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

Hunches and guesses

  1. Oct 26, 2016 #1


    User Avatar
    Gold Member

    How do you get your hunches even if they are out of your circle of expertise, and how often are they right?
  2. jcsd
  3. Oct 26, 2016 #2
    Problem solving skills and intuition are not discipline bound.
  4. Oct 26, 2016 #3
    Experience and observations - justified by a foundation of education.

    As an engineer with not the best grades, I started as a field service engineer. In this role I was lucky enough to see MANY types of failures, and to understand and address them I relied on a good conceptual education over a number of disciplines. When ever we came across a problem I always sought to find a cause(s) based on proper theory. This is not at all limited to engineering - mathematics, statistics, psychology, even English (language) and communications - have all been parts of "fixing" failures I have come across. A good example - is System / Feedback Theory, much of this can be applied to SO many cases in the real world, in ether a numerical case or as part of a process- everything done well has a loop - somewhere.

    Also - a good engineering background, allows you to see many analogies in all of the things around us.

    Today - 30 years after University - the experience of solving those issues, and remembering the theory behind it, has allowed my to become a good analyst of both technical and commercial situations.
  5. Oct 26, 2016 #4

    Jonathan Scott

    User Avatar
    Gold Member

    It's a matter of being able to work with vague knowledge and have a feel for what's more or less likely, rapidly evaluating hypotheses. At work I've often found others stuck trying to solve a problem because there's no logical way to deduce the cause of a problem from the symptoms, whereas I start with a cloud of vague ideas in my head as to what could have gone wrong and rapidly narrow them down to find ones which could account for the symptoms, until I find a plausible match. In particular, I often focus on areas where people are likely to have made mistakes because of complexity or obscurity, and of course I usually consider a single point of failure more likely than multiple points of failure. I then work from both ends, modifying my guesses and checking for confirmation in the symptoms until I know what's happened.

    My hunches usually turn out to be correct, but that's partly because they are start off sufficiently vague to cover all likely possibilities and evolve as I make progress. In my own areas of expertise, the process is often so rapid that it seems almost subconscious, but if I think about it I usually find I had some previous experience which led me in that direction.
  6. Oct 28, 2016 #5

    Jonathan Scott

    User Avatar
    Gold Member

    I seem to have an example in progress:

    Last month I posted an idea of what might have caused the recent SpaceX anomaly in the SpaceX needs us thread, a subject on which I have no specialist experience. Unlike most posters both on that thread and on the Nasa forums, I thought the initial problem was not the explosion as seen but rather a very sudden failure in the helium pressure system, and that the explosion came a few milliseconds afterwards as a result of the second stage being ruptured, causing ejection of a significant amount of LOX and fuel mixture which could then ignite trivially easily, for example as a result of heat from rupturing materials. I also later mentioned that the failure could have been triggered by thermal stresses during LOX loading.

    So far, this hunch has not been confirmed, but as far as I know it is now the main line of inquiry.
  7. Oct 28, 2016 #6
    "I believe in a long, prolonged, derangement of the senses in order to obtain the unknown." Jim Morrison

    While this is not necessarily scientific, I tend to agree.
  8. Nov 3, 2016 #7
    New on the schedule, looks as if your going to be correct on this one. :smile:
    2016 December 18 04:30 LC-39A KSC Falcon 9 EchoStar 23
  9. Nov 4, 2016 #8
    I have a model of the way the human brain works. It teaches me how to solve complex problems and how to come up with brilliant solutions. First you have to accept the notion that your subconscious brain is an amazing pattern recognition machine. I have learned how to use this machine. First you formulate a question. Next you gather information related to that question, then you go for a drive. Forget about the problem, let the hind brain cogitate until it spits out an answer for you to examine. I'll give an example.

    I wanted to know what "intelligence" was, intelligence the verb, not the noun. So I read all about intelligence, about animal intelligence, machine intelligence, computer intelligence, human intelligence, composite intelligence, design intelligence, artificial intelligence, and so on. You make a big messy pile of information. It took years. I read science fiction and fact, math, and computer programming languages, then I stopped. I said to myself, what is intellect? Then I forgot all about it, let the big number crunching part of your brain go to work.

    One day an answer popped into my head. Fully formed. Just like that! The big subconscious hind-brain had come up with an answer and it popped it to my attention. Intelligence is "the ability to extract data, information, knowledge, and wisdom, directly from images." Now the answers that pop out are not always correct, but they are always interesting and important. This is how I think intuition works.

    Does that make any sense to you?

    I think I have a working definition for "intelligence"
  10. Nov 4, 2016 #9


    User Avatar
    Science Advisor

    Pattern matching against schemas.
  11. Nov 8, 2016 #10
    I think the Oxford dictionary may have beaten you to it.
  12. Nov 8, 2016 #11
    I am not going to brag, but my hunches have successfully predicted the winner of every presidential election in my adult life (40 years worth).
  13. Nov 8, 2016 #12


    User Avatar
    Gold Member

    I agree with much of what you said about unconscious computing.

    In your definition of intelligence, why extraction from images
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook

Have something to add?
Draft saved Draft deleted