. It gets everyone into the mood and sets the
terminology and tone for the coming lessons.
Dwight
> On Mar 1 2006 9:43 PM, John Forsberg wrote:
>
> > This stuff is a little too basic for me right now, but I just thought I'd
> > let
> > you know you have an audience. I'll pipe up once something seems unclear or
> > if
> > I
> > think I've got something meaningful to say.
> >
> Gary Carson
> http://www.garycarson.com
>
>
_______________________________________________________________
ML
Gary Carson wrote:
>
> Keno makes sense to me. I'll bet on Keno. But, football makes no sense at all,
> I don't bet on it.
Do you happen to bet on the ponies?
> I gave up on sports in the 9th grade when I was the fastest 220 yard sprinter in
> the school and the coach said I couldn't compete in any meets unless I stopped
> smoking. To him it wasn't about the sport or competition, it was about
> controlling a bunch of kids.
You would concede that smoking is not helpful in becoming a faster
sprinter, right? I don't know this coaches motives (do we truly know
anyone's motives?), but he may have just been a stickler for policy, or
maybe had been in trouble for past liberalities with the students
habits. Who knows? Maybe he actually cared about the health and
well-being of the students.
--
ML
Gary Carson
That's very contrived and not the kind of example I'm thinking about.
Since it's virtually impossible to put playrer on that kind of range of specific
hands the example isn't really a poker example.
What I was talking about is that I'm guessing that If you know something
like that two players have the range of hands AA KK AK and third player has the
range AA KK AK QQ and the action is on you it might not be a good thing to have
AA.
I'm not sure, I havn't actually checked that, but if you change the annual WSOP
question of if every goes all in on the first hand of the WSOP do you call with
AA to if 3 very tight players go all in ahead of you do you call with AA it's
possible the answer is fold.
You can sometimes put opponents on very tight ranges of hands. You're never
going to put them on exactly 88
I'll talk about that some (about why pokerstove is a better tool than twodimes)
when I get to chapter 2.
Thanks though. I don't mean to totally disocunt your example. It demonstrates
clearly why you can't just assume the "best hand" is the best hand.
> > http://www.garycarson.com/
> >
> >
>
>
Gary Carson
http://www.garycarson.com
_______________________________________________________________
Gary Carson
On Mar 2 2006 10:29 AM, ML wrote:
> Gary Carson wrote:
>
> >
> > Keno makes sense to me. I'll bet on Keno. But, football makes no sense at
> > all,
> > I don't bet on it.
>
> Do you happen to bet on the ponies?
Yes, the first gambling writing I did was about horses. Racing Star Weekly and
American Turf Monthly in the 80's.
>
> > I gave up on sports in the 9th grade when I was the fastest 220 yard
> > sprinter in
> > the school and the coach said I couldn't compete in any meets unless I
> > stopped
> > smoking. To him it wasn't about the sport or competition, it was about
> > controlling a bunch of kids.
>
> You would concede that smoking is not helpful in becoming a faster
> sprinter, right?
It doesn't really hurt anything either. It does long term. But, although I did
flunk the 9th grade I wasn't really in jr. high long term.
I don't know this coaches motives (do we truly know
> anyone's motives?), but he may have just been a stickler for policy, or
> maybe had been in trouble for past liberalities with the students
> habits. Who knows? Maybe he actually cared about the health and
> well-being of the students.
I ddin't care what he cared about. His job was to coach the track team, not to
be my daddy.
Of course my attitude about him was a little biased because he was a member of
a secret society of grown men who frequently beat little boys with big sticks.
That a real careing individual.
Don't get me started on wanna be men who coached jr high sports in the 1960's.
They were all fucking losers with no moral direction. Child abusers every
fucking one of them. Almost as bad as priests.
Gary Carson
http://www.garycarson.com
_______________________________________________________________
ML
paddling the boys. I sat in the hallway for half a semester during
metal shop class because I would not let the teacher paddle me for
being out of my seat (never mind the fact that 2 other boys were
fighting and EVERYONE was out of their seat watching).
I really hope that paddling the boys didn't gave these "men" any type
of sexual jollies. Yuck.
I wonder how tough those type "men" are around people their own size.
Later as an adult, I always wanted a chance to meet a couple of those
type teachers somewhere dark and alone to find out how tough they
really were. Never got the chance.
