luck or math...

Discussion in 'General Craps Discussion' started by basicstrategy777, Jul 16, 2010.

  1. basicstrategy777, Jul 16, 2010

    basicstrategy777

    Joined:
    Jul 4, 2010
    Messages:
    7,894
    Likes Received:
    6,044
    Gender:
    Male
    Location:
    CT.
    If you polled 100 knowledgeable math people who were honest, and loved and played craps alot, do you think most would say
    the reason they won during a 2 hour session is because.... A) they were lucky...............or B) their knowledge of the math of the game resulted in them making the lowest vig bets and most likely outcome bets.


    777
     
    #1
    yacraps likes this.
  2. The Midnight Skulker, Jul 17, 2010

    The Midnight Skulker

    Joined:
    Jan 28, 2010
    Messages:
    3,386
    Likes Received:
    2,692
    Gender:
    Male
    Location:
    Idaho, USA
    You did say [em]knowledgeable math[/em] people, right? I would then expect the vast majority of them to list positive variance (aka luck) as the primary reason for their success, and their choice of low vig, high win probability bets as the reason the positive variance they had was able to overcome the house advantage. Of course they might also cite their choice of a betting strategy that was well suited to the events that occurred (e.g. a winning progression that took advantage of a big hand; that one time in over six years, on average, that Eddie's Hardway Parlay came in).
     
    #2
    yacraps likes this.
  3. basicstrategy777, Jul 17, 2010

    basicstrategy777

    Joined:
    Jul 4, 2010
    Messages:
    7,894
    Likes Received:
    6,044
    Gender:
    Male
    Location:
    CT.
    You must pick A or B. You picked both. Grit your teeth and pick one.

    All positive variance is, is deviation to the upside. You say luck is positive variance, but luck is more than that.
    Positive variance has nothing to do with......when you walk up the the tabe....... when you stop playing.......
    when you switch tables.....when the table is loaded with your place bets and you call your bets off and the next roll is a seven........when you have a hunch and bet a bunch and you win.........plus, plus.

    I don't need alot of positve variance betting the center of the table to make more money than you....while you are betting the mathematically correct, low vig, high probability, 6 and 8.

    Math people feel comfortable in what they say because what they say is probably correct and provable ( although throughout history the math folks made a few mistakes and had to reverse course ). The problem is, although they are necessarily are correct, in a 2 hour session, following the math does necessarily mean you will win. There are many factors that go into winning. The math of the game is one of them but I don't think the math is the dominant factor in winning in a 2 hour session. IMHO you have to be lucky.

    777

    post script: I enjoy your writings.
     
    #3
    yacraps likes this.
  4. goatcabin, Jul 17, 2010

    goatcabin

    goatcabin Member

    Joined:
    Feb 3, 2010
    Messages:
    857
    Likes Received:
    63
    No, because the correct answer is both. The probability of overcoming the house edge is a function of the ratio of the expected value/standard deviation - the higher the standard deviation relative to the expected value (loss), the less luck (positive variation) is needed to come out ahead. So, the math tells you how lucky you have to be to come out ahead, and the dice determine whether you do or not.
    Cheers,
    Alan Shank
     
    #4
  5. basicstrategy777, Jul 17, 2010

    basicstrategy777

    Joined:
    Jul 4, 2010
    Messages:
    7,894
    Likes Received:
    6,044
    Gender:
    Male
    Location:
    CT.
    So the chance of me winning can be determined by plugging in the numbers in a math formula. Different numbers yield different answers........but ultimately, the fall of the dice will determine if I will win.

    The math can predict but can't give me the formula for success. It appears, the dice have the final say and not the math.


    777
     
    #5
    yacraps likes this.
  6. goatcabin, Jul 17, 2010

    goatcabin

    goatcabin Member

    Joined:
    Feb 3, 2010
    Messages:
    857
    Likes Received:
    63
    Very true; the dice have the "final" say. But the ramifications of what they "say" are determined by the particular betting strategy. For example, if the dice "say" you come out one standard deviation better than the expected value, that may or may not be enough to overcome the house advantage.

