Variance

Discussion in 'General Craps Discussion' started by KokomoJoe4, Jan 12, 2020.

  1. KokomoJoe4, Jan 12, 2020

    KokomoJoe4

    KokomoJoe4 Member

    Joined:
    Sep 28, 2013
    Messages:
    6,013
    Likes Received:
    5,131
    Dice are designed with six faces, each face showing one to six spots, arranged in such a way that opposing sides spots will add to seven.

    Assuming fairly weighted face surfaces, each of the six faces have an equally likely chance (1/6) to show on top at outcome. Since two dice are used to play craps, summed dice outcomes can range between 2 and 12.

    The table of 36 shows the 36 unique ways top dice outcomes read. It shows that the 2 and the 12 can be made only ONE way, the 3 and 11 TWO ways, the 4 and 10 THREE ways, the 5 and 9 FOUR ways, the 6 and 8 FIVE ways, and the seven, the only number without a sister, SIX ways.

    There will be variability with outcome results. The ONLY thing that is KNOWN about an outcome result BEFORE it occurs is that it will be a number between 2 and 12. The table of 36 shows us what to EXPECT, based on probability. Probability does NOT tell us that these expectations are absolute. Rather, it tells us that seven is the most likely result, and 2 and 12 are least likely, and that the numbers SHOULD appear as predicted.

    Shooters will perform differently when shooting. The same shooter will sometimes perform brilliantly, and other times shoot terribly.

    Different shooters perform differently because of variance associated with a random process. The same shooter will perform well one time and poorly another time for the same reason.

    Math can be used to rate any shooters performance. If results for a particular outcome are better than expected from probability, we have a shooter showing a mathematical advantage.-----> Note that this is NOT impossible, it can and likely does happen.

    However, and this is where the typical DI gets off track, this advantage applies ONLY to the specific wager in question, AND this "mathematical advantage" is subject to change with the input of additional results.
     
    #1
  2. Dave G Ct, Jan 12, 2020

    Dave G Ct

    Dave G Ct Member

    Joined:
    Mar 15, 2015
    Messages:
    7,588
    Likes Received:
    1,162
    Gender:
    Male
    Yes the DI can profit from hitting the number his set implies he has an advantage.But there are DI ers who just want to avoid the 7 and will bet inside or 6&8 and they have no idea what will roll.
     
    #2
  3. Mssthis1, Jan 12, 2020

    Mssthis1

    Mssthis1 Member

    Joined:
    Jul 25, 2016
    Messages:
    3,242
    Likes Received:
    5,154
    When I started considering every result as a 1 in 36 chance of occurring it changed my outlook on mistaking variance as influence in the game. Lets look at the 6 as an example.

    There's 5 ways to toss a 6. 3/3, 5/1, 1/5, 4/2, and 2/4. If you claim to toss more 6's than random you're claiming you can control 10 dice face outcomes. At home on the practice table I feel I have a large enough sample to claim I toss slightly more of 2 face results than random. That's the two dice faces facing the back wall. So in this example I put 3/3 on those faces since I almost always use a hardway set. If I toss a 5/1, 1/5, 4/2, or 2/4 that is not a controlled result in my book but I'll still let then pay me for the bet;)

    Messing around in Wincraps you can see that it isn't that unusual for 200+ tosses to occur with out a 2/2, 6/6, 3/3 etc. It also isn't that unusual to see the same result occur several times in a short sequence of tosses. When you realize that is true for all 36 of the different combinations, even 10's of thousands of rolls probably isn't a big enough sample to diffidently claim influence or biased dice if you're in that fan club.

    Tracking bubbles was an eye opener for me. The Azure at Prairie Meadows still has an SRR of over 6.3 on a 17,000 plus roll sample. The only variable is when I play since I obviously can't track rolls if I'm not there.
     
    #3
    HornHiYo, yacraps and TDVegas like this.
  4. TDVegas, Jan 12, 2020

    TDVegas

    TDVegas Member

    Joined:
    Jan 4, 2014
    Messages:
    17,970
    Likes Received:
    12,339
    Wow....that certainly suggests something could be wrong. 20,000 roll simulations and best was 6.02 to the high side and 5.82 to the low side. The 5.82 would be an almost equivalent to 6.2 in reverse...but a full point higher (6.3) could definitely suggest a malfunction with that many rolls counted....or you are that good;)

    The odd thing is 20 out of 22 came in under 6.0. Strange.

    20,000 ROLLS -- 22 TRIALS
    SRR

    5.82**
    5.82
    5.98
    5.9
    5.85
    5.88
    5.88
    5.88
    6.02*
    5.91
    5.97
    5.92
    5.95
    5.93
    5.92
    5.92
    5.99
    5.93
    5.86
    5.92
    6.01
    5.93

    *record high
    **record low
     
    #4
    Last edited: Jan 12, 2020
  5. TDVegas, Jan 12, 2020

    TDVegas

    TDVegas Member

    Joined:
    Jan 4, 2014
    Messages:
    17,970
    Likes Received:
    12,339
    I'm leaning that way as well. It may come down to a "personal" belief trumps the idea you need a LOT of data to confirm DI. Hell, "money won" might be that personal belief.