--
ML
Gary Carson
On Mar 2 2006 10:26 AM, Dwight Abbott wrote:
>
>
> On Mar 1 2006 7:49 PM, Gary Carson wrote:
>
> > Thanks. It's going to be pretty basic until I get to chapter 3 --
> > descriptive
> > statistics. Most of that is going to be basic except for the discussion
> > about
> > different viewpoints of how to measure dispersion, in particular tail
> > behavior.
> > That's still going to be basic, but will be non-tradional That's a prelude
> > to
> > risk analysis which comes much later. Probably at least the next 3-4 weeks
> > will
> > be pretty basic.
> >
>
> I think a review of the basics is absolutely necessary even for those who have
> seen the material before. It gets everyone into the mood and sets the
> terminology and tone for the coming lessons.
Yes, that's why I'm doing it. It puts everybody into at least similar mindsets
even though backgrounds differe wildly. There's people here with PhD's in math
and some who flunked business math in the 9th grade.
And all this stuff is important for later stuff. The transitivity issue will
come back in Chapter 21 or so when I get to Game Theory. Math geeks who
understand transivity very well seem to just ignore it when they get to game
theory, perfectly willing to assume that it makes sense to talk about optimal
stratagies as being of the form Bet with any hand "h" such that "h" is better
than the optimal threshhold hand "X".
I first got exposed to this common error in game theory models when I was a
graduate student and read an article by Zadah in Operations Research. He
claimed to determine optimal opening hands, by position, in draw poker. Myself
and a friend, neither of us played poker but were interested in the math of
gambling, noticed that he just assumed you would open with hand h and all better
hands.
But, if you open with AAxxx or better, is KhQhJhThx better than AA or not? He
ranked hands by pre-draw EV.
In the example he gave in that article it's not. But, he used a very small
ante. Increase the ante and the one card draw to a straight flush has
more pre-draw EV than the AA (also better implied odds, but he treated the game
as one with a draw but no betting post draw).
Ever since then it just amazes me how many smart people are willing to just
close their eyes to lack of transitivity when they analyze game theory models.
That basic idea that when the pot gets big enough from dead money (whether from
antes or from multi players doesn't matter) you're sometimes better off with a
draw than with the "best hand" was the entire basis for my hold'em book. And,
that all results from the fundemental lack of transitiviey.
Gary Carson
http://www.garycarson.com
_______________________________________________________________
Michael Sullivan
> And the ignnoramous doesn't know the BCS is based on polls.
That must be why I said something like "or rather, a similar system that
doesn't use polls as input". Pay attention, eh?
I assume Gary was suggesting that some maximum-likelihood based computer
algorithm be used to determine the "best team" at the end of the season.
I think he's right that such a system would more consistently produce
the "best" team than a single elimination playoff system. It's also
certainly true that playing a bunch of playoff games and feeding them
into the same algorithm would work even better. But the system might
not end up calling the team that won the playoffs #1.
As a practical solution, these are all silly. 90% of people don't care
who the best team is, what they want is somebody who is easy to get
behind as a "winner" and a playoff provides that in a way that no
mathematical rating system can.
I never said that the existing BCS isn't a pretty foolish algorithm.
In fact, it's possible that it is *so* bad mathematically, that a
traditional 8 or 16 team playoff would actually do better at producing
the best team than just crowning the BCS #1, but I rather doubt it.
Michael
Michael Sullivan
> This stuff is a little too basic for me right now, but I just thought I'd
> let you know you have an audience. I'll pipe up once something seems
> unclear or if I think I've got something meaningful to say.
AOL.
I'm also going to be dropping reading r.g.p. for a while (work is
getting to involve more actual work), but I'll check in now and again
for this series.
Michael
eleaticus
If we must have a BCS then I wish that each year there would be a playoff,
not amongst teams, but candidates for use in the formula.
Using the last regular season poll/ranking/computer numbers, let them each
and all be used to pick the winning bowl teams with a stated margin to be
used as first tie-breaker, second to be picking against a concensus line.
Then use only those candidates the next year that had done best. And if none
do very well, give me the data I need and I'll do it right.
The "'puters", if any of them are any good at all, should stomp the polls
(which couldn't compete on the tie-breaker.)
eleaticus
ee-lee-AT-i-cus
Peg Smith
>Ever since then it just amazes me how many smart people are willing to just
>close their eyes to lack of transitivity when they analyze game theory models.
