TheSwan
First post on this site (thanks for the info so far)-
I've noticed that many people keep quoting the statistic that if you can control the
throws 1 out of 42 or 43 times that you will have reduced the house edge to zero. I'm sure
you've all seen it too.
Please refer me to the source of this myth. I am more than confident that reducing the
number of 7's thrown by 1/241 will reduce the house edge to zero. This assumes that the
1/241 is spread uniformly across all other possible outcomes. Place bets on 6 or 8 yield
the best player's edge if this is done.
If this is true, then this should make former subscribers to the 1-in-43 house of
precision at least 6 times as happy. I'd like to share my thoughts on the subject, but an
Excel spreadsheet and rudimentary statistics would produce similar results.
The greatest part of craps is that you need no Monte Carlo simulation to determine whether
or not it is worth your time to practice dice setting. But with a measly 1/241 increase in
your SSR, it seems that we should be teaching our children how to control dice.
irishsetter
It's not my quote so I cannot verify the veracity of the
statement. However here is the info from the person who did the math:
How Sharpshooter Came to 1 controlled throw in 43 equation....
Solving the problem initially took a lot of work (and pages of calculations), however once
you know the answer and can go right to the 42 and 43 throw scenario, proving it is rather
easy. Consider the following calculations for random at 42 throws:
There are five 6s or 8s in 36 throws, so there will be 42/36 times five 6s or 8s in 42
throws. This equals 35/6, 6s or 8s. There are seven 7s (random) in 42 throws. This means
if you randomly throw the dice 42 times, you will hit (35/6) 6s , for example, and seven
7s. You will win 7/6 units times 35/6 hits = +6.805555 units and you will lose 1 unit
times 7 hits = -7.0000 units for random. you will lose -0.194444 units which is divided by
the number of decisions (35/6 + 7), or -0.1944 x 6/77 = -0.0151515 which is -1.515% (this
verifies random).
Now, when you add "one controlled throw" to the 42 with a guaranteed no 7, there
is now a 5/30 chance of it being a 6 or an 8. You will now win 5/30 times 7/6 more units,
but still lose only 7 units. Wins equal 5/30 (or 1/6) plus 35/36 = 6. 6 x 7/6 = +7 units
won in 43 throws. and seven 7s times one unit means -7 units lost in 43 throws -- a
break-even situation that dictates a 0% house edge. (7 - 7)net units divided by 13
decisions = 0.0000!
Now let's look at your friend's scenario: If you can only control one throw in 82, then 81
are random throws and the one controlled (no 7) roll has a 1/6 chance of hitting a 6 or 8.
Wins = 81 x (5/36) + 1/6 = 11.41667 units times 7/6 payoff = +13.31944 units won. 81
random rolls times (36/6) = 13.5, 7s thrown = -13.5 units lost. So, 13.31944 units won
minus 13.50000 lost = -0.180555 units divided by (11.41667 + 13.50000) decisions (BTW this
works out exactly to 24 and 11/12's), = -0.007246 or -0.7246 %. That means that a thrower
who controls the dice (no 7 guarantee) only one throw in 82 is playing at about a negative
seven-tenths of a percent edge when placing the 6 or the 8 as opposed to the random
-1.515% edge.
You can also look at Krigman's related (Shift of sevens).
TheSwan
Ah, I see. My error was assuming that the 1 controlled
throw out of 43 was distributed evenly across numbers when I should have distributed it
evenly across outcomes. A subtle difference, but enough to make my number wrong.
Now I don't know if dice setting is worth my time. 1 out of 43 seems pretty difficult to
maneuver, even for a graceful shooter like The Swan.
Sure you might be able to get .5% over the house, but the amount of time you invest in
practicing and studying justifies a hobby- not a living.
wrecknball
Actually, 1 of 43 is a piece of cake. The problem with that
number is that it is based on the long term (thousands of throws). If ALL I could do is
alter 1 throw of 43 in a years worth of sessions, no, all this effort would not be worth
it to me. OTOH, let's say you could control even 1 in 10 in the short term, (which I
believe I can do), then it is definitely worth it. Lastly, even if you can overcome the HA
when you're shooting, still means that you are at a disadvantage when others are shooting,
so it pays to shoot solo as often as possible and know how to play the game well. Dice
setting, even for the hobbyist can be very beneficial...
