It's a combination of understanding today's reality (the fundamentals) and who can project, in this vary complex world of many moving parts, the future most accurately.

A simple example, let's say I own the Coco Cola Company and you want to buy it.  I look at my sales and profit projections and calculate NPV's, risks, etc, and I think the value of the company and its future earning stream is $10 billion.  And I think that basically because I see sales growth at 3%.  you do the same analysis and think that young people in China are going to go crazy over coke, and you use all the same numbers I do to calculate value, except that you see sales growing at 15% per year--you value the company at $20 billion.  So you offer me $12, I take it thinking you are a fool.  Some say we have different perceptions--perhaps true right now today.  But 2 years down the road, sales have grown 15%, not 3%, you are a genius and have a company that can likely be sold for your $20 billion.

Both of us had an understanding of the current reality, but you understood a trend better than I, and you capitalized on it.  The consensus of opinion was likely with me, the owner of the business.  sales had been growing 3% per year for years, why should it change?  You went against my opinion and the consensus, and you were right.  The winner often bets against the consensus of opinion.  he may look wrong in the present period,,,,but in time the trends play out, and we see who was right.

Often the best way to make money is to bet against the consensus.  When people are so depressed about a falling stock price that it's hard to get anyone to buy the shares you hold, that often means prices have fallen to rock bottom, and the market is very undervalued.  If you can see that moment, and act on it, fortunes can be made.

Obviously a very simple example holding all kinds of variables constant, which economists love to do.