You've done a good job with your lessons so far, shown by the fact
that two weeks ago I wouldn't have had a clue what that sentence
means, but now I understand it perfectly.
Peg
Gary Carson
team, then worry about who's second best later. That was pretty much the
consensus of experts I knew when I was in college.
Gary Carson
http://www.garycarson.com
_______________________________________________________________
Your Online Poker Community - http://www.recpoker.com
Dwight Abbott
On Mar 2 2006 8:30 AM, Gary Carson wrote:
>
> That's very contrived and not the kind of example I'm thinking about.
>
> Since it's virtually impossible to put playrer on that kind of range of
> specific
> hands the example isn't really a poker example.
>
> What I was talking about is that I'm guessing that If you know something
> like that two players have the range of hands AA KK AK and third player has
> the
> range AA KK AK QQ and the action is on you it might not be a good thing to
> have
> AA.
>
I would think you would always be +EV in this situation. This is based on some
playing aroud I did in Pokerstove. I may not be understanding you completely so
I'll wait for the example.
> I'm not sure, I havn't actually checked that, but if you change the annual
> WSOP
> question of if every goes all in on the first hand of the WSOP do you call
> with
> AA to if 3 very tight players go all in ahead of you do you call with AA it's
> possible the answer is fold.
>
Yeah, the eternal question. I have a hard time seeing how anyone would fold
here though. Whether it is the right choice is possible I guess. Seems like
you would always be +EV, but want to wait for a better edge?
> You can sometimes put opponents on very tight ranges of hands. You're never
> going to put them on exactly 88
>
> I'll talk about that some (about why pokerstove is a better tool than
> twodimes)
> when I get to chapter 2.
>
I started with pokerstove, but it wouldn't let me cut and paste the results from
it. I was using the free version, so maybe if I paid for it, I could cut and
paste.
> Thanks though. I don't mean to totally disocunt your example. It
> demonstrates
> clearly why you can't just assume the "best hand" is the best hand.
>
I actually started out trying to prove that AA would always be the best hand and
was getting ready to tell you I was very doubtful it would ever be subordinate
to any hand preflop. But then I got the idea of putting all of it's outs in
other hands and was surprised to see it worked.
Dwight
> > > <a href="http://www.garycarson.com/" target="_blank">http://www.garycarson.com</a>
> > >
> > >
> >
> >
> Gary Carson
> http://www.garycarson.com
>
>
_______________________________________________________________
Posted using RecPoker.com v2.2 - http://www.recpoker.com
Gary Carson
On Mar 2 2006 2:02 PM, Dwight Abbott wrote:
>
>
> On Mar 2 2006 8:30 AM, Gary Carson wrote:
>
> >
> > That's very contrived and not the kind of example I'm thinking about.
> >
> > Since it's virtually impossible to put playrer on that kind of range of
> > specific
> > hands the example isn't really a poker example.
> >
> > What I was talking about is that I'm guessing that If you know something
> > like that two players have the range of hands AA KK AK and third player has
> > the
> > range AA KK AK QQ and the action is on you it might not be a good thing to
> > have
> > AA.
> >
>
> I would think you would always be +EV in this situation. This is based on
> some
> playing aroud I did in Pokerstove. I may not be understanding you completely
> so
> I'll wait for the example.
You might be right, probably are. I really havn't investigated it thourghyily.
I really have two points. One is that there's no logical reason to conclude one
way or another whether it's plus EV for the AA. It's an emperical question
that has to be investigated pretty much by inumeration rather than logical
thought. And, the second point is that AA might well not be the best hand to
have in the situation, whether it's plus EV or not.
>
> > I'm not sure, I havn't actually checked that, but if you change the annual
> > WSOP
> > question of if every goes all in on the first hand of the WSOP do you call
> > with
> > AA to if 3 very tight players go all in ahead of you do you call with AA
> > it's
> > possible the answer is fold.
> >
>
> Yeah, the eternal question. I have a hard time seeing how anyone would fold
> here though. Whether it is the right choice is possible I guess. Seems like
> you would always be +EV, but want to wait for a better edge?
