Bots on Paradise?
William Bradley
(Para-bot-ic) asked that someone crosspost it here. I have no idea if it is
legitimate, but I thought it may be of interest to many of you. As an aside,
for those of you interested in reading more about the University of
Alberta's work on poker bots, here is their website:
http://www.cs.ualberta.ca/~games/poker
---------------------------------------------------------------
I bot
There have been a lot of posts regarding various "conspiracy" issues at
Paradise and other poker sites. While I can not comment on all the alleged
collusion, fixed deals, shills, I can comment definitely on the use of bots
at Paradise.
I have written and have been using a bot on Paradise. My bot first made its
appearance on the play money tables approximately 8 months ago and has been
in production on the real money tables for approximately 3 months.
Let me state right up front that I am in no way affiliated with Paradise and
have deliberately not included information within this post to ensure that
my identity remains anonymous. I am convinced that Paradise will shut me
down if they had any idea who I am.
I also fully expect that this post, with some of the technical details,
might cause Paradise to change its software to slow me down. I look forward
to any changes or curve balls they throw, as it has been the
Academic/Programming challenge that spawned the bot in the first place.
Many of you may also wonder; why I would write this post and possibly kill
the goose that is laying the golden eggs. I do so for the following reasons:
- While my bot is showing a profit far better than minimum wage, if I were
to put an hourly wage on my time creating it I have probably earned pennies
an hour.
- I originally started this project as an exercise in poker playing
research, there have been many discussions on twoplustwo and R.G.P. about
how there can never be a software poker program that will approach that of a
great poker player. While I do not claim that my bot approaches that of the
great players, it is a winning player. If I can do it in my spare time (3
years) then I am convinced that if the same efforts were put towards poker
as deep blue (the chess program) then poker software would beat the big
guns.
- I want to encourage others to develop bots, I would love the day where
bots can openly compete at the poker houses.
- I want it to be known that when on Paradise, or any other on-line house
that the players might be competing against a bot.
- It's not quite a golden egg. :)
My greatest wish is for Paradise to encourage and allow publicly created
bots with the following conditions: - Create a bot API so that code,
resources and programming is not wasted interfacing the client software.
- Insist all bots have the name bot in their name.
- Create bot free tables so people who do not want to play against bots don'
t have to.
My bot's results
Paradise free tables: 2.8 Big Bets per hour
Paradise .50/$1 tables: 2.1 Big Bets per hour
Paradise $1/$2 tables: 1.3 Big Bets per hour
Paradise $2/$4 tables: 0.9 Big Bets per hour
Paradise $3/$6 tables: 1.1 Big bets per hour
Paradise $5/$10 tables: 0.4 Big bets per hour
It works hard for its money
My bot is strictly a hold'em player. I have only played it on full tables up
to $5/$10, I won't go near 1 on 1, tournament, or 5 handed.
This bot is far from automated, it needs my help to select, enter a game and
sit down, but from that point on it will play. Upon starting up I enter its
seat number, how long it will play, and/or a win/loss stop gap, the bot will
play until the predefined time has expired or the win/loss stop achieved, or
the money it brings to the table reaches zero. (I hate it when that
happens).
The bot determines the cards dealt to it by sampling the pixel values (more
on this later) and then moving the mouse over the appropriate check/call
bet/raise or fold buttons and clicking. It follows the action by reading the
action in the chat box. The chat box is set to "Dealer: Everything".
A typical game looks as follows:
»Dealer: Starting a new hand...
»Dealer: ** DEALING POCKETS **
»Dealer: Player_1 posts the small blind
»Dealer: Player_2 posts the big blind
»Dealer: Player_3 folds
»Dealer: Player_4 calls
»Dealer: Player_5 calls
»Dealer: Player_6 folds
»Dealer: Player_7 calls
»Dealer: Player_8 folds
»Dealer: Player_9 folds
»Dealer: Player_1 folds
»Dealer: Player_2 checks
»Dealer: ** DEALING FLOP **
»Dealer: Player_2 checks
»Dealer: Player_4 bets
»Dealer: Player_4 bets
»Dealer: Player_5 folds
»Dealer: Player_7 calls
»Dealer: Player_2 folds
»Dealer: ** DEALING TURN **
»Dealer: Player_4 bets
»Dealer: Player_7 raises
»Dealer: Player_4 calls
»Dealer: ** DEALING RIVER **
»Dealer: Player_4 checks
»Dealer: Player_7 checks
»Dealer: Player_4 shows a pair of xxxxx
»Dealer: Player_7 shows two pair, xxxxx and xxxxx
»#xxxxxxxx: Player_7 wins ($xxxxx) with two pair, xxxxx and xxxxx
The bot capture the text as it appears in the chat box, tracks the actions
then acts by moving the mouse. As soon as is sees the "Dealer: xxxxxxx
shows" prompt it samples the players shown cards and updates the player's
statistics in the Player access database.
Humorous Anecdotes
Watching the bot play is a somewhat eerie feeling. Sometimes I find myself
talking to it, yelling, smiling or nodding in approval. While testing and
playing of the following have happened.
A maniac player had been raising, my bot had pocket KKs and was quite
content re-raising. A third player was caught in the middle. A king fell on
the turn and my bot continued re-raising. The third player folded. The
maniac folded on the river, at which point the third player accused the bot
and maniac of colluding.
Another time my bot was accused of sucking out, it had middle pair and a
runner-runner straight and flush draw. The Ace fell on the river giving the
opponent trips but the bot a flush. At this point the player started
berating the bot, calling it names and giving it lessons.
During one of the promotions as the winning hand started approaching, many
on the table started folding to the blinds. My bot doesn't understand
folding because of special hand numbers and continued to call/raise. The
table captains were furious and started yelling.
The bot has received a lot of "nh sir", idle conversation or where in
XXXXXXXXXX do I live or work. I just smile.
Every bot has a history.
I originally started this project, not with the intentions of creating a
bot, but to create a poker-playing program and to do research on various
starting hands. Here is basically how the bot started from its inception to
present day.
The software originally started out (about 3 years ago) as a starting hand
analyzer basically tracking how each had did headup, 3 player, 4 player, etc
playing each hand millions of times and tracking its results.
The next generation played out millions of hands this time
folding/calling/raising and tracking the EV for each hand. The database was
then enhanced to track each of the starting hands EV in various positions
and tracking the action of each player who has acted.
I then added a player type database, this player database rated opponents
based on percentages of flops seen, aggression, starting hand and actions
associated with each starting hand. The program then tracked each opponent
and not only acted on its EV tables but adjusted its actions based on what
it had learned about its opponents. I.E. If the program has a marginal to
good hand but the player in front of it is somewhat loose it might raise or
call. However in the identical situation if the player in front is somewhat
tight it would fold.
At this point I started playing various versions of the program against a
table of various versions. Tweaking various values, and kept progressing the
version that consistently outperformed other versions.
Determining cards
The bot learns about cards by sampling pixels for color values, the first
thing the bot needs to do is learn where its cards are located. When
entering a game I provide the bot interface the seat number the cards,
Paradise's client is 800 by 600 and the cards will be displayed in the
following locations.
Relative to the 800 by 600 screen (including the window borders) the upper
left corner of the cards is located at:
Seat 1: Card 1 : x501-y22: Card 2: x517-y24
Seat 2: Card 1: x699-y90: Card 2: x715-y92
Seat 3: Card 1: x710-y202: Card 2: x726-y204
Seat 4: Card 1: x600-y339: Card 2 : x616-y341
Seat 5: Card 1: x452-y354: Card 2 : x468-y356
Etc.
The next step is to determine suits. My Paradise client is configured as a
four-color deck. Relative to the top, left corner or the card I sample the
Red Green Blue (RGB) value of the pixel located at x8-y22 (8 pixels to the
right and 22 pixels down) from the upper left corner.
The values will be 1 of the following:
black - spade - r0 g0 b0
red - heart - r240 g32 b0
green - clubs - r32 g127 b0
blue - diamonds - r15 g32 b 239
Now the bot knows the suits of the two cards dealt.
Determine the value of each card is trickier. Basically I sample from one to
three points on upper left portion of the cards that will have a non-white
value unique to that card.
For example all two's and only two's will have the pixels (relative) to the
upper left corner of the card set will have the pixels in the following
three locations set on: x5-y3 x14-y3 x14-y11, So I sample those three points
if all three are non-white then I know it's a two, if any of them are white
I check for unique pixels that identify a three. And so on...
The Queens are the easiest to detect, the bottom decender of the loop is
unique so a queen can be determined by sampling just one point at location
x12-y18.
Determine the flop, turn and river are done the same way, the five cards are
kept at locations: card 1 - x266 - y150 card 2 - x320 - y150 card 3 - x374 -
y150 card 4 - x428 - y150 card 5 - x482 - y150
As each player shows his/her hand I then retrieve their cards in the above
table and store them in the player database/.
Interfacing with the Paradise Client
The toughest part by far was interfacing with the Paradise client. Those of
you who are familiar with hooking DOS interupts will be familiar with the
technique used. Though much more complex under WIN32 applications the
Windows API can be hooked. To do so I used the APIHOOKS package.
APIHOOKS allows you to hook any of the WIN32 events generated by a process
and redirect it one of your DLL's. Basically you configure what windows
events you want to interupt, when the client generates events your code
executes and you can act upon the even, pass data to your application then
allow the Windows API to continue processing.
Many of the graphics, mouse events have to be hooked and then steered
towards your DLL.
Paradise Bot FAQ
Q. Can I get a copy of the bot?
A. No this bot is not for sale. Maybe one day if Paradise or another site
endorces bots I will sell the engine. But don't hold your breath.
Q. What other sites will it work on?
A. Currently only Paradise, I might port it to other sites but no plans yet.
I understand Party poker has a similar look. Hmmmmmmmmmm.
Q. Whose better you or the bot?
A. I like to think I am however, the bot does better than I do. I tend to
think that the bot doesn't tilt.
Q. What are some flaws in the bot.
A. The bot doesn't check-raise often enough, raise to isolate enough and
some other flaws, but I am continously tweaking the database and code.
Q. Is the bots play predictable.
A. No, the bot uses randomness as well, even though it is driven by EV,
there is no play where it will always do the same play, this is especially
noticible where the EV between calling/folding or foldin/raising is close.
Q. Does the bot play similar to S&M, Jones or Abdul?
A. None however I noticed that at the free tables the starting hands play
closer to Jones, but at the money tables the bot tigtens up, not quite as
tight as S&M though.
