Prediction in Complex Seemingly Random or Chaotic Systems Discussed Philosophically

Can we really predict seemingly chaotic events and systems? Many believe that we can if we just know the exact starting point and all the components along the way, but therein lies the problem, we can’t know. Yes, so many folks attempt to predict the future, including myself, working so hard to find the patterns, the clues, the nuances, and subtle tale tell signs. Let’s talk.

Did you ever see the Hollywood Movie; “Knowing” by Nicholas Cage where he figures out the mathematics of future events, disasters? Even the ability to predict the end of days; or, how about “A Beautiful Mind” about the famous mathematician John Nash, working to find the secret, he never did, went nearly insane trying, still, this notion of trying to understand using math seems to have a logical basis, and many have pursued to their own peril, and yet, maybe they are right.

If you read up on the RAND Corporation, the Think Tank which attempted to quantify everything and make future predictions and strategies for our military during the cold war, you see that the best and brightest minds were hired to work on such seemingly impossible problems. For instance, please read:

1.) “Soldiers of Reason – The RAND Corporation and the Rise of the American Empire,” by Alex Abella, 2008 (uncorrected proof copy).

2.) “Think Tanks,” by Paul Dickenson, 1971.

Now then, on an intellectual note, please indulge me in taking this dialogue to a higher level. You see, Ray Kurzweil a rather prominent engineer, scientist and now futurist explains to us that while it is almost impossible to know the long-term future, it’s relatively easy to get most of your predictions right in the short term – 2, 3, or 5 years for instance. He’s not the first to note this – myself and others in fact have said this before and concur with this observation.

When predicting the weather, we are pretty good at it for 2-3 days, 4-at the most, but beyond that it is silly to entertain the prediction models or to use them to make plans for your family BBQ. I know this, having been in the car washing sector which relies on the weather to determine volume, thus labor needs.

When using algorithms to predict the stock market or individual funds, stocks, ETFs, or commodities, very recent past information and trend is an excellent predictor of near term movements, we’ve all seen how derivative strategies work. By the way check out “The Black-Scholes Formula,” on YouTube (in 5-parts, approximately 10-minutes each).

We can predict some long-term trends provided we know the exact starting point of all interactive events effecting our system or model, and although we know we cannot, just as we cannot know when an anomaly in “white noise” will occur without knowing all the events through space time, like Brownian Motion we can know that the probability of X-number of events of x-number of types should occur until the anomaly makes modification of our computer modeling, that is if our input has been made honestly and with painstaking detail. Even one fudging of the numbers or data for political reasons or otherwise, changes the entire set of events of the mathematically precise world inside the Matrix (Hollywood Movie).

From each inflection point of each anomaly we can predict the “near-term” new trending lines very well, but only for a very brief time period. Thus, it is there we must concentrate our efforts, and that’s where I concentrate mine. Be well, live well, and live longest!