Not always. Here's a contrived lineup where AA is minus EV for an allin preflop
if the blinds are very small
5,104,008 games 0.050 secs 102,080,160 games/sec
Board:
Dead:
equity (%) win (%) tie (%)
Hand 1: 19.9297 % 01.85% 18.08% { AA }
Hand 2: 19.9297 % 01.85% 18.08% { AA }
Hand 3: 18.8467 % 18.78% 00.07% { KhKs }
Hand 4: 16.6011 % 16.54% 00.07% { QhQs }
Hand 5: 24.6928 % 24.63% 00.07% { 6c5c }
>
> > You can sometimes put opponents on very tight ranges of hands. You're never
> > going to put them on exactly 88
> >
> > I'll talk about that some (about why pokerstove is a better tool than
> > twodimes)
> > when I get to chapter 2.
> >
>
> I started with pokerstove, but it wouldn't let me cut and paste the results
> from
> it. I was using the free version, so maybe if I paid for it, I could cut and
> paste.
>
Just hilight the results using the cursor and do a cntl-C to copy to the memory
pad, cntl-V to paste it somewhere.
>
>
> > Thanks though. I don't mean to totally disocunt your example. It
> > demonstrates
> > clearly why you can't just assume the "best hand" is the best hand.
> >
>
> I actually started out trying to prove that AA would always be the best hand
> and
> was getting ready to tell you I was very doubtful it would ever be subordinate
> to any hand preflop. But then I got the idea of putting all of it's outs in
> other hands and was surprised to see it worked.
I'm pretty sure it's going to always be favored if at least one ace remains in
the stub.
>
Gary Carson
http://www.garycarson.com
_______________________________________________________________
Gary Carson
I was wasting time this morning on the internet and ran across an article on
ranking things from the Journal of Statistical Education.
It's only a little technical, probably requires at least a good course in stat
methods as a prerequisite. I've only skimmed it, but I think it's funny. Tons
and tons of masturbation trying to be technically accurate in performing a task
that's really not doable or meaningful anyway.
It describes the procedures they followed in ranking towns for an article on
"healthist towns" for Boston Magazine.
http://www.amstat.org/publications/jse/v13n3/sharpe.html
I think doing such rankings are fun, but you really shouldn't pretend the
ranks mean something. However the process you go thru do get some rankings does
man something and the article really delves into the process.
So, although I laugh at it, and call it masturbation, I think it's a good
article.
On Feb 28 2006 6:36 PM, Gary Carson wrote:
>
>
> Sorry about that. Wrong button.
>
>
> I'm still just talking about data and some of the relevant characteristics
> that
> data might have. Today I'm going to talk about a mathematical property of a
> relationship between different data -- Transitivity.
>
> We ll get back to more discussion of the nomenclature of different types of
> data
> shortly. But, right now we ll look at the importance of a mathematical
> property
> that is often wrongly assumed to hold with all relations.
>
> One mathematical property that we often implicitly assume, but which sometimes
> doesn t hold when dealing with poker related data (or in gambling in general)
> is
> transitivity. That s the property that if A > B and B > C then we know that A
> >
> C.
>
> Whenever we re dealing with processes based on multi-player match-ups, whether
> it s a poker game or a basketball tournament, this property of transitivity
> does
> not hold.
>
> Transitivity is actually a property of a relation, the relation > is the one
> that s either transitive or not, it s not really a property of the data.
> Transitiveity is a necessary property for ranking things in order of size or
> value or importance or some other concept of better or bigger. It s entirely
> possible to be able to say that X is better than Y without being able to
> determine which is best among X, Y, and Z
>
> Because of the need for transitivity, it s often not possible to say who the
> best poker player is, or what the best starting hand is.
>
> A football example is probably the best known common example of an
> intransitive
> relation. It s the well known intransitivity of football match-up results
> that
> makes the BCS rankings preferable to a playoff system. With any kind of
> playoff
> system you can only match up teams two at a time. But, it s entirely possible
> for LSU to have a better team than UT, UT to have a better team than UCLA, but
> UCLA being a better team than LSU. The skill sets that make up a winning team
> are complex and there s no reason for individual match-ups to result in
> meaningful rankings overall.
>
> The same thing happens when you try to rank starting hands in poker. The
> characteristics that give value to a poker hand. Tend to combine in complex
> ways, since the deck is finite the hand you re dealt is not independent of
> other
> players hands and are not independent of cards to come. In my book The
> Complete
> Book of Hold Em Poker I give some characteristics of a starting hand in
> hold em
> which give value to the hand: current poker hand, prospects for improvement,
> position, characteristics of the game, tendency of opponents to make
> mistakes.