Q. Do you think its wrong to play a bot?
A. My thoughts about bots are similar to Black Jack counters. It is not
illegal, only against the rules of the house therefore I take measures to
not get caught. I do believe that the population should know that there is
the potential of playing against bots, thus this post.
Q. How do you hide the fact it's a bot?
A. The bot uses the same account I do. It plays for random periods and not
excessive.
Q. Do you play the bot unattended.
A. Yes, I enter it in a game and do something else all the time.
Q. Whats the longest session it has played
A. About 4 hours.
Q. Will it respond to text chat?
A. No, I am not interested in devoting resources to having it attempt to be
human, plus that would probably give it away. I am satisfied that its
quietness is the same as a large percent of the Paradise population. Q. When
will it play larger, one on one or tourneys.
A. One on One and tourneys are not a priority, however I still need some
more 5-10 experience and tweaking, but it is in the near future.
Q. What are some of its cool features?
A. It will automatically request hand histories every 75-90 hands. Uses hand
histories to update player profiles.
Q. How can I write a bot?
A. Learn to program. Spent thousands of hours programming. Learn to play
poker. Spent thousands of hours playing poker. Read about U of A's pokibot,
grab the source available, examine it, understand it. Read
rec.programming.games
Q. Can you characterise its play?
A. It plays tight, when it has the cards it gets aggressive. It doesn't get
fooled by deception like humans normally are. After an opponent has played
100 or more hands against it the player's profile has a pretty good record
of flops seens, position play, as the player changes mode the database
reflects it.
Q. Can it be beaten?
A. Yes it can and has been, however it tracks how well players have done
against while in the same hand and learns to respect when those players gets
aggressive.
Insair
computer speak?
DrToast
>A fascinating and quite believable post. Or am I just easily taken in by
>computer speak?
No. If this is a hoax it is a damn good one.
DrToast
"Looks like I picked the wrong week to quit sniffing glue."
(Get rid of "THEWEAKESTLINK" to reply)
Tim Dow
>A fascinating and quite believable post. Or am I just easily taken in by
>computer speak?
I am not easily taken in by computer speak, (poker speak is another story
though). Some of the Para-bot-ic post does seem believable but I find a
couple of things about the post quite strange. For instance:
>I also fully expect that this post, with some of the technical details,
>might cause Paradise to change its software to slow me down. I look forward
>to any changes or curve balls they throw, as it has been the
>Academic/Programming challenge that spawned the bot in the first place.
This paragraph makes the whole thing much less believable to me. If Paradise
wants it can implement trivial changes it its client software that will not
only "slow down" this type of bot, but render it useless for at least the
next 50 years or so.
Secondly what technical details? Pixel positions that quite literally anyone
could determine even with zero programming knowledge? Recognizing the cards
using this type of screen scrape approach is a project for a junior high
school student.
>The toughest part by far was interfacing with the Paradise client. Those of
>you who are familiar with hooking DOS interupts will be familiar with the
>technique used. Though much more complex under WIN32 applications the
>Windows API can be hooked. To do so I used the APIHOOKS package.
In my opinion the above is a ludicrous statement. For an experience Windows
programmer there us nothing at all complex about using hooks. The ability to
play poker is what is complicated about writing this bot. Everything else
could be written in a matter of a week or two by most intermediate level
Windows programmers.
Even assuming one started with the University of Alberta's Lokibot source
the improvements so casually claimed, especially in the area of
unpredictability, are, if true, a significant accomplishment and one that
would require an expert in artificial intelligence programming.
The results claimed are as follows:
>Paradise free tables: 2.8 Big Bets per hour
>Paradise .50/$1 tables: 2.1 Big Bets per hour
>Paradise $1/$2 tables: 1.3 Big Bets per hour
>Paradise $2/$4 tables: 0.9 Big Bets per hour
>Paradise $3/$6 tables: 1.1 Big bets per hour
>Paradise $5/$10 tables: 0.4 Big bets per hour
I am certainly not any kind of a Poker expert much less online poker, (which
I have never played) so I have no idea how good the above results really
are. Quite possibly they are in line with the current University of Alberta'
s Lokibot performance.
I think the main thing to remember is that the next steps in computer poker
playing will be exponentially more difficult. Some of the artificial
intelligence problems faced dwarf the problems of a chess playing program.
In my opinion even Go playing programs which have had much more limited
success against the master players after many years of programming effort,
including millions of dollars in rewards offered, may be an easier task.
Jeffrey L. Woods
says...
> A fascinating and quite believable post. Or am I just easily taken in by
> computer speak?
I believe it. It is indeed completely plausible, and not unlikely that
someone with time to burn and a fascination "to see if it could be done"
would do so.
Jeffrey L. Woods
wgbr...@home.com says...
> Q. What are some of its cool features?
> A. It will automatically request hand histories every 75-90 hands. Uses hand
> histories to update player profiles.
This will almost certainly narrow down to a very small number of playesr
who the author is -- at least to Paradise, who has access to their
database of who requests ALL hand histories that they play in. Few, if
any, other players will have such a strong "hit rate" as to always
grabbing histories.
Steve Irwin
"Jeffrey L. Woods" <jlw-n...@telix.com> wrote in message
news:MPG.15e5d3623...@news-server.nc.rr.com...
The bot-programmer mentioned that the bot played on his account. He also
mentioned that the bot has better win rates than he does, implying that the
programmer also plays online, on the same account. That might inadvertently
serve to hide the bot from Paradise, if the programmer is not as good about
requesting hand histories.
Steve Irwin
Richard Kuhler
news:S9Ne7.71144$B37.1...@news1.rdc1.bc.home.com...
<snip>
> Let me state right up front that I am in no way affiliated with Paradise
and
> have deliberately not included information within this post to ensure that
> my identity remains anonymous. I am convinced that Paradise will shut me
> down if they had any idea who I am.
I don't know why Paradise would care. His bot keeps games going,
contributes to the rake and has no secret knowledge that would give it an
advantage. Why would they want to shut it down?
Richard
mhu...@mcloud.slktaho1.nv.home.com
> <snip>
Perception. If I was Paradise I'd want to keep my customers from having
the perception that there were bots online (thus the reason not to make
it easy). OTOH I can't believe that somebody with enough interest to
write a winning low limit bot wouldn't have the discipline to gather
statistics on their own and others non-strategic and non-tactical
actions (betting speed, frequency of hand history requests, size of
hand history requests, lengths of sessions, etc.) not to be able to model
those also. The only reason not to do this was if the programmer was
truely only interested in the challenge of writting a winning bot.
FWIW I found the description of writting the bot as believeable but I
also know that probably 10 thousand people in the US (or more) could
write such a description without having ever gone through the experience.
Its just too obvious. Thus the realism of the description shouldn't be
taken as proof that the author actually has a running bot.
mph
yoyojemojo
Why can't we all just get along?
a war is coming..................
yoyo
Richard Kuhler <pocket...@winning.com> wrote in message
news:1rVe7.239$uh1.1...@typhoon.san.rr.com...
Lee Daniel Crocker
> though). Some of the Para-bot-ic post does seem believable but I find a
> couple of things about the post quite strange. For instance:
>
> >I also fully expect that this post, with some of the technical details,
> >might cause Paradise to change its software to slow me down. I look forward
> >to any changes or curve balls they throw, as it has been the
> >Academic/Programming challenge that spawned the bot in the first place.
>
> This paragraph makes the whole thing much less believable to me. If Paradise
> wants it can implement trivial changes it its client software that will not
> only "slow down" this type of bot, but render it useless for at least the
> next 50 years or so.
Poppycock. If you honestly think that, then you know nothing about AI and
networking. Yes, there are a lot of curves they could throw, and very many
of them might put the 'bot author beyond his capability to keep up, but it
could never prevent bots entirely. At some point, it has to send bits over
an anonymous wire that put bits on the screen that a human can identify, and
there's simply no way to do that without allowing bots to identify them as
well. Humans are much better at pattern-matching than bots, but recognizing
cards and buttons and text is not exactly a challenge.
And even if I'm wrong about that, that in no way makes the post unbelievable,
because its quite clear that the author believes it and has every reason to.
> Secondly what technical details? Pixel positions that quite literally anyone
> could determine even with zero programming knowledge? Recognizing the cards
> using this type of screen scrape approach is a project for a junior high
> school student.
>
> >The toughest part by far was interfacing with the Paradise client. Those of
> >you who are familiar with hooking DOS interupts will be familiar with the
> >technique used. Though much more complex under WIN32 applications the
> >Windows API can be hooked. To do so I used the APIHOOKS package.
>
> In my opinion the above is a ludicrous statement. For an experience Windows
> programmer there us nothing at all complex about using hooks. The ability to
> play poker is what is complicated about writing this bot. Everything else
> could be written in a matter of a week or two by most intermediate level
> Windows programmers.
When did he ever say he was an experienced programmer? Little details like
this are exactly what makes the post believable. The fact that he had difficulty
with something that many Windows programmers don't, and that he took pride in
something relatively simple, is entirely consistent with someone who has been
programming only a short time, and entirely believable. There's no reason at
all to think an experienced programmer would be required to do this. It wouldn't
surprize me a bit if this kid is 19.
> I think the main thing to remember is that the next steps in computer poker
> playing will be exponentially more difficult. Some of the artificial
> intelligence problems faced dwarf the problems of a chess playing program.
> In my opinion even Go playing programs which have had much more limited
> success against the master players after many years of programming effort,
> including millions of dollars in rewards offered, may be an easier task.
That may be true, but anyone who doubts that computers will eventually play
Go--and Poker--better than any human doesn't understand the issue. Chess
only took a few decades; since then, computers have gotten bigger and faster,
and software has gotten better and better. Go might take only another decade
or two, and Poker probably not much longer.
--
Lee Daniel Crocker <l...@piclab.com>
<http://www.piclab.com/lee/>
David Sklansky
A Probable Lie
Posted By: David Sklansky <Dskl...@aol.com>
Date: Thursday, 16 August 2001, at 5:51 p.m.
In Response To: Re: Bot and paid for (goodbeet)
There are several reasons, some which I cannot divulge. Ones I can are that
there is not enough difference between play money and .50-1.00 results.
Another reason is that while it is simple to use game theory headup,
multiway cannot be programmed by anybody who is not an expert himself. But
again the biggest reason I seriously doubt this poster is related to
confidential information given to me which is highly reliable.
PS could someone please repost this on RGP
AL - Dork to some
>From: "David Sklansky" fran...@netzero.net
> the biggest reason I seriously doubt this poster is related to
>confidential information given to me which is highly reliable
why even post if you're gonna write this cryptic crap?