    No time now, but I will illustrate what I mean with an example.
    Cheers,
    Alan Shank
     
    #6
  7. goatcabin, Jul 18, 2010

    goatcabin

    goatcabin Member

    Joined:
    Feb 3, 2010
    Messages:
    857
    Likes Received:
    63
    Contrary to what BasicStrategy777 says, the math does not "predict". One can use math or simulation to describe the distribution of possible outcomes for a given strategy. For example, if you set up a particular strategy using a WinCraps auto-bet file and run 10,000, 20,000, 30,000 sessions, millions of dice rolls, the program keeps track of several parameters of each session and can draw you a graph of, say, all the net outcomes. This gives you the same kind of information as knowing the dice probabilities, like:

    probability of rolling a seven: = .1667 = probability of busting a $200 bankroll before two hours
    probability of rolling a twelve = .0278 = probability of winning more than double your bankroll
    etc.,etc.

    We all start with a "handicap", the expected loss for a session, which is always edge * action. In order to overcome that handicap, we have to experience positive variance equal to or more than that handicap. For a given handicap (expected loss), the size of the standard deviation of our possible session outcomes determines what degree of positive variance is required to equal or exceed the handicap. Here are three different bets, 60 resolutions each, with roughly the same bet handles:

    Code:
    bet                      ev      SD ev/SD
    Any 7   for $11    -$110.00 $158.77 .6928
    Boxcars for $11    -$ 91.67 $434.07 .2112
    $5 pass, double odds -$4.24 $110.66 .0383
    
    In each case, the probability of coming out one standard deviation (or two, or three, etc.) better (or worse) than expectation is the same. For the Any 7, I have to come out about .7 of a standard deviation better than the expectation in order to break even; for Boxcars, I have to come out about .2 SD better; whereas, for the passline with double odds, I only have to exceed the expectation by about .04 of a standard deviation. Those numbers can be translated into probabilities, so we find that the "Big Red" bettor has about a 25% chance of breaking even or better for 60 bets, while the Boxcar bettor has about a 42% chance and the pass/odds bettor about a 48% chance.

    You can think of it this way (thanks to "The Midnight Skulker" for this analogy):
    The situation for the "Big Red" bettor is like drawing from a bin of green and red ping-pong balls, 75% of which are red (losing sessions) and 25% are green (my favorite color, hence representing winning sessions). The Boxcars bettor is drawing from a bin with 42% green and 58% red balls, while the pass/odds bettor's bin has 48% green, 52% red.

    This doesn't guarantee that the passline bettor will do better in any given session than either of the other bettors. In other words, it doesn't predict any specific result, any more than knowing the dice probabilities predicts the next roll.

    Let's examine the outcomes for each of these bettors, given differing degrees of good and bad luck:

    Code:
    bettor     -3 SD   -2 SD  -1 SD    ev  +1 SD   +2 SD   +3 SD
    Any 7      -$586   -$428  -$259 -$110   +$49   +$208   +$366
    Boxcars   -$1394   -$960  -$526 -$ 92  +$342   +$776  +$1211
    Pass/odds  -$336   -$226  -$115 -$  4  +$106   +$217   +$328
    
    You can see that there's a point in between the ev and +1 SD where the Boxcars passes the pass/odds; it's at +.27 SD, which corresponds to a probability of .39; however, on "the other side of the tracks" that high variance is killing Mr. Boxcars. Even the Any 7 bettor gets ahead of Mr. Passline up over +2 SD (2.20), the probability of which is .0139.