    I've seen variance enough in short spurts to say..."wow", many times.
     
    #5
    yacraps and Mssthis1 like this.
  6. TDVegas, Jan 12, 2020

    TDVegas

    TDVegas Member

    Joined:
    Jan 4, 2014
    Messages:
    17,970
    Likes Received:
    12,339
    I'm willing to say no one is going to come up with random roller expectation over 180 or 360 rolls or higher....random chucker or DI. We will ALL have skewed probability. Obviously I could be wrong and 1 person hits random probability "on the nose"....but I highly doubt it.

    Look at the 20,000 roll trials. Not 1 of them hit SRR exactly at 6.0...and that's 20k rolls. If you reduce the rolls down to 180 or 360....the variance is going to sky rocket for ANY shooter.
     
    #6
  7. von duck, Jan 12, 2020

    von duck

    von duck Member

    Joined:
    Feb 10, 2017
    Messages:
    11,561
    Likes Received:
    2,792
    Gender:
    Male
    No!, the "variance" does not sky rocket, when you decrease the number of trials. THE PERCENTAGE OF THE VARIANCE SKYROCKETS. :cool:
     
    #7
  8. twodicebilly, Jan 12, 2020

    twodicebilly

    twodicebilly Member

    Joined:
    May 9, 2016
    Messages:
    4,328
    Likes Received:
    1,523
    Gender:
    Male
    TDv

    what player was that , that you got that data from, was it Rick, or Bob or Jerry
    or even me ?????

    If the data did not come from a player......it means nothing in terms of making a statement
    that in the long run only a certain thing can happy.

    Until you come to grips with the notion that players play in the short run nothing will ever make sense.
    GTC indicates to "all" new guys and that is hundreds, that you only bet the 6 & 8 and pass line
    with what ever odds you can afford... those are the bets with the lowest house advantage...when did
    you ever see one win like that …..consistently.... cant be done.

    If variance can happen by accident and it means nothing in terms of the normal rations of numbers
    that are supposed to come up, then variance created by a guy that plays a couple of times a month
    has no affect. Both are possible.

    TDB
     
    #8
  9. von duck, Jan 12, 2020

    von duck

    von duck Member

    Joined:
    Feb 10, 2017
    Messages:
    11,561
    Likes Received:
    2,792
    Gender:
    Male
    Duh! Strange.
     
    #9
  10. TDVegas, Jan 12, 2020

    TDVegas

    TDVegas Member

    Joined:
    Jan 4, 2014
    Messages:
    17,970
    Likes Received:
    12,339
    Well, Tomp offered you the chance. You felt it was like being a dancing monkey and refused. So, end of trial offer. Tomp was forced to go about other methods to get some data. Namely, computer sims.

    What do you think makes up long term average? Short term results added up.

    This is why after 20,000 rolls, imo, your SRR (or anyone's) will likely average out between 5.7 to 6.3....Give or take.

    You will have hands where the SRR is >8. You will also have hands where it is <4 and everything in between.

    We can agree to disagree about where a player will stand long term, as this cannot and will never never be answered on a craps forum.

    How would you like to provide verifiable or confirmable roll data here?
     
    #10
    Last edited: Jan 12, 2020

  11. FredP

    FredP Member

    Joined:
    Dec 4, 2011
    Messages:
    2,681
    Likes Received:
    1,016
    Gender:
    Male
    Location:
    Northern NJ, Clinton area
    Please define the "percentage of the variance".
     
    #11
    yacraps likes this.
  12. von duck, Jan 12, 2020

    von duck

    von duck Member

    Joined:
    Feb 10, 2017
    Messages:
    11,561
    Likes Received:
    2,792
    Gender:
    Male
    For simplicity, let's use a coin flip, to make this demonstration.
    If you flip a coin 4 times, and there are 3 heads and one tail, then you would have received 150% of expectation, on heads, and 50% of expectation on tails, with a difference of 2 flips, 3/1. If you continue to flip until you have 100 flips, and the difference is still 2 flips 51/49, then the variation would be the same, one more than expected, and one less than expected, but the percentage of expectation, is now only off by two percentage points for each face, 51/49 vs 50/50 expectation. :cool: Ma'am.
     
    #12
    Last edited: Jan 12, 2020
  13. yacraps, Jan 12, 2020

    yacraps

    yacraps Member

    Joined:
    Sep 24, 2011
    Messages:
    4,405
    Likes Received:
    4,744
    Gender:
    Male
    Location:
    san diego
    Great point and post Koko...some golden info here
     
    #13
  14. twodicebilly, Jan 12, 2020

    twodicebilly

    twodicebilly Member

    Joined:
    May 9, 2016
    Messages:
    4,328
    Likes Received:
    1,523
    Gender:
    Male
    TDV


    I have many times, you just did not like it.