> These are not simple, independent characteristics.
>
> In general it s not possible to logically deduce how these characteristics
> that
> give value to a hand combine or interact to give value, it s an empirical
> issue
> dependent on exact combination of levels of the various characteristics of
> inputs.
>
> We can go ahead and use PokerStove right now to look at transitivity in
> attempts
> to rank poker hands.
>
> Let s just look at three hold em hands AsKd, JdTd, and 2h2c.
>
> Which is the best hand of those three? Some people would say the 2h2c because
> it s a pair and it beats both the non pair hands. But, that line of thinking
> presumes that it s possible to have a showdown with 2 card hands. Of course
> that doesn t happen, you never have a showdown until all the cards have been
> dealt, a hold em hand at the showdown is the best 5 cards out of 7. So,
> although it s possible 2h2c is the best hand it s not because a pair beats no
> pair, it s would be because it s the most likely to end up the best hand when
> all the cards are out. That s where PokerStove comes in handy.
>
> Set up your PokerStove with the three hands (be sure and specify the suits).
> If
> you don t have a copy of PokerStove then get one, it s a free download at
> http://www.pokerstove.com/.
>
> You should get AsKd with 36% chance of winning, JdTd 34% and 2h2c with 30%.
> So, of the three, if you think of better in terms of highest chance of ending
> up
> the winner, the best hand is AsKd.
>
> But, let s look at how they stack up if we just compare AsKd and 2h2c, without
> a
> third hand. Rerun PokerStove after removing the JdTd hand and you should get
> AsKd with 47% chance of winning, 2h2c 53%. Now what was the worst hand when
> all
> three are compared simultaneously has become the best hand.
>
> The point of this is that the concept of best when measured by pair-wise
> match-ups does not have the mathematical property known as the transitive
> property. The implication of that is that you can t create a meaningful
> ranking
> of hands.
>
> Next week I'm going to talk some more about transitivity. I'm still on the
> 1st
> chapter and will be for another 2-3 weeks.
>
> I've made some changes to Chapter 1 and posted it on my website for download.
> I'm also putting some links to supplimentary material on the site.
>
> CommentS? Questions?
>
> On Feb 28 2006 6:30 PM, Gary Carson wrote:
>
> >
> > Gary Carson
> > http://www.garycarson.com
> >
> >
> Gary Carson
> http://www.garycarson.com
>
>
>
Gary Carson
http://www.garycarson.com
_______________________________________________________________
The Largest Online Poker Community - http://www.recpoker.com
da pickle
> I think doing such rankings are fun, but you really shouldn't pretend the
> ranks mean something. However the process you go thru do get some rankings
> does
> man something and the article really delves into the process.
>
> So, although I laugh at it, and call it masturbation, I think it's a good
> article.
Sometimes, I think that the tweaks start when "my town" does not get the
right result.
I worked for the Division of Business and Economic Research in graduate
school doing population calculations for the non-Census years for the US
Census Bureau for all the parishes in Louisiana. Lots of school age
population calculations and recalculations. Pushing and shoving and
tweaking. Interesting, but I do not claim to have known what I was helping
with then or now.
Gary Carson
On Mar 3 2006 10:03 AM, da pickle wrote:
> I worked for the Division of Business and Economic Research in graduate
> school doing population calculations for the non-Census years for the US
> Census Bureau for all the parishes in Louisiana.
At LSU? When was that? Who was the director then?
Gary Carson
http://www.garycarson.com
_______________________________________________________________
Your Online Poker Community - http://www.recpoker.com
da pickle
>
> On Mar 3 2006 10:03 AM, da pickle wrote:
>
>> I worked for the Division of Business and Economic Research in graduate
>> school doing population calculations for the non-Census years for the US
>> Census Bureau for all the parishes in Louisiana.
>
> At LSU? When was that? Who was the director then?
Louisiana Tech, they had the contract with the fed in the late 60's. The
director was Dr. Forrest Pollard. I published my first paper with him.
"Trends in Louisiana Civilian Employment, By Industry Division 1960-1967,"
with Dr. Forrest H. Pollard,
The Louisiana Economy, Louisiana Polytechnic Institute, Ruston, Louisiana,
December, 1968.