    So, the bets that one makes are a very significant component in determining one's chances; the dice, however, ultimately pick one ping-pong ball out of the bin, so to speak.
    Cheers,
    Alan Shank
     
    #7
  8. basicstrategy777, Jul 19, 2010

    basicstrategy777

    Joined:
    Jul 4, 2010
    Messages:
    7,894
    Likes Received:
    6,044
    Gender:
    Male
    Location:
    CT.
    Thanks for the explanation Alan.

    To me, the math predicts the chance of an event happening; however, it cannot predict something with certainty.

    Your last paragraph pretty much sums it up.

    I guess you side with Damon Runyon when he said..." The race is not always to the swift, nor the battle to the strong---but that's the way to bet."

    777
     
    #8
    yacraps likes this.
  9. goatcabin, Jul 19, 2010

    goatcabin

    goatcabin Member

    Joined:
    Feb 3, 2010
    Messages:
    857
    Likes Received:
    63
    I still don't agree with your use of the word "predict", which to me means "foretell". The math "describes" the probability of an event happening. When I say, "The probability of a seven showing on the next roll is 1/6", that's not a prediction, because it cannot be proven right or wrong by rolling the dice. We know it's true, or at least we very strongly believe it's true, because we're told the dice are machined very carefully so that each side has an equal chance of coming up, and we can count combinations that add up to seven.
    Cheers,
    Alan Shank
     
    #9

  10. arrgy

    arrgy Member

    Joined:
    Dec 12, 2009
    Messages:
    45
    Likes Received:
    0
    A math person would say you got lucky.

    I would say (for me anyway) I am skilled. In other words I very rarely win because of luck its because of skill.

    I would say (for a LOT of other people) they got lucky.
     
    #10
  11. basicstrategy777, Jul 21, 2010

    basicstrategy777

    Joined:
    Jul 4, 2010
    Messages:
    7,894
    Likes Received:
    6,044
    Gender:
    Male
    Location:
    CT.
    I never knew that craps was a game of skill. I always thought there was no strategy that could beat the game, hence, some luck was required.

    777
     
    #11
    yacraps likes this.

  12. arrgy

    arrgy Member

    Joined:
    Dec 12, 2009
    Messages:
    45
    Likes Received:
    0
    When i mean skill, I mean more of discipline, money management, not chasing, etc. That takes skill.
     
    #12
  13. rayfromtheway, Jul 22, 2010

    rayfromtheway

    rayfromtheway Member

    Joined:
    Jul 18, 2010
    Messages:
    1
    Likes Received:
    0
    Alan (goatcabin, or anyone who could answer),

    a question for you please:

    am using "my craps game" computer simulator and it records/tallies "points made" and "seven outs". the question is: disregarding the come out roll of seven or craps, what is the house edge for points made to seven outs?

    crude probability figuring by me, says 18.78788%. that seems awful high, or am i WAY off base?

    i guess what i'm asking is what is the ratio of " points made/seven outs " ?

    thanks and maybe take it easy on me (newbie).

    respectfully,
    ray
     
    #13

  14. arrgy

    arrgy Member

    Joined:
    Dec 12, 2009
    Messages:
    45
    Likes Received:
    0
    For 6,8 six ways to win and five ways to lose gives the house (or don't player) 9.091 edge
    For 5,9 six ways to win and four ways to lose gives the house 20.00 edge
    For 4,10 six ways to win and three ways to lose gives the house 33.333 edge
     
    #14
  15. goatcabin, Jul 22, 2010

    goatcabin

    goatcabin Member

    Joined:
    Feb 3, 2010
    Messages:
    857
    Likes Received:
    63
    That figure is, indeed, the weighted average house edge on the pass/come after a point has been established. If it seems high, keep in mind that on the comeout roll the player has a 33.3% advantage. Since the bet is resolved on the comeout 1/3 of the time, you get:

    Code:
    1/3 *  .333   =  .111111
    2/3 * -.1878 = -.125252
    ----------------------------
    ------------       -.01414
    
    "It all comes out in the wash!"
    Cheers,
    Alan Shank
     
    #15
  16. basicstrategy777, Jul 22, 2010

    basicstrategy777

    Joined:
    Jul 4, 2010
    Messages:
    7,894
    Likes Received:
    6,044
    Gender:
    Male
    Location:
    CT.
    Without the comeout roll, the odds are 3 to 2 against the Pass line.