Tim Dow
>> This paragraph makes the whole thing much less believable to me. If
Paradise
>> wants it can implement trivial changes it its client software that will
not
>> only "slow down" this type of bot, but render it useless for at least the
>> next 50 years or so.
>
>Poppycock. If you honestly think that, then you know nothing about AI and
>networking. Yes, there are a lot of curves they could throw, and very many
>of them might put the 'bot author beyond his capability to keep up, but it
>could never prevent bots entirely. At some point, it has to send bits over
>an anonymous wire that put bits on the screen that a human can identify,
and
>there's simply no way to do that without allowing bots to identify them as
>well. Humans are much better at pattern-matching than bots, but
recognizing
>cards and buttons and text is not exactly a challenge.
>
>And even if I'm wrong about that, that in no way makes the post
unbelievable,
>because its quite clear that the author believes it and has every reason
to.
Actually I posted a response on this issue because I have 25 years of
experience in the field of artificial intelligence. If it were about Poker
strategy I would have kept quite and tried to learn something. Maybe not bad
advice for you regarding AI? As far as networking is concerned it has no
relevance to this discussion.
Its not clear to me how someone could be sure enough of his position to
refer to someone else's as "Poppycock" and then in the next paragraph admit
that he might be wrong, but I will humor you and give you an example showing
that you are, in fact, quite wrong.
If Paradise wanted to stop bots from playing cards (which I doubt they do) I
can think of at least 3 completely different approach's that, as I said,
would eliminate them for at least 50 years or so. The simplest but also the
most intrusive would be to ask a question in a pop up window at random
times. It seems to me if people wanted to be assured of playing at a bot
free table this small inconvenience might be acceptable. Especially as it
could be asked infrequently, only before the deal. Only once in every 12
hours or so of play, because the first wrong answer (subject to follow-up
verification) would be cause for cancellation of the account. Another
advantage of this method is that it could be programmed in a matter of
hours.
Because the bot would not know what question to expect next, or even what
type of question to expect, even a question such as Which hand is higher? A.
Trips B. Flush C. Pair would be completely impossible for the bot to answer.
The possibilities of course are endless. What is 2 - 2, then maybe a
completely different question; What is two minus two? A multiple choice
question showing, for instance an elephant in a small popup window asking
What is this animal? A. Tiger B. Elephant C. Poppycock would be even more
effective. I believe my 50 year estimate is conservative for developing a
bot that can answer these type of random questions. This method is only one
possibility there are other less intrusive but more complicated methods.
>When did he ever say he was an experienced programmer? Little details like
>this are exactly what makes the post believable. The fact that he had
difficulty
>with something that many Windows programmers don't, and that he took pride
in
>something relatively simple, is entirely consistent with someone who has
been
>programming only a short time, and entirely believable. There's no reason
at
>all to think an experienced programmer would be required to do this. It
wouldn't
>surprize me a bit if this kid is 19.
In fact I was not trying to argue that the his post was not believable but
instead to point out some things that I personally thought were
inconsistent. Actually I think there is certainly a possibility that the bot
exists as he says. As I said, what I don't have a clue about is how good the
poker playing at Paradise $5/$10 tables actually is. If that was not clear
than I apologize. As far as him being an experienced programmer I thought he
implied quite clearly that he had been programming for thousands of hours
when he wrote:
>>Q. How can I write a bot?
>>A. Learn to program. Spent thousands of hours programming. Learn to play
>>poker. Spent thousands of hours playing poker. Read about U of A's
pokibot,
>>grab the source available, examine it, understand it. Read
>>rec.programming.games
Again, my point had nothing to do with his programming skills or lack
thereof but was simply to point out that writing the part of the bot that he
stated was the most complex is actually trivial, but writing a poker
program that can beat even an average casino player consistently over the
long haul is something so complex that it has not been accomplished yet.
>That may be true, but anyone who doubts that computers will eventually play
>Go--and Poker--better than any human doesn't understand the issue. Chess
>only took a few decades; since then, computers have gotten bigger and
faster,
>and software has gotten better and better. Go might take only another
decade
>or two, and Poker probably not much longer.
As I said there are very few people who understand the issue better than I
do. I'm surprised chess took as long as it did, it is a VERY different
kettle of fish from Poker. I don't think I mentioned a time frame in which I
though a computer will beat the number one Go or Poker player, two decades
is a long way away.
Since you brought it up I will say that I don't think Computer Go will be
much beyond the 5 Dan professional level in another decade, and will not
come close to a 9 Dan professional in another 2 decades. I believe this even
though I also expect we will have computers running at over 100 trillion
instructions per second. Brute force alone will not solve these problems.
That said I certainly don't disagree that it will "eventually happen", not
did I say I did. It is worth noting that there has been about 2.5 million
dollars in prizes outstanding for at least the last decade for the first 5
Dan program so there will certainly be some incentive to keep trying.
I think Poker is more complicated in some ways than Go but haven't studied
all the issues as thoroughly. Suffice it to say I don't believe a computer
will consistently win the World Series of Poker in our lifetime.
Dsklansky
Excuse me? Do you think I would lie about this? Wouldn't you rather have me
tell you something I am virtually sure of, due to an impeccable source, than to
say nothing at all?
By the way, all of this stuff should be irrelevant to good players. Nine bots
that could each make three bucks an hour in a 3-6, would lose sixty bucks an
hour to me playing 20-40. I am quite positive that no one is near programming a
computer to beat even a fairly tough 10-20 ring game, including U. of Alberta,
Mike Caro, Wayne Russell, or Bob Wilson.
AL - Dork to some
>
>By the way, all of this stuff should be irrelevant to good players
This is true... I'd sure rather play a bot, who has some sense of order to its
play, than some of the erratic low level players.
Linda K Sherman
> ...Read rec.programming.games...
Well, there ain't no such newsgroup. And there ain't no
rec.games.programming neither. There *is* a rec.games.computer that has
several subgroups but a real programmer who reads these groups would not
make such a boneheaded error with the name. Also, if he can read this
mythical rec.programming.games group, he could certainly post his own
message here and not have to ask someone to do it for him.
I have other reasons to believe this is complete bullshit, but it's too
late at night for me to stay up any longer writing in depth about this
issue. I will say that the original message has all the characteristics
of a textbook Usenet troll, except that it was posted on a website.
I'm not saying developing such a bot can't be done, but I'm just about
100% sure that this guy didn't do it.
Linda
--
Linda K Sherman
To email me, replace ".dimsbam." with '@' in this address:
linsherman.dimsbam.worldnet.att.net
Mike Caro
What???
David, it isn't even hard. It would be a scary proposition, though,
because in a demonstration it would be easy for the players to join
together to beat up on the bot. Give me an honest game and six months
and I'm betting my computer player could soundly beat that "fairly
tough" 10-20 game -- especially if it's seven-card stud.
I agree that superior players could usually earn more money from the
same field, though -- but that's only because they can use psychology
and read tells. (Some day, maybe the technology will be there so we
can program that, too. Right now I can only do it in very limited
ways.)
A purely mechanical player who is oblivious to psychology and tells
probably won't fare anywhere near as well as a properly programmed
"artificially intelligent" player -- who has the advantage of always
"remembering" to do everything right, to the extent it's been
instructed. Most winning players simply don't remember to do
everything right in the short time allotted for them to make decisions
-- even those who think they do.
It's simply MUCH harder to program heads-up, no-limit (which is what
my Orac played in 1984) than a ring game where the bet sizes are
pre-established and most of the wars take place between the top tier
of hands that are often easier to evaluate.
Straight Flushes,
Mike Caro
Dark Giant
lot of programming hassle to capture the cards. Its my understanding that
each card is represented by an image file(gif or jpg??..not sure..never
bothered to look) in the main folder. Why not just replace all 52 image
files in the folder with alternate files that would be more bot friendly
that send suit and rank information directly in one step? Just an idea from
a 1990 era Commodore 64 basic programmer who hasn't compiled a program in 10
years....lol.
As a side note, I have often wondered if Paradise and most other sites
don't list the cards dealt themselves in the dealer chat to make it a bit
more difficult for bot programmers. I have been frustrated by this myself at
times. I will be out of a hand and working in another screen and come back
just as the next hand is starting. I will see who won and occassionally out
of curiosity will want to see how the hand played out. I'll hit the dealer
chat to reconstruct the action of the just completed hand to catch up on
what I missed. Its not at all very useful when the flop, turn and river
card dealt are not listed in the dealer chat. Of course I could request a
hand history but that takes a few minutes to arrive and my curiosity will
have passed by that point.
Dark Giant
"William Bradley" <wgbr...@home.com> wrote in message
news:S9Ne7.71144$B37.1...@news1.rdc1.bc.home.com...
Badger
> There is at least one solid reason to suspect a hoax here. He wrote:
>
> > ...Read rec.programming.games...
>
> Well, there ain't no such newsgroup. And there ain't no
> rec.games.programming neither. There *is* a rec.games.computer
There is also an alt.programming.games
> that has
> several subgroups but a real programmer who reads these groups would not
> make such a boneheaded error with the name. Also, if he can read this
> mythical rec.programming.games group, he could certainly post his own
> message here and not have to ask someone to do it for him.
It seemed pretty clear the person didn't know Usenet from a fishing net.
Chris Vinall
"Linda K Sherman" <poke...@worldnet.att.net> wrote in message
news:3B7CC321...@worldnet.att.net...
> There is at least one solid reason to suspect a hoax here. He wrote:
>
> > ...Read rec.programming.games...
>
> Well, there ain't no such newsgroup. And there ain't no
> rec.games.programming neither. There *is* a rec.games.computer that has
> several subgroups but a real programmer who reads these groups would not
> make such a boneheaded error with the name. Also, if he can read this
> mythical rec.programming.games group, he could certainly post his own
> message here and not have to ask someone to do it for him.
Also I checked out alt.games.programming, which actually exists, and it's
devoted, unsurprisingly, to computer game programming as opposed to
programming players of real-life games.
Chris
Dsklansky
>because in a demonstration it would be easy for the players to join
>together to beat up on the bot. Give me an honest game and six months
>and I'm betting my computer player could soundly beat that "fairly
>tough" 10-20 game -- especially if it's seven-card stud.