Gary Carson
Okay. I was working in the Division of Economic Research (or it might have been
Division of Business and Economic Research) at LSU in the early/mid 70's. I
wrote a few articles for Louisiana Business Review back then. Mostly on
industrial structure of the Lousisna economy. I don't even have the citations
of them anymore. We maintained a lot of employment and price data by parish, I
don't think we did any population statistics (although I wrote some stuff with
Roger Burford on population migration in Lousiana, again I don't have the
citations for them and don't even remember where they were published)
I thought you'd said before ou went to Lousisan Tech. I was patrolling around
Tiger Island in 68 (and it wasn't the LSU Tigers).
On Mar 3 2006 11:23 AM, da pickle wrote:
> "Gary Carson" wrote in message
Gary Carson
http://www.garycarson.com
_______________________________________________________________
The Largest Online Poker Community - http://www.recpoker.com
da pickle
> I thought you'd said before ou went to Lousisan Tech. I was patrolling
> around
> Tiger Island in 68 (and it wasn't the LSU Tigers).
BSEE 67
MBA 68
January 69 through January 73, AC130 Gunships.
Dwight Abbott
Good example. I hadn't thought if someone else has AA. If they do, you lose a
lot of value due to split pots.
> >
> > > You can sometimes put opponents on very tight ranges of hands. You're
> > > never
> > > going to put them on exactly 88
> > >
> > > I'll talk about that some (about why pokerstove is a better tool than
> > > twodimes)
> > > when I get to chapter 2.
> > >
> >
> > I started with pokerstove, but it wouldn't let me cut and paste the results
> > from
> > it. I was using the free version, so maybe if I paid for it, I could cut and
> > paste.
> >
>
> Just hilight the results using the cursor and do a cntl-C to copy to the
> memory
> pad, cntl-V to paste it somewhere.
>
That works good. Thanks.
> >
> >
> > > Thanks though. I don't mean to totally disocunt your example. It
> > > demonstrates
> > > clearly why you can't just assume the "best hand" is the best hand.
> > >
> >
> > I actually started out trying to prove that AA would always be the best hand
> > and
> > was getting ready to tell you I was very doubtful it would ever be
> > subordinate
> > to any hand preflop. But then I got the idea of putting all of it's outs in
> > other hands and was surprised to see it worked.
>
> I'm pretty sure it's going to always be favored if at least one ace remains in
> the stub.
>
I think you're right. I might play around a little tho to see if I can find a
way to leave an ace unaccounted for.
Dwight
NoSelfControl
[snipped]
> Let s just look at three hold em hands AsKd, JdTd, and 2h2c.
> You should get AsKd with 36% chance of winning, JdTd 34% and 2h2c with 30%.
> So, of the three, if you think of better in terms of highest chance of ending up
> the winner, the best hand is AsKd.
>
> But, let s look at how they stack up if we just compare AsKd and 2h2c, without a
> third hand. Rerun PokerStove after removing the JdTd hand and you should get
> AsKd with 47% chance of winning, 2h2c 53%. Now what was the worst hand when all
> three are compared simultaneously has become the best hand.
> CommentS? Questions?
As stated, this example does not show lack of transitivity. It shows
lack of a property that is known in social choice and voting theory as
independence of irrelevant alternatives. 22 beats AK heads up, but the
presence of JT acts a spoiler and AK is better.
Transivity only refers to binary relations - eg heads up matches in the
context of poker hands. Viewed this way, the three hands do in fact
provide an example of intransivity heads up.
(i) AsKd beats JdTd (60%-40%)
(ii) JdTd beats 2h2c (54%-46%)
(iii) 2h2c beats AsKd (53%-47%)
Dwight Abbott
Take a look at this. It's very interesting how many ways AA can become the
preflop underdog. Granted most of them extremely unlikely, but now when someone
says it's the "Best hand in the game." My reply would be "Not necessarily..."
