What a difference more “realistic” guidance makes.

As CFO Peter Oppenheimer explained the new rules last January, Apple would no longer give analysts “conservative” forecasts. Rather than tell them what the company was reasonably confident it could achieve the following quarter, it would offer a range of guidance that reflected what it believed it was likely to achieve.

The distinction between “belief” and “reasonable confidence” is a subtle one, but judging from the results, most of the 62 analysts (professionals and amateurs) who responded to our polling understood what Oppenheimer was getting at.

Two years ago, the pros missed the company’s March quarter revenue and earnings by a combined average of 11%. The amateurs did considerably better, missing by only 2.5%.

This quarter, 26 analysts came within 2.5% and the average error for the entire group of 62 was 3.7%.

There were, however, some outliers who didn’t get the memo. Gabelli’s Hendi Susanto forecast revenues of $47.7 billion, $4.7 billion higher than the top of Oppenheimer’s $41-43 billion guidance range. On the other extreme, Citicorp’s Glen Yeung and Rajesh Agarwal of the independent Braeburn Group came up half a billion or more short of Oppenheimer’s goal posts. You’ll find their names at or near the bottom of this quarter’s rankings.

Among those who took Apple’s new rules seriously, several deserve special mention:

• Merrill Lynch’s Scott Craig, who came in first in the race for best revenue and earnings estimates, having missed Apple’s $10.09 EPS by a single penny.
• Canaccord Genuity’s T. Michael Walkley and the Braeburn Group’s Michael Cranston, who finished  in second and third places, respectively.
• The ISI Group’s Brian Marshall, who had the best average estimate when unit sales and gross margin were included.
• J.P. Morgan’s Mark Moscowitz, Wells Fargo’s Maynard Um and the Braeburn Group’s Matt Lew, who didn’t place but managed to predict Apple’s $43.603 billion revenue with an accuracy of 0.11% or better.

As for the original impetus for the Earnings Smackdown — to highlight how much better the amateurs had been doing than the pros at predicting Apple’s results — that’s all but disappeared. In Tuesday’s results, the two groups were nearly indistinguishable.

More important, Apple seems to be sending much clearer signals to Wall Street. Whether that results in more realistic valuations for the company remains to be seen.

Below the fold: Our annotated master spreadsheet, with the best estimates highlighted in bright green, the second and third best in light green, the worst in red and the second and third worst in pink.

Thanks one last time to Posts at Eventide‘s Robert Paul Leitao for pulling together the Braeburn Group numbers.