My reason for doubting this is not because I don't think it is theoretically
possible but rather because of the number of trees involved. It is much more
than in chess. For one thing how a position was gotten to matters a lot in
poker but not at all in chess. In others words what transpired up to that point
is essential to know to play wel,l but only in poker. Another problem in multi
way games is that the computer must not only read hands but also read an
opponents thoughts about not only the computer's hand but third party hands as
well. Game Theory fixes this problem only in two handed games.
In spite of the above I would not be surprised if the team that programmed Deep
Blue could make a winning 10-20 player using their supercomputers and hundreds
of years of man hours. But I think it would take them at least five years . And
you more than six.
aaron
loved your letter in the review-journal concerning UNLV.
aaron
dave keiser
> A fascinating and quite believable post. Or am I just easily taken in by
> computer speak?
I don't know but they dropped a cookie on me when i read their
post. Cheak my internet clean-up before i exit my browser. Guess their
doing research but i don't remember someone dropping a cookie on me in
a rgc post. Can't eat cookies high blood-lipids.
Pokerbase
detection schemes. Just pay some kid (or do it yourself) $15 bucks an
hour to monitor your bots and answer these infrequent pop up questions
as they come. If you've got 30 bots grinding out $4 an hour in 3-6
games, you will still be making clost to a $100 an hour after
expenses.
If your bots were programmed to subtlety collude with each other when
they found themselves in the same game, I think that advantage would
make it relatively easy to program a bot that could beat semi-tough
games (with 2-3 bots in each game). Remember, with a bot you don't
need to make 2bb an hour in order to make a lot of money. In fact,
you would probably would want to win at a much lower rate in order to
stay under the radar.
I doubt that there is anyone actually doing this, except perhaps for
the intellectual challenge. Frankly, a person with the skills to
build a system like this could probably make more money in a lot of
other ways.
"Tim Dow" <te...@example.net> wrote in message news:<C_0f7.2971$j_2....@newscontent-01.sprint.ca>...
Janne Raevaara
Badger <PlayWinning[RemoveThis]Po...@earthlink.net> wrote in message
news:oq4f7.16930$ZM2.1...@newsread2.prod.itd.earthlink.net...
> "Linda K Sherman" <poke...@worldnet.att.net> wrote...
> > There is at least one solid reason to suspect a hoax here. He wrote:
> >
> > > ...Read rec.programming.games...
> >
> > Well, there ain't no such newsgroup. And there ain't no
> > rec.games.programming neither. There *is* a rec.games.computer
>
> There is also an alt.programming.games
Actually it's alt.games.programming, ant there also seems to be
rec.games.programmer, but I don't subscribe to it, so I don't know if it's
really about games programming...
Janne Raevaara
Dark Giant <raym...@yahoo.com> wrote in message
news:y04f7.111546$EP6.29...@news1.rdc2.pa.home.com...
> A very interesting post. It seems to me that the programmer goes through
a
> lot of programming hassle to capture the cards. Its my understanding that
> each card is represented by an image file(gif or jpg??..not sure..never
> bothered to look) in the main folder. Why not just replace all 52 image
> files in the folder with alternate files that would be more bot friendly
> that send suit and rank information directly in one step? Just an idea
from
> a 1990 era Commodore 64 basic programmer who hasn't compiled a program in
10
> years....lol.
I have given some thought to coding a paradise bot, and that's how I'd
probably do it if I had the time... Of course, PP client might already
calculate some check sums from it's own graphics and notify the server if
they are altered. And if it's not already done, it could be easily
implemented if they suspected such bots to be in use...
Jeffrey L. Woods
dskl...@aol.com says...
> >why even post if you're gonna write this cryptic crap?
>
> Excuse me? Do you think I would lie about this? Wouldn't you rather have me
> tell you something I am virtually sure of, due to an impeccable source, than to
> say nothing at all?
Unless you're going to say WHAT you know, and to back it up, no, I don't
want to hear the equivalent of:
"I know something you don't know... I know something you don't know..."
Nat Silver
> Sklansky writes:
> > Al says:
> > >why even post if you're gonna write this cryptic crap?
> > Excuse me? Do you think I would lie about this? Wouldn't you rather have
me
> > tell you something I am virtually sure of, due to an impeccable source,
than to
> > say nothing at all?
> Unless you're going to say WHAT you know, and to back it up, no, I don't
> want to hear the equivalent of: "I know something you don't know...
It's Sklansky putting his credibility on the line, giving
you his word: "Do you think I would lie about this?"
In case you are operating under some delusion,
that's saying a lot more than "I know something
you don't know."
Jeffrey L. Woods
mat...@worldnet.att.net says...
I don't care if it's Jesus Christ or The Oracle of Delphi -- if he has
something credible to say, then he should SAY it. If not, then he
shouldn't even hint at it -- it is incredibly annoying, and smarmy
elitism, to throw that kind of Nyah Nyah Nyah Nyah Nyah out there, and
ask us to take it on the faith of the almighty Sklansky.
Tim Dow
>You could easily work around this problem, and I suspect other bot
>detection schemes. Just pay some kid (or do it yourself) $15 bucks an
>hour to monitor your bots and answer these infrequent pop up questions
>as they come. If you've got 30 bots grinding out $4 an hour in 3-6
>games, you will still be making clost to a $100 an hour after
>expenses.
Your method for working around the bot prevention scheme I outlined would
certainly work, but since the kid has to sit there all the time the program
would no longer fit the description of a bot. It would instead be a kind of
"Poker Assistant".
This brings up the point I have been trying to make all along. The real
issue is how good is your software at playing the game of Poker? At such
time as Poker playing software progresses to the point where it can
consistently beat good players, no sensible player will play online, bots or
no bots. For all I know they already don't. I know I personally would not
play online for money even today.
Blackjack has without question been in this boat since the beginning of
online casinos. I think I'm safe in saying (I hope I am) that no blackjack
player with even a basic understanding of the game would play online
blackjack and expect to win in the long run. I have seen debates about
whether or not they shuffle after every hand but as far as I'm concerned the
answer is obvious; they have to.
Its not like I think either the question method or any other bot detection
method will actually be implemented, but I will point out that there are
methods that could be used to eliminate bots without asking questions. As I
alluded to in my earlier post the others would just be a bit more
complicated. A second method is inadvertently suggested by the author of the
following gem, which was the subject of my prior response:
>>Poppycock. If you honestly think that, then you know nothing about AI and
>>networking. Yes, there are a lot of curves they could throw, and very
many
>>of them might put the 'bot author beyond his capability to keep up, but it
>>could never prevent bots entirely. At some point, it has to send bits
over
>>an anonymous wire that put bits on the screen that a human can identify,
and
>>there's simply no way to do that without allowing bots to identify them as
>>well.
If you truly "put the 'bot author beyond his capability to keep up" you have
eliminated bots. I won't get into the details of how this might be
accomplished because it really is a moot point in my opinion. The final
solution to evade detection would be to run the "Poker Assistant" I spoke of
earlier in a separate computer and simply pay the kid you spoke of to type
the hands in.
>If your bots were programmed to subtlety collude with each other when
>they found themselves in the same game, I think that advantage would
>make it relatively easy to program a bot that could beat semi-tough
>games (with 2-3 bots in each game). Remember, with a bot you don't
>need to make 2bb an hour in order to make a lot of money. In fact,
>you would probably would want to win at a much lower rate in order to
>stay under the radar.
Collusion is certainly possible, even among humans, and is another reason I
personally would not play online poker for money. I'm getting a bit far from
what I know for certain, but I think the key here is your word subtle. It is
easy to say bots could be "programmed to subtlety collude with each other"
but much harder to do. That said it would certainly be possible. I think
that blatant collusion would be a pretty easy thing for an online Poker
house to detect using computer software, but they would need to walk a fine
line to avoid accusing players who are simply inept of collusion. What is
less clear to me is how much effort they would want to put into it, after
all they are not the ones losing money.
David R Huberman
> rec.games.programmer, but I don't subscribe to it, so I don't know if it's
> really about games programming...
Supernews, the largest usenet provider in the world, does not carry
rec.games.programmer or rec.programmer.games, therefore they do not
exist in the mainstream usenet feed.
/david
Dsklansky
>something credible to say, then he should SAY it.
What's wrong with you? I did say it. Namely that the bot story is almost
certainly untrue. The fact that I can't divulge the specific reason why I
believe this, might be unsatisifying, but still better than if I said nothing
at all. My statement should make people have large doubts about the original
post, unless they think I am an idiot or a liar. I admit the doubts would be
larger still if I was free to post an explanation. But I am not. One thing I
was not trying to do was brag I am privy to information that others aren't
privy to. But I am. And if half a loaf isn't better than none to you , I am
sure it is to quite a lot of the readers out there.
lvdlrs
"Jeffrey L. Woods" wrote:
I don't care if it's Jesus Christ or The Oracle of Delphi -- if he has
something credible to say, then he should SAY it. If not, then he
shouldn't even hint at it -- it is incredibly annoying, and smarmy
elitism, to throw that kind of Nyah Nyah Nyah Nyah Nyah out there, and
ask us to take it on the faith of the almighty Sklansky.
Me:
I would care if it was Jesus Christ. "Nyah?" Sounds like the bait used
in the movie Jurassic Park. Don't let him get your goat. Eat it first.
Signed T-Rex
Henry Estes
<20010817122459...@mb-fz.aol.com>, Dsklansky wrote..."
David, it's not the half-a-loaf thing. It's that the original poster, should he
desire, would have difficulty replying to what you know.
Henry Estes
A. Prock
>>David, it isn't even hard. It would be a scary proposition, though,
>>because in a demonstration it would be easy for the players to join
>>together to beat up on the bot. Give me an honest game and six months
>>and I'm betting my computer player could soundly beat that "fairly
>>tough" 10-20 game -- especially if it's seven-card stud.
>
>My reason for doubting this is not because I don't think it is theoretically
>possible but rather because of the number of trees involved. It is much more
>than in chess.
David, David, David...
Are you just making this stuff up?
Chess match:
50 moves
~10 possible branches per move
~10^50 state space
10 handed Poker:
52 C 25 hands and boards ~= 10^15
4 rounds, ~10 actions per round, 3 branches per action = 3^40 actions
~10^35 state space
This is easy math, don't wimp out on us. Chess clearly has
a larger state space.
>In spite of the above I would not be surprised if the team that programmed Deep
>Blue could make a winning 10-20 player using their supercomputers and hundreds
>of years of man hours.
This may or may not be true, but I'm certain that it's already been
done by the U Alberta team.
- Andrew
VPOKERL
>Deep Blue could make a winning 10-20 player using their supercomputers and
>hundreds of years of man hours.