TheSwan
Wrecking Ball,
One out of 43 isn't a "piece of cake" I imagine. What's more, The Swan doesn't
think that your capability to "throw 1 out of 10 in the short run" carries any
weight in this argument. The Swan and his Cygnets could care less what you do in the short
run- we believe in style, grace and the law of large numbers. Not to take away from your
precision-shooting ability; on the contrary, I'm sure you're blessed with a fine toss. But
either you are not choosing a large enough sample, or your boast of 1/10 less of an SSR
needs to be further investigated/exploited.
If you can throw 1/10 in the long run, meet The Swan in Windsor on March 8. We should make
a mockery of those silly Canadian warblers. You will know it is The Swan when you see me
shoot.
irishsetter
Moderator's note: The tone of this
thread is beginning to seem disingenuous but I'll see where it leads for the time being.
I've never met ANYONE who would have thrown the dice enough in a casino setting to qualify
for "the long run."
Let's say you have someone who plays 5 times a week, handles the dice each day 3 times,
and plays 50 weeks each year. Furthermore, we'll go way out on a limb and say he averages
over that period of time 10 throws each hand. That would give said shooter 7500 in casino
throws per year. Certainly I don't know what WOULD qualify as the long run, but I do know
that 7500 doesn't qualify.
For nearly all craps players, The law of large numbers is an esoteric theory. The casinos
live and die by it, but the average player need not. (I didn't say to IGNORE it, I said
doesn't need to live and die by it)
W-ball didn't say he had an SRR of 10, he said that he could control 1 in 10 throws. If
you'd like to see ME control 1 in 10 throws (at least), feel free to sign up for the
upcoming class. Of course that still wouldn't qualify, as the sample size would be
inadequate for your purposes.
TheSwan
sorry irish, i didn't mean to come off like an angry new
poster. i actually think your methods might work, and at least your site entertains my
fancies. it's always good to question your instructors, n'est-ce pas?
I completely agree that a single shooter in five years of moderate rolling could never
throw enough to provide a significantly large sample of rolls. But I'm still wondering
how, exactly, you would show me that you're controlling the dice in ten throws without
showing me you are controlling 100,000 out of 1,000,000. Smoke and mirrors, I tell you! A
fancy show is not enough to convince The Swan and his Cygnets!
InsideIke
Hey Swanie,
The law of large numbers discussion seems pointless to me. The casinos, for the most part,
don't get too wacked about dicesetting because they live by the law of large numbers as
they are in the game 24 hours a day seven days a week for an infinite duration. The
statistical sample that represents a precision shooter's career is negligible to the
casinos, but everything to the shooter. Now, if you and your f***ing cygnets were all at
the same table 24/7, I am sure the casinos would be disturbed for more than one reason.
NC
Disingenuous, no I think we are really way off the beaten
path. This is mathematic proof ( created by the PARR group I think ) that precision
shooting works, and yes in the long run ( at least 43 throws ), I don't know where the
comparison to large numbers or what they are actually comes from. This was more of a
convincing proof if your interested in numbers, that precision shooting or dice control
worked.
This was a selling point, not a technical point I believe. If you controlled 1 in 43 of
your throws you have taken off the house edge. This does not mean you can incorporate this
math into your playing, but could possibly attract someone else to try dice setting if
they could understand the math. All the mathematics proved was that controlling 1 in 43
brought your SRR above 6 therefore improving your play. That states that 1 in 43 merely
proves ( an SRR a fraction above 6) and that the higher the SRR the more beneficial your
play. Not to say that it would not take more than a high SRR to make money, form most of
us it would. 1 in 43 is not a highly technical discussion. Not that we are highly
technical.
NC
Disingenuous, no I think we are really way off the beaten
path. This is mathematic proof ( created by the PARR group I think ) that precision
shooting works, and yes in the long run ( at least 43 throws ), I don't know where the
comparison to large numbers or what they are actually comes from. This was more of a
convincing proof if your interested in numbers, that precision shooting or dice control
worked.
