When soothsayers talk about the future, best take their words with a grain of salt.

Here, for instance, is what I said last May when search engine Google was planning to go public at a price as high as $135 a share, an astronomical figure relative to the company’s earnings:

“Google appears to have set a high price based on ample evidence that investors are drooling over the stock. But investors don’t drool forever, as we saw when the tech-stock bubble of the late ’90s burst in 2000.”

If I were buying Google shares, I said, “I wouldn’t bet more than I could afford to lose.”

I wasn’t the only skeptic, and Google eventually went public in August at a more modest $85 a share.

But it turns out the worriers were wrong.

On April 22, Google shares soared 5.7 percent to $215.81 after the company reported a nearly sixfold increase in first-quarter profits, with sales nearly doubling. It closed Friday at $220.

There is a lesson here that also applies to the stock market’s wild swings in the past couple of weeks: For all our research and analysis, predicting whether stocks will go up or down is about as scientific as flipping a coin.

This isn’t a new lesson, but we need to relearn it over and over, else we get carried along as the crowd chases one fad after another.

Keep in mind that stock prices are mainly governed by the pro traders who do most of the buying and selling. A few weeks ago, these wise heads looked at sour economic and inflation news and drove the Dow down 4.8 percent in six trading sessions.

Then, on April 21, they focused on some surprisingly good corporate earnings news and gave stocks their biggest one-day surge in two years — more than 2 percent.

If pros could forecast companies’ long-term performance, they could calculate appropriate share prices, and shares would just stay there. But the pros can’t do that — there are too many uncertainties, so stocks whipsaw. Ordinary investors can’t, either.

Another Google lesson: From time to time there is a breakthrough company that justifies investors’ enthusiasm.

But these are the exceptions that prove the rule: Most “paradigm shifts” are mere wishful thinking. For every Google, IBM or Microsoft, there are dozens of once-hot stocks that crashed.

Back in the late ’90s Internet bubble, investors convinced themselves that shares in profitless tech companies should sell for sky-high prices because, in the paradigm shift called the “New Economy,” earnings and sales figures no longer mattered. That was wrong, and the bubble burst.

Google, however, shows how companies can benefit on those rare occasions when paradigm shifts are real. In this one, advertisers are spurning traditional media such as newspapers and TV to move to the Internet. And Google has a peerless search engine to draw viewers to those ads.

But unlike the Internet high-flyers that tanked a few years ago, Google has delivered good earnings and sales. Investors evaluating the stock don’t have to rely on new-fangled gauges such as “hits” and “page views.”

Sure, Google is trading at a price about 85 times the past 12 months’ earnings, well above the stock market’s historical average of around 15. That makes Google risky.

But the company has a track record of stupendous earnings growth and it is the premier player in an international market with lots of room to grow. And the price-to-earnings ratio is a more modest 55 if you use analysts’ earnings projections for the rest of this year — especially since past forecasts have been low.

So Google may be, in many ways, the perfect stock: Its product is on the cutting edge, yet investors can focus on old-fashioned data they can understand — earnings and sales. Analysts recently said Google shares could go to $250 or higher by the end of the year.

I still wouldn’t pour money into Google unless I could stomach a big loss. But that’s nothing against Google — it’s a good rule with any stock.

After all, the analysts are often wrong. And Google has raised investors’ expectations so high it will be easy to disappoint them.