Bullshit. Just what are you basing this brazen statement on, your programming
experience? Make a winning 10-20 player against what table type? Loose
passive idiots that are also for a practicle purposes programmend themselves
Could they build one that could beat the 10-20 Holdem game at the Mirage in
Vegas?
Why do Casino's limit there Video Poker games to probabilistic machinations.
Could one of the resons be that they know there is no way for them to gain an
edge? Do not use the excuse that they make money from live games. there are
plenty of Casinos that have no poker rooms. I'm fairly sure that a poker
playing machine would pique their interest given the number of poker players in
this country.
Blah to your theory of progamming!
Vince
RMITCHCOLL
>Bullshit. Just what are you basing this brazen statement on, your
>programming
>experience?
I am sure gonna miss the Vince! forum :-)
Randy Collack
To stay up-to-date on all MARGE happenings, please subscribe to our mailing
list. To subscribe send e-mail to majo...@conjelco.com. In the body type
subscribe marge <your e-mail>
MARGE will be held in Biloxi, Mississippi November 8th-11th.
mhu...@mcloud.slktaho1.nv.home.com
> Why do Casino's limit there Video Poker games to probabilistic machinations.
I believe Nevada gaming law currently says they have to.
> Blah to your theory of progamming!
I'm interested in what your background is in software engineering, computer
science, or mathematics is. You seem very comfortable bashing others
in these areas.
mph
bob nerkul
VPOKERL
Some of you're posts are as funny as the column'ss you write. Just like this
one. here we have poor Wilson software trying in vain for years to turn out a
competent simulator and Mike Caro claiming it's a piece of cake. why Mr.
wilson O'l Mike could do with no problem in six months. you guys must be
stupid!
Vince
VPOKERL
>
Wake up there are casinos in other states. Do you think the Pequat Indians
care about Nevad State law.
>I'm interested in what your background is in software engineering, computer
>science, or mathematics is.
You first.
vince
Richard Kuhler
news:20010817171746...@mb-mk.aol.com...
No offense to Mr. Wilson or his software but whether he's done it or not
says very little about whether it can be done.
Richard
David R Huberman
> >Supernews, the largest usenet provider in the world, does not carry
> >rec.games.programmer or rec.programmer.games, therefore they do not
> >exist in the mainstream usenet feed.
bob nerkul responded:
> groups.google.com has rec.games.programmer, therefore it is
> mainstream.
No it doesn't. Not according to:
http://groups.google.com/groups?hl=en&safe=off&group=rec.games
groups.google.com lists 31 2nd level rec.games. groups, of which
rec.games.programmer is not one.
/david
mhu...@mcloud.slktaho1.nv.home.com
> Wake up there are casinos in other states. Do you think the Pequat Indians
> care about Nevad State law.
Granted. But I've seen allegations that lots of them _don't_ deals cards
in a uniform manner. FWIW there are cases in NV where VP manufacturers have
changed the unformity of the cards dealt. Additionally, at least from
what I. Nelson Rose writes, the jurisdictions that have taken an interest
in providing some control over gambling have copied a lot of what NV has
done.
>>I'm interested in what your background is in software engineering, computer
>>science, or mathematics is.
>
> You first.
Your the one throwing around claims without any support. In any case I have
a AB and MS in Mathematics and have been making money writting code since I
was a freshman in high school (20 years).
mph
dave keiser
> >
> > It's Sklansky putting his credibility on the line, giving
> > you his word: "Do you think I would lie about this?"
> > In case you are operating under some delusion,
> > that's saying a lot more than "I know something
> > you don't know."
>
> I don't care if it's Jesus Christ or The Oracle of Delphi -- if he has
> something credible to say, then he should SAY it. If not, then he
> shouldn't even hint at it -- it is incredibly annoying, and smarmy
> elitism, to throw that kind of Nyah Nyah Nyah Nyah Nyah out there, and
> ask us to take it on the faith of the almighty Sklansky.
Here i am to help you out Jeff baby, put the pacifier in your
mouth while i heat up some Gerber's.
Sometimes people cannot break confidences if they give their word
or it would cause more troble than its worth. It sounds to me like
Dave was trying to aleive concerns of the rgp posters,giving us
consideration and access to his inside knowledge. This is Dave S., A
genius who probably carries a Phi Beta Capa chain around in his pocket
but is too humble to wear it like most elitist would, and could be a
member of Mensa if he so desired.
I've seen him for many years and sometimes give him a dig but
he's one of the class people that make up our sub-culture. He's
written books, worked for casinos, and used to give odds on the spot
in the old days in pot and no- limit games. If someone wasn't up on
the correct odds he would call for Dave if he was in the room. No
angles i ever saw. He's the real deal and on the square.
Now what do you want for dinner, carrots o.k., now lets put your
bib on and enjoy.
VPOKERL
>says very little about whether it can be done.
Well if he's done it, it certainly would say a lot about whether it can be
done. Don't you think?
vince
Badger
Earthlink carries rec.games.programmer. Supernews sucks as a Usenet
provider.
VPOKERL
since I was a freshman in high school (20 years).
I flunked high school trig. But I love Blaise Pascal and some of his buddies.
If you are such a proficient software engineer why don't you just show us with
all of your programming knowledge "How to progam a computer to win at Casino
poker". You and Mike Caro might want to team up. Just think of all the money
you will make. Sklansky says it can be done so why don't you three big brains
prove it someway, somehow, somewhere, someday. I shtink cause you can't.
>Your the one throwing around claims without any support
Beg your pardon? I believe it is Cao and Sklansky and others of that ilk
including you that are claiming something without support. I am "refuting"
your claims with proof. The proof is the "negative". There just ain't no such
animal as you and others "claim". So what do you do? You say, yes, there is no
software now but there could be. Yeah,, yup sure, ha, yuh, ha, yup, aha! Hey
what is that MS? Is that just more BS?
vince
Mike Caro
I already DID it, Vince. About 17 years ago. Check "Science 85,"
"Video Magazine," the "Ripley's Believe It or Not!" TV show on ABC,
and about 35 other sources. I'm not claiming it's easy for anyone
else, though.
Straight Flushes,
Mike Caro
Dsklansky
>genius who probably carries a Phi Beta Capa chain around in his pocket
>but is too humble
Thanks for the kind words. You might have gone a bit overboard though.
AL - Dork to some
>From: dskl...@aol.com (Dsklansky)
>>but is too humble
>
>Thanks for the kind words. You might have gone a bit overboard though.
I agree...
from your posts, you are far from humble!
dave keiser
tee-shirt. Let's try down to earth.
Sean Duffy
Actually it doesn't even matter if something's listed on Google,
because that still has nothing to do with whether or not a group
"exists." I'm not sure if Google is modeled on the old Dejanews
format, but under that format you could just crosspost to any
fictional group, say "rec.gambling.poker.gary-carson," and it would
show up on Google. For a big 8 group to "exist" it needs to be
created via the proper channels, passing a vote in news.groups, etc.
-Sean
mhu...@mcloud.slktaho1.nv.home.com
>>I have a AB and MS in Mathematics and have been making money writting code
> since I was a freshman in high school (20 years).
> I flunked high school trig. But I love Blaise Pascal and some of his buddies.
FWIW I don't think degrees and experience are everything, but I do think showing
that you've spend some time doing some level of research in the area that
your trying to comment on is important. Personally I was hoping you would
tell us you were the ghost of Alan Turing hangin out in the Castro and decided
to come forth and explain to us why we were all mistaken. Instead all we
have is a loudmouth.
> If you are such a proficient software engineer why don't you just show us with
> all of your programming knowledge "How to progam a computer to win at Casino
> poker".
The issue is far less software engineering and far more computer science and
mathematics. But then if you'd bother to know what you were talking about you
would know that.
> You and Mike Caro might want to team up. Just think of all the money
> you will make. Sklansky says it can be done so why don't you three big brains
> prove it someway, somehow, somewhere, someday. I shtink cause you can't.
For a poker player you sure don't seem to understand price. While I'm
fairly sure building a program to beat a paradise 10-20 game wouldn't be
impossible, I don't think it would be easy and I'm not convinced that
by the time you got it built that you wouldn't be bucking fairly extensive
counter measures which would then cost you more to overcome.
>>Your the one throwing around claims without any support
> Beg your pardon? I believe it is Cao and Sklansky and others of that ilk
> including you that are claiming something without support. I am "refuting"
> your claims with proof.
Orac?
> The proof is the "negative". There just ain't no such
> animal as you and others "claim". So what do you do? You say, yes, there is no
> software now but there could be. Yeah,, yup sure, ha, yuh, ha, yup, aha! Hey
> what is that MS? Is that just more BS?
Mason? Doug Grant? A combo of the two?
mph
VPOKERL
Thanks for the compliment or is "It takes one to know one" more appropriate.
Either way, all you offer in the way of proof that poker can or will someday be
beaten by a computer is your "loudmuth" flapping it's jaws. Or did I miss
somehting? Did you offer anything that resembled proof? If you want to sling
bull it's fine with me.
"The issue is far less software engineering and far more computer science and
mathematics."
You don't know that. You are purely speculating. You have ABSOLUTELY NO IDEA
what it will take for a computer to be developed that can be a winning poker
playing. Please prove me wrong, Bigmouth!
>But then if you'd bother to know what you were talking about you would know
that.>
This is much more appropriate a statement towards a bigmouth like you that
claims to know what it takes to win at poker. I don't claim that I can develop
nor do I claim to know what it takes to develop a computer to win at poker, You
did. Maybe you should look at yourself in the mirror when you make silly
comments like the above. Looks like I was right about what your BS stands for.
"I'm fairly sure building a program to beat a paradise 10-20 game wouldn't be
impossible,"
There you go again big mouth. What study have you done?
>Mason?
Gosh I hope your right. I don't know Doug Grant but if I have a choice between
you and him....
Bye.
vince
A. Prock
>>loudmouth.
>
>Thanks for the compliment or is "It takes one to know one" more appropriate.
*plonk*
- Andrew
Lee Daniel Crocker
> Lee Daniel Crocker wrote:
>
> >> This paragraph makes the whole thing much less believable to me. If
> Paradise
> >> wants it can implement trivial changes it its client software that will
> not
> >> only "slow down" this type of bot, but render it useless for at least the
> >> next 50 years or so.
> >
> >Poppycock. If you honestly think that, then you know nothing about AI and
> >networking. Yes, there are a lot of curves they could throw, and very many
> >of them might put the 'bot author beyond his capability to keep up, but it
> >could never prevent bots entirely. At some point, it has to send bits over
> >an anonymous wire that put bits on the screen that a human can identify,
> and
> >there's simply no way to do that without allowing bots to identify them as
> >well. Humans are much better at pattern-matching than bots, but
> recognizing
> >cards and buttons and text is not exactly a challenge.
> >
> >And even if I'm wrong about that, that in no way makes the post
> unbelievable,
> >because its quite clear that the author believes it and has every reason
> to.
>
> Actually I posted a response on this issue because I have 25 years of
> experience in the field of artificial intelligence. If it were about Poker
> strategy I would have kept quite and tried to learn something. Maybe not bad
> advice for you regarding AI? As far as networking is concerned it has no
> relevance to this discussion.
>
> Its not clear to me how someone could be sure enough of his position to
> refer to someone else's as "Poppycock" and then in the next paragraph admit
> that he might be wrong, but I will humor you and give you an example showing
> that you are, in fact, quite wrong.
I don't want to get into a credential contest here (mainly because I have
no formal education--I am a self-taught programmer), but I'll just say that
I do know something about the field; I've had lunch with Marvin Minsky and
spoken with Ray Kurzweil; I've attended conferences where AI was a major
topic, and I've been programming computers since 1979. But again, whether or
not I am correct about my assessment of AI, what I called "Poppycock" was
your assertion that the poster can't be sincere because /he/ believes that
he can circumvent Paradise's attempts to thwart him. That's clearly relevant,
and you should be able to understand that. I did not mean to imply that
your opinion of AI is nonsense (though I do happen to disagree with that also).
> If Paradise wanted to stop bots from playing cards (which I doubt they do) I
> can think of at least 3 completely different approach's that, as I said,
> would eliminate them for at least 50 years or so. The simplest but also the
> most intrusive would be to ask a question in a pop up window at random
> times. It seems to me if people wanted to be assured of playing at a bot
> free table this small inconvenience might be acceptable.
Yes, Turing-test stuff would be the easiest way to thwart a bot; I don't
think it would be practical, though. Not only would it annoy players, but
it would require human staff to evaluate the replies--after all, if the
replies are judged by algorithms with below-human intelligence, then human
intelligence is not necessary to fabricate passing responses. The cost of
human staff to evaluate these replies would rival the cost of real dealers
in a casino, taking away the online's biggest advantage.
Your sample questions are good, and would indeed be beyond the present
capabilities of AI (your "50 years" guess is still overly pessimistic
from what I've seen, but I'll certainly grant you 10). If they did hire
some people to create a few hundred of these questions, and then had bots
ask them and evaluate responses (which would be the only way to save
costs), then the players would be further pissed off by the failures of
the bots to recognize some valid answers ("zeroe"), so we're back to
needing human staff.
> Again, my point had nothing to do with his programming skills or lack
> thereof but was simply to point out that writing the part of the bot that he
> stated was the most complex is actually trivial, but writing a poker
> program that can beat even an average casino player consistently over the
> long haul is something so complex that it has not been accomplished yet.
Loki already beats an "average casino player" easily. It can't beat a good
player, but it racks up thousands on IRC, and doesn't lose it all back to
us good players very fast. Yes, writing a bot that beats good players is
hard, and tough it _has_ been attampted many times (don't forget about
Turbo Texas Hold'em and the like), it has not been entirely successful.
But beating idiots isn't that hard. :-)
Dennis Altman
be very important. The bot could scan the lists of all tables, and then select
the table where he has had the most profit against the given players. This was
not brought up at all by the writer. If you are going to have a database of
players, this is how I would use it to the fullest extent.
--
Sincerely,
Dennis Altman
Wedmart.com
http://www.wedmart.com
888-802-2229
Please send paypal payments to auct...@wedmart.com
Lee Daniel Crocker
> My statement should make people have large doubts about the original
> post, unless they think I am an idiot or a liar.
Or if they think one of (1) you aren't an expert on computer programming
competent to evaluate the post, or (2) you have some personal or financial
interest in the story being false.
I have no reason to suspect either of those things, but then I have no
particular reason to doubt them either. What "credibility" you have here
applies to poker strategy, and only that. Any other claim you make
without any evidence or details holds no more water than the same claims
made by the likes of CGA.
--
Lee Daniel Crocker <l...@piclab.com>
<http://www.piclab.com/lee/>
Eric Taylor
Dsklansky <dskl...@aol.com> wrote:
>What's wrong with you? I did say it. Namely that the bot story is almost
>certainly untrue. The fact that I can't divulge the specific reason why I
I think you're wrong. You have demonstrated you have very little
knowledge about computer AI. So I have to speculate your secret
information is something you have heard second hand and take on faith.
In which case I do not trust it at all. Analyzing these 'bots doesn't
require a poker expert. They require a computer expert. You are not
one.
Mason Malmuth went so far in the two + two board to say that he didn't
even read the article. And he says it's impossible. Without reading
the article or any of the messages. Now that's rich.
--- edt
Jonathan Kaplan
>
>...Analyzing these 'bots doesn't
>require a poker expert. They require a computer expert....
actually, a successful bot (against average mid to low limit players) is going
to require BOTH a poker expert AND a computer expert, more probably groups of
people working together. i think it can and probably will be done. i am sure
people are working on it right now.
but a bot that is going to be successful against great players, especially in
games with numerous participants, is going to require even more skill than
"expert" in at least one of these fields, and probably again, in both.
and a bot that is going to be as good as the best humans in no limit holdem or
pot limit omaha, well, that bot might be possible. maybe. but the effort
required to do this is going to be incredibly costly, in money and man-hours.
i'll believe it is possible when i see it.
maybe we'll see, someday.
i am willing to bet big money that i will win the WSOP championship before any
bot will. and i know how unlikely it is that i will ever win the big thing.
smile
Jonathan
[i have been hanging out in the backgammon newsgroup some, that group is
dominated by posts about BG bots. it seems the technology is high enough that
bots will beat almost everyone else who plays backgammon, in the long run.
i think that poker bots are still far away from that level of skill. but then,
backgammon is a much much more "definable" game.]
no matter where you go, there you are...
Dsklansky
>knowledge about computer AI. So I have to speculate your secret
>information is something you have heard second hand and take on faith.
>
>In which case I do not trust it at all. Analyzing these 'bots doesn't
>require a poker expert. They require a computer expert. You are not
>one.
>
>Mason Malmuth went so far in the two + two board to say that he didn't
>even read the article. And he says it's impossible. Without reading
>the article or any of the messages. Now that's rich.
>
>--- edt
>
a computer to play ring games well. He and I know of two well known examples
where they don't. But my opinion comes from reliable information that I can't
divulge rather than computer expertise.. But I'll bet on it. Those who think
that doesn't matter have a big problem.
A. Prock
> Mason's conclusions come from his knowledge of how difficult it is to program
>a computer to play ring games well.
Mason probably couldn't code up "Hello, World" without a manual.
I'm fairly certain he's never coded up an artificial neural
network. In fact, I'm almost positive that he doesn't understand
the U of A papers that have been published in AIII.
His opinion is fairly worthless when it comes to computer poker.
>He and I know of two well known examples
>where they don't.
I know of two well known examples where they do.
>But my opinion comes from reliable information that I can't
>divulge rather than computer expertise.
What? Chuck matched the IP address of the original poster with
that of "Ray Springfield" or some other torll? Sure, that
wouldn't surprise me. On the other hand, that doesn't begin to
address the phenomenon of the heads-up bots that populate Paradise
Poker.
>But I'll bet on it. Those who think
>that doesn't matter have a big problem.
Your opinion on computer poker is pretty much a non-sequitor
in my book. This is an area where you have pretty much no
prior practical experience. Someone like Mike Caro, or Darse
Billings has the credibility to talk about computer poker
intelligently.
You don't.
- Andrew
swami
> >But I'll bet on it. Those who think
> >that doesn't matter have a big problem.
>
> Your opinion on computer poker is pretty much a non-sequitor
> in my book. This is an area where you have pretty much no
> prior practical experience. Someone like Mike Caro, or Darse
> Billings has the credibility to talk about computer poker
> intelligently.
>
> You don't.
Of course Sklansky is not as smart as he thinks he is (neither am I),
and of course his expertise does not extend to every domain. The "big
problem" refers to his implicit claim that his intelligence consists
of the ability to evaluate evidence in the context of probability.
This claim is too general to be strictly true, but without seeing the
evidence yourself, it's hard to discount specific applications of it.
Even though you reject Sklansky as an authority on computer poker, and
may be annoyed at his silence re: his "evidence" against bots, just
ask yourself if you would want to fade his action.
If I knew that he knew the limits of his expertise, I might be happier
about the situation, but that's just gratification for my ego. I
wouldn't take the other side in either case.
Dsklansky
that he would believe Billings or Caro. Suppose my information was that Caro or
Billings told me so.
AL - jus AL
>From: dskl...@aol.com (Dsklansky)
>Suppose my information was that Caro
that and 2 cents....
dave keiser
To SYeates485: Maby he is drunk, i don't see any reason to jump
into a good thread with information flying around. Doesn't make sense.
I will be an addict the rest of my life but haven't been a tweaker
since Jan.4,1999.
Eric Taylor
Hi David,
If the information is from Billings or Caro, then you would be
expressing an opinion based on second hand information about a computer
AI field of which you have no experience, taking it purely on faith
rather than actual knowledge and experience of the field involved.
Which is exactly what I presumed you were doing.
Look, I'll believe first hand information.
For instance, if you have a coroner who autopsies a body (in this case
the body is a 'bot), then decides that there was foul play involved,
sure I would believe him, he's an expert in the field.
However, if the person talking about foul play is the friend of the
friend of the coroner, well, come on now, really, you expect me to
believe it? If so I got a nice bridge to sell you.
You see, it doesn't matter if the source of your information is Caro or
Billings or any computer expert, the problem is that it is second hand,
you yourself are not an expert, and thus it is not reliable. Any
history researcher, policeman, judge, or even a parent trying to figure
out what happened to the car can tell you how unreliable information is
when you hear it second or third hand.
Second or third hand information is worthless enough that you should
discount it entirely if you have first hand evidence of which we have
plenty, the author of the 'bot in the original article, and several
computer AI experts who have weighed in (on both sides of the issue,
though it is mostly slanted towards believing the original author) both
on two + two and here.
--- edt
SlmShdy957
just compare wut he knows to wut u know
Larry W. (Wayno) Phillips
>i dont know how u guys can be dissing poker experts like sklansky
>just compare wut he knows to wut u know
Is that you David? It might interest you to know that
we can't be fooled by really, really, really bad spelling.
