Econometrics

There's no such thing as a sure thing.
　
   Econometrics combines economic theory, statistics and mathematics in an attempt to reduce human behavior to a formula that will accurately predict future business events and behavior.
  
   Hah!

   Econometrics does not deal in certainties, however much politicians and investors may want.
   It's a useful tool, but like any tool, it's only as good as the user. And since the user, in this case, may be the same person who designed and created the tool, that means the prediction it yields is only as good as the designer.
   Or as they say in the computer biz, GIGO -- garbage in, garbage out.
   So it's like reading a tout sheet at a race track, or a corporate earnings report, or an investment fund prospectus. But past performance is no guarantee of future growth/profit/victory.
   This is not to say that econometrics has no value. But there are limits, and it's important to understand what those limits are. Understanding these limits, and knowing how to use the information within these limits is key to making useful predictions. Pollsters do it regularly, and the better ones point to the "margin of error" in their analyses. A two-point margin of error when one candidate has 51 percent and the other 49 percent could well reverse the conclusion.
　
   Flip a coin, and what's the probability it will come up heads? Answer: 50 percent. And no how many times you toss that coin, each time you toss it, the probability of coming up heads remains 50 percent.
   The prediction becomes more complicated when dealing with several coins in the toss, but there remains only one variable for each coin: heads or tails. Move on to a pair of dice, and you increase the number of variables: six sets of spots on each die.
   So if you want to, you can deal with the variables in a regularized mathematical way, just as gamblers do with a deck of cards or a roulette wheel. In fact, the science of probability was devised by the 18th Century French mathematician and philosopher Blaise Pascal, whose patrons happened to be ardent gamblers and card players.
   The fun begins when you apply mathematical principles to human behavior.

   Statistics has to do with gathering data (numbers) and organizing them into patterns. Probability has to do with what those numbers and patterns portend for the future. When you stir statistics and probability into the human stew that is economic theory, you get a spicy taste of what may happen in the future.
   Then again, it may not.

   Economics is the study of what people do with what they've got, or what's available to them and whether and how they use it.
   Econometrics may be the most challenging of all academic disciplines, because it tries to identify the many relevant variables of human behaviour and sort them using the strict formulas of statistics to come up with a reasonable probability of future action.
   The problem is in deciding what is relevant, and what is reasonable.

   Lotsa luck with that one.