    After seeing me post for a very long time that I think the table you play on makes
    a big difference, you cant possibly think I would allow you or anyone pick a table
    I am supposed to play on.



    TDB
     
    #14
    eagleeye2 likes this.
  15. KokomoJoe4, Jan 12, 2020

    KokomoJoe4

    KokomoJoe4 Member

    Joined:
    Sep 28, 2013
    Messages:
    6,013
    Likes Received:
    5,131
    Other than a result between 2 and 12, nobody has any idea what they will roll. (Except possibly LID and the UCDT).

    The DI can profit if he is lucky, as can the random roller.

    You are absolutely on the wrong track regarding "hitting numbers his set implies" Dave, because dice ABSOLUTELY do NOT stay on axis. MAYBE this DI rolls 4's or 10's more frequently than 3/36 of the time. Maybe he shoots out of the 2V, but 4's and 10's can and do appear in every axial array - HW, H6, V6, X6, 3V and 2V. Because this is true, there is more to the dice than the table of 36, but that is another matter.

    IF the DI always bets one of the outside numbers, and IF he does better than making it more than expected, he will be ahead of the game with respect to that number. Any other bet he makes while shooting that doesn't exceed expectation drags him down.
     
    #15
    yacraps likes this.
  16. TDVegas, Jan 12, 2020

    TDVegas

    TDVegas Member

    Joined:
    Jan 4, 2014
    Messages:
    17,970
    Likes Received:
    12,339
    I understand you want to be believed without question or doubt....but on a subject about a person able to skew dice results out of random, on a 500 year old game...I'm afraid it's going to need a bit more than "here are my numbers, believe what it means".

    If I was a DI, my own data would be subject to the same scrutiny, doubt or need for 3rd party confirmation....especially if i claimed it indicated something out of the ordinary. If much more than ordinary (random)....most definitely confirmation or validation needs to be a part of the conversation. For me, you or anyone, imo.

    This takes NOTHING away from what you believe. Only what can or cannot be proven here....which is nothing. Me included.
    The bohunk put no restriction on what table could be used. It was a matter of convenience for both parties and if their schedules could meet. Any table was generally acceptable as long as it conformed to the basics..felt, backing, size.
     
    #16
    Last edited: Jan 12, 2020
    yacraps likes this.
  17. twodicebilly, Jan 12, 2020

    twodicebilly

    twodicebilly Member

    Joined:
    May 9, 2016
    Messages:
    4,328
    Likes Received:
    1,523
    Gender:
    Male
    tdv

    it should be clear by now what I think of you and your opinion, nothing has changed in two years,
    I cant understand why a guy that knows nothing talks so much about what he does not understand.

    SEE

    TDB
     
    #17
    eagleeye2 likes this.
  18. twodicebilly, Jan 12, 2020

    twodicebilly

    twodicebilly Member

    Joined:
    May 9, 2016
    Messages:
    4,328
    Likes Received:
    1,523
    Gender:
    Male
     
    #18
  19. twodicebilly, Jan 12, 2020

    twodicebilly

    twodicebilly Member

    Joined:
    May 9, 2016
    Messages:
    4,328
    Likes Received:
    1,523
    Gender:
    Male
    koko

    you said this...…..
    IF the DI always bets one of the outside numbers, and IF he does better than making it more than expected, he will be ahead of the game with respect to that number. Any other bet he makes while shooting that doesn't exceed expectation drags him down.[/QUOTE]

    you understand that is not a true statement for all people that work with the dice.....you know that...….nothing is certain, so when
    seeking to benefit from influence you are playing the percentages, what is likely to happen, if my aunt had balls she would be my
    uncle.. I am not now going to start thinking about what is that happened.....

    Most guys play differently, some like to see what their set is doing and try to change that set to improve their chances of getting
    certain numbers they may need for a fire bet or an ATS bet....your using the best information you have..makes sense to me.
    Now I play to extend rolls, the longer the roll the better chance I have of getting paid on numbers and getting 2-3 numbers with
    bets high enough to make some money, and with the little Joe shot their are a couple of things you can. But again nothing is
    certain, your trying to get yourself the best chance.

    You post against the possibility of DI all the time, yet you tend to quote some of the information of Linaway....did you ever ask
    your self what good is his information if you cant take advantage of that. That would be like going back to the idea that
    using the five count puts you on the better rollers.....all it does is decrease the amount you are betting.

    TDB

     
    #19
  20. basicstrategy777, Jan 13, 2020

    basicstrategy777

    Joined:
    Jul 4, 2010
    Messages:
    9,971
    Likes Received:
    9,550
    Gender:
    Male
    Location:
    CT.
    I thought the 5 count saved you from losing money on early 7 outs....is that not helpful ?........the longer you are at the table the closer you are to a long roll...which should result in you winning money........is that not true ?

    777
     
    #20
    eagleeye2 and yacraps like this.