Random Realization: Probability as a scam

Today, during a routine discussion session with friends, a very important observation was made.It was about how the scientific experimentation process was inherently flawed because it tends to associate cause and effect relationships to events many times solely based on the premise  of their simultaneous or subsequent occurrence. In many cases, we seldom try to ascertain whether the event is the ‘actual cause’  of the other. The reasoning we give in most cases  is that since event A and event B have occurred 1000 times together or one after the other, there is no reason for A not to be the cause of B i.e  A and B will always occur together no matter how many times we do the experiment.

It was here that I realized how big a scam, probability as a concept is, because it seeks to cover up the aforementioned huge gap in the scientific process. In fact, probability is ‘probably’ the biggest scam invented to cover up  science’s arse. Whenever you are not sure, just say that it is with a 90 or a 95 percent probability that this is the reason for this. So you have the remaining 5-10 percent to save you in case of a goof up.

P.S: This might seem a trivial realization, but still it randomly occurred to me and hence I penned it down.

16 thoughts on “Random Realization: Probability as a scam

      1. What I am trying to is that “probability” differs in a scientific context, and it’s genraal usage. An experiment, by definition is scientific, involving procedures that take into consideration all important variables that might affect the relation between two events. I don’t think any experiment makes sweeping statements regarding the occurrence of two events. Whenever say, an event A occurs and then B happens, science does not start making assumptions right away. Infact, we try to see and work out the trajectories lead to B. And then, we apply our understanding to explain other related phenomena. And most probably whenever an old scientist says, I’m 90% sure, (Atleast I’d like to believe) he is only giving us the benefit of doubt.
        Maybe if you can give me some examples, I can understand more clearly what you are trying to say.

  1. Let me give you an example. We say that an increase in number of vehicles has led to an increase in pollution levels. This might be a true concept in all probability. Please don’t consider this as a baseline example but use this to think of examples of your own to understand what I am trying to say.
    Now an increase in vehicles is one event and increase in pollution levels is another event. How do you know for sure that A causes B. it could be pure coincidence that A and B have occurred together many times in history and hence we give a cause and effect relationship to both. The very act of A being the “cause” of B has never been proved. You may argue that yes it has been proved, but what we have actually done is again associated intermediate events.

    1. I think a sweeping generalization that merely sees correlation and assigns a causation is found only in GRE Analytical Writing questions, not in real Science. For the most part, scientists who do this would be trashed unless they can substantiate their conclusions with some scientific logic/reasoning. That logic may very well turn out to be wrong, or it may turn out that we saw what was perhaps a simplified form of the problem and gave a simplified theory which might not hold in other test cases (Classical v/s Relativistic, for instance). But, some amount of thought process always goes in, if it’s useful science.

      Of course, there is this thing that all of Science can be discarded as mere data fitting. There is no ‘real’ proof of Newton’s second law except that it has worked for zillions of experiments and can be used to predict results of a zillion more. But does God (nature, that is) actually divide the force by the mass to determine how fast the velocity of the body ought to change? We would probably never know.

      1. Agreed. I was just exploring the second perspective after I was fed up of reports in TOI of the type: “Reseach shows that X may be linked to Y”.being quoted for arguments and debates Although I am not sure, I still suspect that many of the cause effect relationships that are found in science have evolved into cause effect through a chain of many intermediate discrete associations. The number of these associations is probably so large that we tend to think of it as a smooth continuous mechanism which we term as the cause an effect relationship.
        P’S: This may be totally wrong, but then after all it is just a suspicion.

  2. Oh, when you read media reporting of scientific research, you have to bear in mind that the report was written by a person who mostly does not understand science and was written so that people would want to read it, which means that 1) it’s dumbed down 2) it’s sensationalized. In general, if you have to choose between an irresponsible scientist and a stupid reporter, the null hypothesis should be stupid reporter.

    Usually, scientists make very careful claims. For instance, they might say something like, ‘Our study shows that there isn’t enough evidence to conclusively prove that there is absolutely no connection between rice intake penis size.’ What will make it to the papers that day is ‘Small penis? Eat rice!’