This was a selling point, not a technical point I believe. If you controlled 1 in 43 of
your throws you have taken off the house edge. This does not mean you can incorporate this
math into your playing, but could possibly attract someone else to try dice setting if
they could understand the math. All the mathematics proved was that controlling 1 in 43
brought your SRR above 6 therefore improving your play. That states that 1 in 43 merely
proves ( an SRR a fraction above 6) and that the higher the SRR the more beneficial your
play. Not to say that it would not take more than a high SRR to make money, form most of
us it would. 1 in 43 is not a highly technical discussion. Not that we are highly
technical.
irishsetter
Swan,
My irony must have been missed. The point is, be it at one of our seminars, or
wreckingball showing up at windsor, there will be no way to "prove" one
controlled throw in ten, as the sample size will be too small.
I have no desire to "prove" anything in fact. I'm not a dice influencing
missionary, I need no dice groupies. Come to the site, learn dice influencing
techniques...or not is my opinion. Whatever floats your boat.
However, if you decide that you'd like to take on an experiment, let me know. Here's one
that was proposed on another board was to take 20 precision shooters and 20 random rollers
and have them each throw a relevant sample (say 40,000) of throws, then analyze the data.
At 4 hours a day, 100 throws an hour, that'd take about 100 days for each shooter. In the
name of research, I'd discount my hourly fee to $25/hour...plus per diem of course. (i'd
have to take time from work to perform such an important experiment) Let me know when it's
all set up.
TheSwan
There's no dispute among any of us that 1 out of 43
controlled dice (i.e. no 7) reduces the house edge to zero if you're placing the 6 or 8. I
will temporarily lay my doubt to rest as to how this should or could be proved
empirically. For the moment, I will remain crapnostic. I hope to make more friends here
this way.
I am more concerned with the application of test data and optimizing play given a player's
skill set. If we continue along the lines of controlling our rolls such that a 7 doesn't
come up, at a certain level of skill it no longer makes sense to place a bet on the 6 or
8. In the simplest form of right betting, placing the 4 or 10 replaces the 6 or 8 as the
optimal bet at around 13-17% controlled SRR. (I have the actual percentage at home, but it
is figured no differently than Irish's reply to my original post.)
I'd rather not open up a can of worms regarding betting strategies. I'm talking purely on
the simplest terms; if you guys are dice-setting monsters (I am merely waterfowl), your
bets might include more 4 and 10 place bets.
I'd really like to know what kind of control you think you have! Are we talking 1 out of
43, thanks for the free room? Or are we talking 1 out of 6, thanks for financing my new
H2? Or are we talking, 'Shut up, The Swan, nobody can ever know their percentage
control.'? I really have no idea whom I'm dealing with. I'm just the graceful bird who
shoots crap.
heavy
The one in forty-three number has been bandied about for
years. I've tossed it out myself on occasion and I think it's good enough.
I make no claims to be a math whiz - and my calculations may well be off - but here's how
I THINK it shakes out when you take an average thirty-six number roll distribution and
play with that a bit. Let's say instead of six sevens you roll five. And instead of five
sixes you roll six. What's that do to your payout schedule? Well, if you can do it
consistently you have an advantage on the eight place bet because the house pays you 7-6
while the "adjusted" sharpshooter odds would be even money. And you have an even
larger edge on the six, obviously, because they're paying you 7-6 when the adjusted
sharpshooter odds should be 5-6.
Well, they're paying you 7-5 on the place bet on the five and nine - and the
"adjusted" sharpshooter odds are 4-5. So you actually have an advantage there as
well.
My point is - you don't have to control one in forty three to influence the odds of the
game. It's actually much less IF you KNOW where the numbers are going to fall. Personally,
I think the sevens to rolls ratio number is more important than the one-in-forty-three
number figure we toss around. After all - what if the one-in-forty three number you
control is a result of tossing one less ace-duce and one more yo. Doesn't do you a hell of
a lot of good. Or does it?
Better still, perhaps we should track sevens to inside numbers rolled as our friend the
Dice Doc suggested awhile back. That has even more value to most of us.
I digress.
With an SRR of 1:7 in a double odds game you'll have an advantage of around 8% on the pass
line. I won't go through the math - SS did that in his book and, since he's the engineer
and I'm not - I'll go with that.
