Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Does Probability Exist?

  1. Dec 3, 2003 #1
    If a six-sided die is rolled 6 times, probability dictates that each number will appear once. Any number has a 1/6 probability of being rolled. But, the die is not necessarily subject to probability, because probability itself, in this case at least, is only a function of the known universal laws. The only things that determine which number is rolled are gravity, the speed and angle at which the die is thrown, the initial spin of the die, air resistance, surface it lands on, etc, not a purely random system we call probability.
    Likewise, the results of flipping a hundred pennies will eb based on the same system, and if all information is known, we can determine whether it will be heads or tails. This destroys probability as an actual function, but limits it to an excuse for our inability to know all these variables at all times. I've heard theories state that when a dice is rolled, all numbers are rolled, but at different probabilities. So all outcomes affect "everything", so to speak, due to causality. But, if a machine were constructed in a constant vacuum chamber with constant gravity, and this machine could launch a number of dice at the EXACT same angle, speed, spin, etc, we could theoretically determine whether or not the outcome will be 1, 2, 3, 4, 5, or 6.
    Is probability really just a way of explaining what we cant easilly determine with a simple equation, or am I missing something? If every event is determined on a set of laws, then wouldnt this mean that probability is just meaningless? How can all outcomes happen if only one is allowed by a constant set of laws? The only way probability can really exist is if universal laws act differently on every particle in the universe at different times, and even this would amke probability a function of laws, not events.
  2. jcsd
  3. Dec 4, 2003 #2
    This is incorrect. Therefore, I can safely ignore the rest of your post as it seems to be based on this "fact".
  4. Dec 4, 2003 #3
    Then please, instruct me as to the correct view of this situation, so that I may benefit from it
  5. Dec 4, 2003 #4


    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    If I rolled a die 1,000 times and got one every single time, what would you say?

    - Warren
  6. Dec 4, 2003 #5


    User Avatar
    Science Advisor

    Perhaps you need to rethink your moniker!
  7. Dec 5, 2003 #6
    I'd say something like: Wow, that was quite unlikely to happen (probability ~ 10-779)! From there I would conclude with more than reasonable confidence that the die was fixed...

    @HallsofIvy: what's a moniker?
    Last edited: Dec 5, 2003
  8. Dec 5, 2003 #7
    I think we are getting off track from SmarterThanGod's main question (paraphrasing): If everything that happens to a rolled die is a function (solely) of universal physical laws, then to what extent is "probability" really in play?

    I think this is a respectable question and deserving of a much more respectful dialog than the one currently underway. It has in fact been one asked, in various forms, many times on these fora. I have to say that I am in complete sympathy with SmarterThanGod on this one. This view (that the universe completely described by the myriad laws of nature - and nothing else) is Determinism.

    Determinism, as far as I know (being one myself), is seriously and credibly attacked only by certain investigators in the realm of quantum mechanics, where behaviors of subatomic "particles" are posited as being determined only "probabilistically" - without underlying, determining causation. I have been trying to follow the many threads on the physics forum related to this specific subject (admitting that so much of the math and thinking is over my head, being a rather simple biology guy) and have found at least small comfort with the realization that there are at least some professional investigators (in addition to several apparently well-informed writers on this forum) who are looking for underlying causation - not content to just take "ultimately probabilistic" for an answer.

    I personally hope such investigation proves fruitful. As I have said before, how anyone can believe in a universe where some things arise from no cause whatsoever and not go insane is beyond me. More importantly, I believe the present situation is somewhat unscientific in that the apparent majority of subatomic physicists are quite confident in asserting that there are no causative factors at the quantal level. It seems a bit like mysticism to me. (But again, what do I know. :))
  9. Dec 5, 2003 #8
    you know probability does not mean that things arise from no cause. Probability means you have more possible outcomes. like when you thow a dice you get as a result a cetain face upwards. Acausality means that when you throw a dice, the dice will transform into a red dragon that will burn your left hand....
  10. Dec 5, 2003 #9
    Would that it were so, but I have pressed this point many times to those accepting the majority view, and the claim is that at least some aspects of subatomic behavior are truly uncaused. I could live (more easily) with thinking along the lines you have just expressed, but I have found no relief.

    This dichotomoy of "belief" (ultimate causality vs ulitmate probablistic behavior) is substantial. Your view is, mercifully, less so.
  11. Dec 5, 2003 #10

    could you elaborate on that...
  12. Dec 5, 2003 #11
    I can offer some often stated examples of sub-atomic behaviors said to be truly "uncaused" (no hidden variables at work): the propensity of a particular atom for radioactive decay as vs another, the propensity of one particular atom to "tunnel" through a barrier as vs another; "collapses of wavefunctions" e.g. Heisenberg Uncertainty Principle.