Text results appended to pokerstove.txt
278,256 games 0.005 secs 55,651,200 games/sec
Board:
Dead:
equity (%) win (%) tie (%)
Hand 1: 21.9420 % 21.83% 00.11% { AcAd }
Hand 2: 03.7806 % 03.67% 00.11% { AhQc }
Hand 3: 00.9852 % 00.01% 00.97% { KcKd }
Hand 4: 24.1962 % 24.16% 00.04% { 8h8s }
Hand 5: 07.3735 % 07.34% 00.04% { 2c2d }
Hand 6: 09.7347 % 09.70% 00.04% { 3c3d }
Hand 7: 12.8663 % 12.83% 00.04% { 4c4d }
Hand 8: 16.1202 % 16.08% 00.04% { 5c5d }
Hand 9: 03.0013 % 02.03% 00.97% { KhKs }
Dwight
> > >
> > Gary Carson
> > <a href="http://www.garycarson.com" target="_blank">http://www.garycarson.com</a>
Chris in Texas
On Mar 3 2006 9:17 AM, Gary Carson wrote:
>
> I was wasting time this morning on the internet and ran across an article on
> ranking things from the Journal of Statistical Education.
>
> It's only a little technical, probably requires at least a good course in stat
> methods as a prerequisite. I've only skimmed it, but I think it's funny.
> Tons
> and tons of masturbation trying to be technically accurate in performing a
> task
> that's really not doable or meaningful anyway.
>
> It describes the procedures they followed in ranking towns for an article on
> "healthist towns" for Boston Magazine.
>
I probably don't have the stats background to understand the article, but I
agree with you on the many flawed prodecures developed/followed in such
rankings.
Two examples - my town, Laredo, Texas, was recently ranked "happiest city" in
USA by Men's Health I believe. That was based on the fact that we have the
lowest per capita rate of anti-depressant purchases at local pharmacies. What
the rankers ignored is a majority of our population is uninsured, thus they buy
medicine across the river in Mexico, which isn't factored into their rankings.
Another one - many years ago Texas monthly ranked Laredo as the most expensive
Texas city to buy groceries. They took the annual sales as reported by grocery
stores on sales tax reports and divided by the cities estimated population.
Laredo's avg. was TWICE the 2nd place city, and 2 through 10 were fairly tightly
bunched. I recall many locals complaining that the San Antonio based HEB chain
was gouging us on food prices, until someone wrote a letter to the editor of the
local paper explaining that the methodolgy used by the magazine didn't account
for the fact that local stores sell ALOT to shoppers from Mexico.
Just 2 examples that we need to be careful on starting assumptions.
Chris
eleaticus
"Chris in Texas" <4307...@recpoker.com> wrote in message
news:1141412279$748...@recpoker.com...
eleaticus
"Chris in Texas" <4307...@recpoker.com> wrote in message
news:1141412279$748...@recpoker.com...
>
>
>
Gary Carson
On Mar 3 2006 12:48 PM, Dwight Abbott wrote:
>
> equity (%) win (%) tie (%)
> Hand 1: 21.9420 % 21.83% 00.11% { AcAd }
> Hand 2: 03.7806 % 03.67% 00.11% { AhQc }
> Hand 3: 00.9852 % 00.01% 00.97% { KcKd }
> Hand 4: 24.1962 % 24.16% 00.04% { 8h8s }
> Hand 5: 07.3735 % 07.34% 00.04% { 2c2d }
> Hand 6: 09.7347 % 09.70% 00.04% { 3c3d }
> Hand 7: 12.8663 % 12.83% 00.04% { 4c4d }
> Hand 8: 16.1202 % 16.08% 00.04% { 5c5d }
> Hand 9: 03.0013 % 02.03% 00.97% { KhKs }
>
>
Yes, that's a very interesting one. AA isn't the best hand, but still has
enough value for a raise.
There is no way I would have guessed 88 is actually ahead from just looking at
the hands.
Gary Carson
The hills just North of DaNang had a bunch of mortar positions that fired across
the harbor into the base from time to time. We were anchored in the harbor one
night in Spring, 68 when they were shooting a lot and some gunships went in
after them.
It was a hell of a light show. We sat on the fantail drinking coffee and
watching the lights. I did that a lot, sitting around drinking coffee watching
somebody else shooting it up. It was excellent training for a later career as a
corporate management dweeb.
But the mortars in the hills story is funny to me becuase of why were in the
harbor. Some Marines were hitting the beach on the other side of the hills and
we were providing fire support for them. We had three 5 inch guns for that, and
also had a 3 inch gun and two .50 cal. automatic weapons mounted on the bridge
(and some ASW stuff that wouldn't have helped much).