    I dunno about other fields, but I see this happen all the time in linguistics.

    1. Agreed again. But the point I want to make is more about the “infinite chain of discrete occurrences being possibly misinterpreted as a continuous cause and effect mechanism”. It takes a lot of shredding the ‘science is all powerful’ ego to appreciate the value in the above statement, which is very difficult for people like you and me. Take a few minutes of silence, disassociate yourself from science [i.e imagine you are a Xaviers arts student( :P), that is the way I did it ] and then try to see reason. I am sure you will get my point.
      A purely academic discussion anyway, so don’t bother to do the above if not interested.

  3. You might find this interesting — http://norvig.com/chomsky.html

    It deals with essentially the same thing.

    And I don’t see how probability covers up science’s ass. The scientific method is all about observing, making a theory to explain the observation, and testing it on new cases to see if it holds. The best of our theories are simply explanations of what exists. I don’t know how confidently we can say that our theory is what the ‘real shit’ is i.e. our theory is not merely a curve that fits nature’s curve pretty closely, but is actually the same equation. If probability helps us in coming out with a theory that matches nature’s curve pretty well, I don’t see how it is lesser than Newton claiming with almost 100% confidence the proportionality constant in the second law is 1.

    Of course, I am not condoning brainless statistical analysis to dis/prove everything. But it might be one of the many things that give you insight into a particular phenomenon.

    1. Well, Newton did not claim with 100% confidence that the k in F=kma is 1!!
      Its just that you could define a Force’ = 27*m*a or anything like that and accordingly the definition of force changes…..remember, force is not a fundamental quantity…it is derived from mass and length and time…..

      1. Yeah yeah, the issue is not that it is ‘1’. The issue is that is one particular value. In your case, it would have been 27.

        I agree that a long, tenuous chain of so-called causations/correlations do not make a theory. My only point was that we have (as of now) no real way of knowing what ‘really’ the universe works on. We don’t have access to the operating manual. We can merely perform experiments, fit results, give reasons to explain those results in a consistent and productive manner etc. In this way, we might come up with a framework that correctly predicts whatever we can throw on it, but that is still not the ‘real’ framework. Classical mechanics, for example, can fit in this category. Until we started worrying about high-enough velocities, the data fit too well.

  4. Interesting. Again, I am not against or for the concept of probability. Look at my statement,

    infinite chain of discrete occurrences being “”possibly”(yeah even I need the concept to argue against the concept) misinterpreted as a continuous cause and effect mechanism

    I was merely trying to look at it from a different angle.In fact, I completely agree with whatever you say. I am only trying to make you see merit in the above statement as well, as I consider it an equally valid thought.

  5. Argh!.. I hate labs. I missed a lot.
    But anyway…
    I agree with you. I think there are numerous cases where completely independent events are “correlated” by fitting arbit graphs. And we must always approach such claims with caution.
    You can sort of see this approach in both An Inconvenient Truth and The Great Global Warming Swindle (Two documentaries that argue on the same question). A lot of arguments are presented in terms of graphs and lines that overlap. We might not have the big picture but at the same time, we can never know what part of reality is being obscured to prove a point. Now, I don’t know how right or wrong they are but I agree with you that this attitude has become increasingly common, especially in the media.
    As long as you are not generalizing this attitude to all sciences, I’m good.

  6. @Krishnan………The objective of theories is not just to explain the ‘truth’ (in fact it shouldn’t be, because of what you pointed out)…..It is actually mainly to develop a conclusion on certain phenomena which when replicated or applied to other similar cases would yield analogous results…..
    As Antariksh correctly pointed out, this may continue to work till we reach a case that we thought was analogous but is different as in the case of relativity….
    Even then, we use not relativistic Newton’s Laws and Galilean Transformations which we know are not the ‘truth’, because its not the truth that is our major concern….
    Basically, science will always be an approximation…….

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s