--- others have tried this ploy
Signed,
We Are Smart
Dsklansky
>Billings or any computer expert, the problem is that it is second hand,
>you yourself are not an expert, and thus it is not reliable.
That's ridiculous. Assuming hypothetically that I am parroting the opinion of a
computer expert you would trust, the only problem would be if I was lied to or
misunderstood him. And if I was sure that I wasn't, than it would only be my
opinion as to the honesty of my source rather than my computer expertise that
would matter.
m...@acm.org
>>You see, it doesn't matter if the source of your information is Caro or
>>Billings or any computer expert, the problem is that it is second hand,
>>you yourself are not an expert, and thus it is not reliable.
> That's ridiculous.
Depending on the form and complexity of the information the quoted
statement could be true.
> Assuming hypothetically that I am parroting the opinion of a
> computer expert you would trust, the only problem would be if I was
> lied to or misunderstood him.
Trust and infallibility are different. For example (on a related but
different topic) ou might tell me the Bruce Schnider told you that the
encryption used was unbreakable in reasonable game time. That would
increase the probability that you were correct to the point that Bruce
is correct, but not beyond.
> And if I was sure that I wasn't, than it would only be my
> opinion as to the honesty of my source rather than my computer expertise that
> would matter.
Up to the level of correctness of your source and since we don't know
who your source is its hard to apply a reliability number to it.
Furthermore the more complicated what you were told (were you told "it
can't exist" or were you told "fact 1, fact 2, fact 3" and left to
come to your own conclusions) the more likely you could have
misunderstood it or drawn incorrect conclusions from it.
Personally I think your statement that the bot didn't exist is one of
the less interesting in this thread. In the hierarchy of information
that I knew before the original post and what I've learned since the
original post your claim sits in the pot of corroborating but hard to qualify
information.
From your standpoint I don't see getting very upset because people are
unwilling to accept a fairly nebulous statement unconditionally. You
threw out as much as you could but left a lot to faith. In a group of
people who think relatively logically you should _expect_ a lot of
people to look on such information with questioning.
mph
Jaeger T. Cat
<m...@acm.org> wrote:
>
>From your standpoint I don't see getting very upset because people are
>unwilling to accept a fairly nebulous statement unconditionally. You
>threw out as much as you could but left a lot to faith. In a group of
>people who think relatively logically you should _expect_ a lot of
>people to look on such information with questioning.
>
But, Michael, you just don't understand. It's _David_ _Fucking_
_Sklansky_ telling you this. Bow down before him, and accept
his knowledge.
After all, he's got a website and poker forum.
Snort.
Dsklansky
>unwilling to accept a fairly nebulous statement unconditionally. You
>threw out as much as you could but left a lot to faith. In a group of
We have some major reading comprehension problems here. I myself didn't say it
unconditionally. I have problems only with those who won't alter their opinion
of the probability the poster spoke the truth given my statement. And what
makes you think that computer expertise comes into play in this at all? Maybe
I know the poster personally and he told me about his planned hoax. One thing
is for sure. If I am willing to lay 3-1 the post was a fraud, then so too
should you be.
AL - jus AL
>From: dskl...@aol.com (Dsklansky)
> Maybe
>I know the poster personally and he told me about his planned hoax.
maybe this...maybe that!!
why bother typing this crap?
You continue to throw this cryptic nonsense out there.
>If I am willing to lay 3-1 the post was a fraud, then so too
>should you be.
if we all bet the same, there'd be none to lay off.
Jonathan Kaplan
>
>>From your standpoint I don't see getting very upset because people are
>>unwilling to accept a fairly nebulous statement unconditionally. You
>>threw out as much as you could but left a lot to faith. In a group of
>
>We have some major reading comprehension problems here. I myself didn't say it
>unconditionally. I have problems only with those who won't alter their opinion
>of the probability the poster spoke the truth given my statement.
i agree that, given your statement (and all other details so far), an
intelligent gambler should alter the probability somewhat, in the direction of
the views you state. the question is, how much? David, you seem to act like
people should alter their probability assessment very far, even completely, in
the direction you state. others (i forget specifically who), say they shouldnt
alter probability assessment at all, because of the second handedness of the
info, and your lack of computer expertise. i think somewhere in between (but
mathematically, closer to David's side), is correct.
David S., when you make the kind of statement you have made (including the
"willing to wager" aspect), i will alter in your direction. but not completely,
because of the same concerns that others have expressed.
and you really shouldnt expect others to alter their assessment of probability
all the way, either. (probably you dont really expect people to do that, but the
way you write things some times gives the impression that you do.)
decisionmaking is a very complex art.
>...And what
>makes you think that computer expertise comes into play in this at all? Maybe
>I know the poster personally and he told me about his planned hoax. One thing
>is for sure. If I am willing to lay 3-1 the post was a fraud, then so too
>should you be.
in fact, given this discussion (in toto), i am now willing to lay 11-4 that the
post was a fraud....grin
Jonathan
A. Prock
>If I am willing to lay 3-1 the post was a fraud, then so too
>should you be.
And if you were willing to jump off a bridge...
?
- Andrew
Jim Geary
expressed a willingness to do so, the hypothetical offered would seem to
underscore the validity of his willingness to assert other things.
Jim Geary
jimgeary.com - something to bore everyone
John M. Davis
> [...] I have problems only with those who won't alter their opinion
> of the probability the poster spoke the truth given my statement.
Appeals to authority without evidence are rarely convincing. When the
appealer is the authority they are amusing. Sometimes it is best to
say nothing...
> Maybe I know the poster personally and he told me about his planned
> hoax.
Maybe you are the poster. Maybe you are the bot -- and you don't work.
> One thing is for sure. If I am willing to lay 3-1 the post was a
> fraud, then so too should you be.
You are a silly bunny. If you want blind devotion, that's why God
created 2+2.
Cheers.
"If you want to start explaining things that you yourself don't have a
good understanding of; well that's why God created RGP." -- David Sklansky
Lee Daniel Crocker
> And what makes you think that computer expertise comes into play in this at all?
> Maybe I know the poster personally and he told me about his planned hoax. One thing
> is for sure. If I am willing to lay 3-1 the post was a fraud, then so too
> should you be.
In all seriousness, that _is_ in fact the most convincing thing I've read on
this thread. As a computer expert, my informed opinion is that the claims of
the post are entirely plausible; you are, as you point out, not claiming
otherwise. You are merely claiming that this particular posting--even if it
is entirely plausible--is a fraud, and while you won't give any real details
or evidence for that claim, the fact that you are willing to lay odds on it
does say something. Even those of us who don't hang on your every word know
that we'd rather be betting with you than against you. :-)
Larry W. (Wayno) Phillips
wrote:
>dskl...@aol.com (Dsklansky) writes:
>
>> [...] I have problems only with those who won't alter their opinion
>> of the probability the poster spoke the truth given my statement.
>
>Appeals to authority without evidence are rarely convincing. When the
>appealer is the authority they are amusing. Sometimes it is best to
>say nothing...
>
>> Maybe I know the poster personally and he told me about his planned
>> hoax.
>
>Maybe you are the poster. Maybe you are the bot -- and you don't work.
>
>> One thing is for sure. If I am willing to lay 3-1 the post was a
>> fraud, then so too should you be.
>
>You are a silly bunny. If you want blind devotion, that's why God
>created 2+2.
>
>Cheers.
"The longer I live the more I see that I am never wrong
about anything, and that all the pains that I have so
humbly taken to verify my notions have only wasted
my time."
George Bernard Shaw
John M. Davis
[ re Sklansky ]
> You are merely claiming that this particular posting--even if it is
> entirely plausible--is a fraud [...]
Arguable.
| A Probable Lie
| Posted By: David Sklansky <Dskl...@aol.com>
| Date: Thursday, 16 August 2001, at 5:51 p.m.
| In Response To: Re: Bot and paid for (goodbeet)
| There are several reasons, some which I cannot divulge. Ones I can are that
| there is not enough difference between play money and .50-1.00 results.
| Another reason is that while it is simple to use game theory headup,
| multiway cannot be programmed by anybody who is not an expert himself. But
| again the biggest reason I seriously doubt this poster is related to
| confidential information given to me which is highly reliable.
| PS could someone please repost this on RGP
m...@acm.org
>>You see, it doesn't matter if the source of your information is Caro or
>>Billings or any computer expert, the problem is that it is second hand,
>>you yourself are not an expert, and thus it is not reliable.
>
>computer expert you would trust, the only problem would be if I was lied to or
>opinion as to the honesty of my source rather than my computer expertise that
>would matter.
Dsklansky <dskl...@aol.com> wrote:
>>From your standpoint I don't see getting very upset because people are
>>unwilling to accept a fairly nebulous statement unconditionally. You
>>threw out as much as you could but left a lot to faith. In a group of
> We have some major reading comprehension problems here. I myself didn't say it
> unconditionally. I have problems only with those who won't alter their opinion
> of the probability the poster spoke the truth given my statement.
I don't know if you find it interesting to debate semantics here, if
you really believe "we" have a reading comprehension problem, or if
you want to muddy the waters. I never said you said "it
unconditionally". I said "people are unwilling to accept a fairly
nebulous statement unconditionally". So far you've wanted to be
considered the perfect conduit of information. No errors introduced in
your statements. Thats unconditional belief.
> And what
> makes you think that computer expertise comes into play in this at all? Maybe
> I know the poster personally and he told me about his planned hoax. One thing
> is for sure.
I don't know if computer expertise comes into play in your statement
because you havn't given me enough data to evaluate your statement to
that level of detail.
FWIW, from a debate standpoint this series of hypothetical situations,
most of which contain wild assumptions, do nothing for your argument.
They appear as if you want to cover up the core issue. If you want to
tie down some of these assumptions then maybe more people will
increase their level of belief in you.
> If I am willing to lay 3-1 the post was a fraud, then so too
> should you be.
Given only the information you've posted even if I believe you 100%
(that is, you are not an error source) I wouldn't do that. I'm not a
religious man. Faith is hard for me. Tell me something that convinces
me your source should be believed to that level (identity,
credentials, method of determination, etc.) and then I'll give
credence to following your lead. Just the fact that your evaluation
of his credentials leads to make that bet isn't enough for me.
So, would you lay 3:1 that the post was a fraud or what just another
hypothetical intended to redirect the discussion?
mph
Dsklansky
>>should you be.