But they didn't have us return fire to the mortars because we couldn't get any
ground spotters into the hills. Gunships didn't need spotters. Shit, they put
out so many tracers they didn't need illuminaiton explosives.
Gary Carson
http://www.garycarson.com
_______________________________________________________________
* New Release: RecPoker.com v2.2 - http://www.recpoker.com
Gary Carson
Okay. It took me a minute, but I think I understand what you're saying. Your
saying that what I did above was compare AK to (JT 22), and then observed that
it beats the two hands but doesn't beat 22.
Yes, I think you're right, that's not really lack of transitivity.
Although the part I had about pairwise comparisons of AK JT 22 not being
transitive, part of what I had isn't really about lack of transitivity.
Good catch, I'll have to rewrite that. I havn't read anything about social
choice theory since I was an undergrad. Maybe I need to catch up on some of
that stuff.
That independence of irrelavant alternatives comes up a lot in poker and I need
to read up on that. I was planning a later chapter to get into hand
comparisions in more detail later and cover that topic, but I didn't know the
name for the phenomana. I knew it came up in voting theory but hadn't really
thought about it.
Gary Carson
http://www.garycarson.com
_______________________________________________________________
eleaticus
Now, you think the process' output ridiculous, which the article certainly
makes clear is absolutely true.
But, you are supposedly knowledgeable about scales.
How far down the article did you have to go to note the first self-reported
crime against measurement and mathematical logic?
eleaticus
ee-lee-AT-i-cus
"Gary Carson" <garyc...@alumni.northwestern.edu> wrote in message
news:1141399054$748...@recpoker.com...
Gary Carson
On Mar 3 2006 5:35 PM, eleaticus wrote:
> Interesting and nauseating reference, Gary.
>
> Now, you think the process' output ridiculous, which the article certainly
> makes clear is absolutely true.
>
> But, you are supposedly knowledgeable about scales.
What I'm knowledgable about is what I mean when I use a particular word.
As you previously pointed out, not everybody means exactly the same thing as I
do when they use the word scale.
Do you have a point that's any deeper than that?
>
> How far down the article did you have to go to note the first self-reported
> crime against measurement and mathematical logic?
I have no clue what you're talking about.
If you'd actually read the post I made you might note that I mentioned I havn't
read it yet.
Are you suggesting that he makes so many mistakes that I shouldn't bother
reading it?
Give it up. I agree. You're smarter than anybody and know more words too.
John Forsberg
On Mar 3 2006 4:17 PM, Gary Carson wrote:
> So, although I laugh at it, and call it masturbation, I think it's a good
> article.
What's not good about masturbation?
eleaticus
"Gary Carson" <garyc...@alumni.northwestern.edu> wrote in message
news:1141431217$748...@recpoker.com...
>
>
>
> On Mar 3 2006 5:35 PM, eleaticus wrote:
>
> > Interesting and nauseating reference, Gary.
> >
> > Now, you think the process' output ridiculous, which the article
certainly
> > makes clear is absolutely true.
> >
> > But, you are supposedly knowledgeable about scales.
>
> What I'm knowledgable about is what I mean when I use a particular word.
tsk tsk, I guess my wording was not pleasant for you?
> As you previously pointed out, not everybody means exactly the same thing
as I
> do when they use the word scale.
>
> Do you have a point that's any deeper than that?
Sure, but I thought you might actually have some insight.
Was that really so dumb of me?
> > How far down the article did you have to go to note the first
self-reported
> > crime against measurement and mathematical logic?
>
> I have no clue what you're talking about.
>
> If you'd actually read the post I made you might note that I mentioned I
havn't
> read it yet.
Gary:
"It's only a little technical, probably requires at least a good course in
stat
methods as a prerequisite. I've only skimmed it, but I think it's funny.
Tons
and tons of masturbation trying to be technically accurate in performing a
task
that's really not doable or meaningful anyway."
And, what in hell was your basis for saying:
"So, although I laugh at it, and call it masturbation, I think it's a good
article."
Good article that you didn't read?
rofflmfao!
> Are you suggesting that he makes so many mistakes that I shouldn't bother
> reading it?
It is not just 'his' mistakes but the whole 'masturbation' thing, which
reminds me of kids playing mud pies in a pig sty.
> Give it up. I agree. You're smarter than anybody and know more words too.
Well, I guess I should apologize for using the word 'scale'!
eleaticus
ee-lee-AT-i-cus