In that same 2X odds game with an SRR of 1:8 your edge over the house rises to around 16%.
At 1:9 it's almost 23%.
At 1:10 it's over 28%.
In Biloxi last weekend my SRR was a tad over 1:16 with MANY hands tossed. That's almost
double what my LONG TERM SRR runs in practice sessions - just under 1:9 which is still
quite respectable.
But the fact is I don't make a whole lot of my money on the pass line. I make MOST of my
money at place betting and prop betting. In one session last weekend I took down almost
$600 on pressed prop action when I threw six consecutive horn numbers (and bet on five our
of six of them).
I can capitalize on that because I track my signature trends with various dice sets and
bet accordingly. And unless you are willing to invest the time and effort to do so over
the long haul you'll struggle - at best - with precision shooting - and quite possibly
give up on it entirely.
Should your betting strategy include the four and ten? Absolutely - if the shooter is
pre-setting the V-2 and the four and ten are among his/her signature numbers. I can assure
you that when Irish, Site, Roadrunner, or any of a dozen other players from the boards I
step up to the tables with from time to time are setting the V-2, I'm all over the four
and ten. But this is not determined by the number of SEVENS they do or don't throw - it is
determined by the number of fours and tens they throw.
As for whether you're shooting for a room comp - or a car payment - I consider anything
that I'd ordinarily spend money on that I don't HAVE to spend money on - e.g. a hotel room
- a "win." But what I really like is coming home from the casino and making a
deposit in the bank. There's really no reason why you can't do both.
aparadim
Removed at the request of the poster.
TheSwan
I was afraid someone was going to tell The Swan to spend
money on practicing craps. I was hoping to join this message board to get input on
standard deviation, skew and kertosis of a few simple strategies. Aparadim, I read
most of your articles, and I appreicate your objectivity towards all subjects. You are
very honest and you give credit where its due. Correct me if I'm mistaken, but
Sharpshooter's proof that started this whole topic off is in fact undisputable. I
mean, it's a proof!
The Swan likes this message board. I hope that in my next post, I can offer more than my
disputes about disputability. Oh, and The Swan likes risks, so perhaps he has more friends
than are coming forward on this site. I'd like to end this post- I'm embarrassed to read
that I even thought 1/241 could get you to even-odds.
NC
TheSwan,
Debunking precision shooting, fact or legend? I don't blame you one bit for questioning
these things. Any good skeptic would, like water dowsing, and alien life forms there is
little scientific evidence to prove it does work or exist. No scientific tests in casino
conditions have ever been preformed by an objective 3rd party. Although we hear stories of
some dice thrower actually using dice control to there advantage they are almost always
hidden behind a book or seminar at some cost. We do discuss these things openly and freely
on this web site thanks to Irishsetter, but at best precision shooting depends on
theoretical math, and statistics developed by craps players throwing on practice boxes in
their own basements and garages. If we waited for a real scientific test to be run on how
to control the random nature of the dice, we could be waiting the rest of our lives with
no chance to try it live in casino action. Therefore we have taken it upon ourselves to
prove dice control's validity ourselves. At the rate that most people lose in casino, it
can't hurt to give it a try. By no means have we proven it to the world!
I do think I see and alien creature or two in the men's room of the local casino from one
time or the other.
aparadim
Removed at the request of the poster.
Dylanfreak
Maybe the question about 1 out of 43 rolls is which is the
number to place a bet on so that you can take advantage of the influence of the dice for
that one number??
aparadim
Removed at the request of the poster.
Dylanfreak
Aparadim is as slick as they come ---Whatever he says
believe---whenever you happen upon a Precision shooter at the table , at the least , you
must recognize the set he is using and then since you may not have laid eyes on him
before, you must pay attention to what numbers he has wagers on---by doing that you can
get close to what his signature numbers are----But at all times you must proceed with
caution and qualify any shooter through a five count, making a pass, tossing repeat
numbers----anything which prevents a loss from jumping in a on a shooter, random or
precision, within a roll or two-----Just my thoughts
Click Here
to Go Back to the Message Board Archive Table of Contents