>
>if we all bet the same, there'd be none to lay off.
It is likely that there isn't a professional gambler in the whole world who
would not take a piece of any substantial bet I made laying three to one on
anything. Sight unseen. The truth is that the great majority of the readers
here substantially changed their assessment based on my comments. Those few who
want to be stubborn are just showing that they don't know how to properly
adjust opinions given less than perfect evidence. Hope you get in one of my
poker games.
John M. Davis
> The truth is that the great majority of the readers here
> substantially changed their assessment based on my comments.
Don't get me wrong. I'm *for* wild unsubstantiated claims -- Take, for
instance, your knowledge of how the great majority of readers
initially viewed the post, and then changed their views after reading
your comments. I guess when you are the center of the universe this
sort of omniscience comes with the territory.
Given that you can confidently speak for a majority of people that
you've never seen or met, then I'm in what must be a tiny minority --
I skimmed the initial post skeptically, written as it is in the
breathless prose that made Dave Rhodes famous -- and then read your
follow-up as content free bluster about sekrit information, multi-game
theoretic player fooferoo, and how somebody who can beat .25/.50
holdem would need to learn new skills to play the 20/40 at Mirage or
some such -- I don't remember.
Cheers!
AL - jus AL
>From: dskl...@aol.com (Dsklansky)
>here substantially changed their assessment based on my comments.
your comments were cryptic and vague. I'd base my assessment more on fact.
>Those few who
>want to be stubborn are just showing that they don't know how to properly
>adjust opinions
it isn't a question of stubbornness..it is a question of following someone
because they say you should. I'd follow someones advice based on the
arguments/facts etc that they provided.
>given less than perfect evidence
none is quite a bit less...no?
> Hope you get in one of my
>poker games.
only if ya play for quarters!! haha
Jaeger T. Cat
Dsklansky <dskl...@aol.com> wrote:
>
>anything. Sight unseen. The truth is that the great majority of the readers
>here substantially changed their assessment based on my comments. Those few who
Let's see if we can count the logical errors here:
1. confusing "those who post" with "those who read"
2. confusing "those posting that their position changed" with "those who read"
3. equating "those who don't post" with "those who agree with me".
4. making a broad, unsubstantiated generalization, without proof (unless
you've got a RGP "readers survey" to back up your claim.
Face it, David, you're completely full of shit on this issue.
Badger
> I don't know if you find it interesting to debate semantics here, if
> you really believe "we" have a reading comprehension problem, or if
> you want to muddy the waters. I never said you said "it
> unconditionally". I said "people are unwilling to accept a fairly
> nebulous statement unconditionally". So far you've wanted to be
> considered the perfect conduit of information. No errors introduced in
> your statements. Thats unconditional belief.
David seems to forget that he has no credibility at all on technology
matters. He has even previously stated that his reputation only extends to
poker strategy in the endorsements he makes. "Game Night with David
Sklansky", or whatever that thing was called, render all his statements on
this topic worthless. He believed what somebody told him then, and it was
crap.
He might be right in what he posted this time or he might not, but just
because he said it is at *best* neutral to the truth of it.
--
Steve Badger
http://www.playwinningpoker.com
m...@acm.org
> <m...@acm.org> wrote...
>> I don't know if you find it interesting to debate semantics here, if
>> you really believe "we" have a reading comprehension problem, or if
>> you want to muddy the waters. I never said you said "it
>> unconditionally". I said "people are unwilling to accept a fairly
>> nebulous statement unconditionally". So far you've wanted to be
>> considered the perfect conduit of information. No errors introduced in
>> your statements. Thats unconditional belief.
> David seems to forget that he has no credibility at all on technology
> matters.
> He has even previously stated that his reputation only extends to
> poker strategy in the endorsements he makes. "Game Night with David
> Sklansky", or whatever that thing was called, render all his statements on
> this topic worthless. He believed what somebody told him then, and it was
> crap.
Actually I interpreted that incident as more David not believing that
an endorsement of a bogus produce carried any reprecussions to his
credibility if he explained he just did it for money. Thats more of a
social and marketing mistake. I believe its relevant to his current
proclaimation but not in the way you suggest.
Take his current claim that _if_ (gotta remember he never said he
_would_ lay 3:1) he would lay 3:1 then so should you. That in itself
is flawed logically. By his earlier actions we don't know if he might
also have an agreement with the Paradise such that he gets paid if he
discovers a security issue including operating bots. In that case the
bet is just an arbitrage to him.
> He might be right in what he posted this time or he might not, but just
> because he said it is at *best* neutral to the truth of it.
I actually think there is some truth in what David has said. The
question in my mind is where the lines between knowledge, supposition,
and implication are. Does David know this is a hoax through and
through (what he has been implying but not stating)? Maybe David
"knows" that the original poster was lieing about which online card
room was being used (in order to provide cover). Now David can
resolutely claim he would lay 3:1 that the post was a "hoax" while at
the same time not knowing if the bulk of the original post was true or
not.
So instead of saying "at *best*" I'd probably say "at *worst*", but
that might just be splitting hairs.
mph
SlmShdy957
i dont know how u guys can be dissing poker experts like sklansky just
compare wut he knows to wut u know
>
> Is that you David? It might interest you to know that
> we can't be fooled by really, really, really bad spelling.
> --- others have tried this ploy
no this is not david sklansky
this is actually a 13 year old who would like to be a poker pro
someday
i hav read many of his books and i hav come to a conclusion that he is
a genius
maybe he is not an expert on computers but he would still know some
basic stuff to come to a reasonable conclusion
AL - jus AL
>
>I posted:
>i dont know how u guys can be dissing poker experts like sklansky just
>compare wut he knows to wut u know
why aren't you in school?
Eric Taylor
>unconditionally. I have problems only with those who won't alter their opinion
>of the probability the poster spoke the truth given my statement. And what
>makes you think that computer expertise comes into play in this at all? Maybe
>I know the poster personally and he told me about his planned hoax. One thing
>is for sure. If I am willing to lay 3-1 the post was a fraud, then so too
First of all, I don't think that you know the 'bot poster personally.
The reason is that you have carefully not given a hint of where your
secret information came from. If you blurted out that you might know
the poster, then the original poster would get mad, and your secret
would be out. Of course, you could be double thinking believing that
we would think you would think that you would never say this, but the
probability of this is negligible because you have demonstrated that you
think anyone who doesn't believe your word is stupid.
Your method of reasoning makes me doubt you more not less. You see,
your vehement defense of hearsay makes me think more and more that you
never had any first hand evidence at all, that in fact your evidence is
itself heresay. I can imagine it already.
The following is a fictionalized encounter, please don't mistake it for
reality or you might get hurt:
<MMalmuth> Hey those 'bots are fake.
<Dsklansky> Really? What is your evidence?
<MMalmuth> It's secret.
<Dsklansky> It doesn't matter, you are smart, so I should take your word.
Hey world the 'bots are fake!
If you just came out and admitted it would be difficult for us to
believe you, I would be more tempted to believe you might have some
first hand evidence but the more you tell us that your word is good
enough for us to stake a 3 to 1 bet the more I think you never had any
evidence at all, just the word of some poker pro friend of yours.
Heresay evidence is just as worthless in real life as it is in a court
of law.
--- edt
Larry W. (Wayno) Phillips
>no this is not david sklansky
Sure David. Nice try
>this is actually a 13 year old
>
>and i hav come to a conclusion that he is
>a genius
All the best,
Tubby Shady
Dsklansky
>matters. He has even previously stated that his reputation only extends to
>poker strategy in the endorsements he makes. "Game Night with David
>Sklansky", or whatever that thing was called, render all his statements on
>this topic worthless. He believed what somebody told him then, and it was
>crap.
>
>He might be right in what he posted this time or he might not, but just
>because he said it is at *best* neutral to the truth of it.
>--
>Steve Badger
>http://www.playwinningpoker.com
Okay I'm laying you $4000 to $1000 that the post was a hoax. Yes or no?
AL - jus AL
>
>Okay I'm laying you $4000 to $1000 that the post was a hoax. Yes or no?
>
ok..let's see... hmmm..
perpetrate a hoax, then bet it's a hoax?
or.... know the hoaxer, or just bein obnoxious.... whichever it is..this thread
is moronic.
Making that offer of a bet doesn't add to your credibility, or make your
remarks any less cryptic and annoying.
Dsklansky
>perpetrate a hoax, then bet it's a hoax?
>or.... know the hoaxer, or just bein obnoxious.... whichever it is..this
>thread
>is moronic.
>
>Making that offer of a bet doesn't add to your credibility, or make your
>remarks any less cryptic and annoying.
>
So you are saying that my offering the bet means I might have pepetrated the
hoax myself to beat Badger out of a thousand bucks? Rather than taking umbrage,
I will point out even if that were true you are admitting that my offer of the
bet has increased the chances that the post was a hoax. QED
John M. Davis
> So you are saying that my offering the bet means I might have
> pepetrated the hoax myself to beat Badger out of a thousand bucks?
> Rather than taking umbrage, I will point out even if that were true
> you are admitting that my offer of the bet has increased the chances
> that the post was a hoax. QED
I'll lay four to one that you'll continue to lay three to one rather
than expand upon your mumbullery (while wearing five-oh-ones) that one
of the reasons the bot couldn't exist is because beating a play-money
or .50/1.00 game is so different in kind from the higher limits.
If you recall, that was one of the non-sekrit reasons for your
dismissal of the post...
AL - jus AL
>So you are saying that my offering the bet means I might have pepetrated the
>hoax myself to beat Badger out of a thousand bucks?
one of many possibilities since your reasoning has yet to be clarified.
>Rather than taking umbrage,
why would you think it an insult? It is based on reasonable thinking since you
have not substantiated your stance.
>I will point out even if that were true you are admitting that my offer of
>the
>bet has increased the chances that the post was a hoax.
Whether a hoax or not isn't the question. The question is whether I should
believe it a hoax..based solely on the fact that you say it is. Odds,
probabilities, etc... I'll take your word for, but why should I believe you
based on nothing?
>. QED
QED ....
http://www.amazon.com/exec/obidos/ISBN%3D0691024170/103-5820832-9333451
m...@acm.org
You responded to my post but quoted badger. Assuming you mean me I'd
have to first understand 1) what your proposing as a bet and 2) how
you propose to resolve the bet. I'm most likely wouldn't take it
unless you made more specific what part you were going to prove was a
hoax. If you going to stick to your current language I'm certain you
could find at least one part of the original post that was deceitful.
mph