    The important thing to remember is there is absolutely no difference among the points or Do or Don't.


    777
     
    #16
    yacraps likes this.
  17. The Midnight Skulker, Jul 22, 2010

    The Midnight Skulker

    Joined:
    Jan 28, 2010
    Messages:
    3,386
    Likes Received:
    2,692
    Gender:
    Male
    Location:
    Idaho, USA
    I thought I did: "I would ... expect the vast majority of [knowledgeable math people] to list positive variance (aka luck) as the primary reason for their success, and their choice of low vig, high win probability bets as the reason the positive variance they had was able to overcome the house advantage." (Emphasis added) Since there is no bet or combination of bets that offer the player a positive expectation, the only way a player can win is to experience a positive deviation from that expectation. By making low vig, high win probability bets the player minimizes how much positive deviation he/she needs to win, but some positive deviation will always be necessary.
    Good luck is positive variance, but one can also have bad luck, which will result in losses greater than expected. It's a two-way street, but in a gambling context luck is variance, nothing more.
    Correct, except for that last one. When you win a big hunch bet you have been lucky.
    Oh? As Alan demonstrated, if we start with equal bankrolls, and you make 60 bets on 12 while I make 60 Pass Line bets with double odds, you are expected to end up with more money than me less than two times out of five.
    Following the math gives you the best chance to win, but you still need to get lucky in order to actually win. For a given betting strategy over a given period of time, the math tells you how lucky you have to get. The dice determine whether you attain that level of luck.
    And our encounter in the anticeptic [sic] lab is one of my favorites from the Before Time. You up for another game of Russian Roulette?


    (Edited 25 Apr 2017 to substitute actual underlining for the HTML tags that did not translate properly when the forum software was upgraded.)
     
    #17
    Last edited: Apr 25, 2017
    yacraps likes this.
  18. goatcabin, Jul 22, 2010

    goatcabin

    goatcabin Member

    Joined:
    Feb 3, 2010
    Messages:
    857
    Likes Received:
    63
    Huh? That's completely wrong. The 4/10 have the lowest win probability (1/3), then the 5/9 (2/5) , then the 6/8 (5/11). Since the payoffs are all even money, those differences are substantial. If you take odds, then the true payoffs make those bets equal in expectation (0).

    As to the do vs. don't, the difference amount to just one unit out of 1980 - trivial.
    Cheers,
    Alan Shank
     
    #18
  19. basicstrategy777, Jul 22, 2010

    basicstrategy777

    Joined:
    Jul 4, 2010
    Messages:
    7,894
    Likes Received:
    6,044
    Gender:
    Male
    Location:
    CT.
    If one side (Do/Don't) had an advantage over the other side, people would all play the side that had the advantage.

    If one box number had an advantage over another box number, people would play ( take/lay odds ) on the box number with the advantage.

    Sorry you didn't realize what I was saying......I thought you would understand.


    777
     
    #19
    yacraps likes this.
  20. goatcabin, Jul 22, 2010

    goatcabin

    goatcabin Member

    Joined:
    Feb 3, 2010
    Messages:
    857
    Likes Received:
    63
    Well, as I wrote, the advantage is miniscule, and most people avoid the "darkside" for social reasons.

    Since the thread was about the difference between the player advantage on the comeout and the house advantage (on the flat bet) on the points, I addressed that. I also pointed out that, for the odds, the payoffs equalize the different win probabilities. It's ONLY on the odds bets that the point numbers are equal, not for the flat pass/come/DP/DC or for place/buy/lay bets.
    Cheers,
    Alan Shank
     
    #20