    As I have indicated above, many folks infinitely more adept at this subject have trotted these out as examples. I myself would welcome any other interpretation or "admission" of a potential lack of sufficient knowledge as to the claims. I wish I had the time to add more detail and background to the examples I give here, but my time here is limited (I'm supposed to be working :)

    Maybe this can at least further this conversation.
  13. Dec 5, 2003 #12
    All of these are SEEMINGLY uncaused, but it is still possible that there are laws that dictate how and why these things happen, we just havent discovered them yet. I find it absurd that things happen according to chance, and I believe that everything is subject to universal laws. Things cant just happen for no reason at all, even if they seem to. There are unknown laws governing everything that happens, even laws that can predict quantum foam. I believe that once these laws are fully understood, probability will serve no important purpose at all. Any other determinists out there with similar views, besides Opiner? Im interested to hear everyones thoughts on this.
  14. Dec 5, 2003 #13


    User Avatar
    Science Advisor
    Homework Helper

    My understanding is that QM can be described as a system with 'hidden variables' as long as the variables are not 'nice' where nice relates to notions of measurability and probability.

    So, AFAIK, there are acceptable models for QM where the universe is deterministic with an unknown starting state instead of being non-deterministic. In effect, these systems trade one paradox for another. (E.G. nonlocality vs. the Branach-Tarski paradox)

    However, in those cases the measurements that are made are 'random' in the sense that their values are 'unknowable' before the measurement is made.
  15. Dec 5, 2003 #14
    Even with a completely deterministic model of physics that allowed you to determine the future based on the past, you can't get rid of probability. From a practical standpoint, we would still say that the probability of rolling 2 with a six-sided die is 1/6, because the amount of work in deterministically figuring out whether a 2 is going to be rolled would be astronomical.

    From a philosophical point of view, a completely deterministic theory might be satisifying. But without knowing everything about everything in the universe, you'll only be able to figure out the probability that your theory is correct - and it won't be 1. So you'll always be stuck with probability.
  16. Dec 5, 2003 #15
    This is true, but this was just an example i used to explain my point that probability is based on a deterministic set of rules.
  17. Dec 5, 2003 #16
    The point I was trying to make was the from a practical point of view, you can never eliminate probability since it's just too useful of a heuristic to give up. And from a philosophical point of view you can't eliminate it without discarding inductive logic. So this reduces to a question of what assumptions about the universe we are willing to make.
  18. Dec 12, 2003 #17
    I know this, but Im saying that probability will eventually be limited to a means of trying to predict what we are too lazy to figure out, such as how to roll a 2, or flipp 100 pennies heads-up.

    Ha are we going to go back and forth quoting each other and arguing that we dont see the other's point?
  19. Dec 17, 2003 #18
    First, you can have an effect without cause. Where did the universe come from if not from nothing? This is like having a conclusion with the premise being false. That is still allowed. You just can't have a true premise without a conclusion. As long as the appearing of some particle out of nothing does not contradict anything else, it is permissible. I believe that this is where "virtual particles" come from. They pop out of nothing.

    Secondly, you object to probabilities because you assume that it is contradictory to logic. Deductive logic has propositions that are either 100% true or 100% false. inductive logic has events with probability anywhere between 0 and 100%. You think these are mutually exclusive. But there is a space that can describe both deductive and inductive logic. See the home page in my profile. Basically, deductive logic is only concerned with the regions in this "sample space" with which we can construct conjunctions and disjunctions (AND's and OR's). Predicate logic is concerned with whether there are ANY samples (not interested in the exact number) within regions representing propositions with certain properties. And probabilities are concerned with the relative number of samples within regions representing "events". The difference in these disciplines of logic only depends on how much information you gather from a function which describes the density of samples in this space.

    It is interesting, when we (I) consider how such a region of an event in sample space can grow with time, such growing events have a boundary that looks very much like the propagation of a string in Superstring theory. If this does prove to be the same thing as string theory, then we will have a physics derived solely from the dictates of logic. A Top Down theory of everything.
  20. Dec 20, 2003 #19


    User Avatar

    I will answer your question with a simple quote :
    "As far as the laws of mathematics refer to reality, they are not certain. As far as the laws of mathematics are certain, They do not refer to reality"
    Albert Einstein
  21. Dec 22, 2003 #20
    What is that supposed to mean?
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook