While not exactly scratched on the back of envelopes, asteroid impact odds are based on frustratingly small data sets and terribly loose formulas that leave wide margins of error. The result has frequently been public fright and confusion that some scientists say should never have happened.

More than once, the calculations have provided for scary headlines that leave indelible impressions of dread in the public mind even as the risk faded to zero when more data was gathered.

Now, scientists tell us, they've got a much more rock-solid case for a possible impact.

Asteroid 1950 DA might or might not hit Earth in the year 2880, according to a report that will appear in the April 4 issue of the journal Science. The maximum odds are 1-in-300.

But, for the first time, the risk is being given as a range of possibilities. The rock might well have a zero chance of hitting us, depending primarily on which way its axis is tilted.

Either way, 1950 DA represents a unique situation in which time and luck have combined to allow more precise projections of a potentially threatening asteroid's future than ever before.

When astronomers spot a previously undetected object in the sky, their first order of business is to see if it moves on subsequent nights in relation to the relatively stable background of stars. Typically from just a few nights of data, they then calculate the asteroid's path.

By considering several factors, including the gravitational nudges of the Sun and various planets, astronomers and software experts combine efforts to make a preliminary calculation of the object's orbits as far into the future as possible.

But often the rock zooms beyond view before a strong case can be built.

The estimates are sometimes off by so much that an asteroid predicted to be targeting Mars might one day hit Earth instead. Or, conversely, an asteroid that seemed to be on course to possibly hit Earth will, after its path is watched for weeks or months, be found to pose absolutely no threat.

So imagine the delight of the asteroid hunters when they rediscovered asteroid 1950 DA a little more than a year ago and realized they had data on the rock's path that spanned five decades. The long time frame, along with new radar observations of the asteroid's surface, allowed a crack team of leading experts to consider a host of factors they would not normally be able to include when projecting an object's orbit into the future.

"It's the most thoroughly investigated case so far, that pulls in factors not previously considered in an integrated way," says Jon Giorgini, an engineer at NASA's Jet Propulsion Laboratory, of asteroid 1950 DA.

Giorgini's team included other JPL researchers along with scientists from Caltech, Washington State University, the University of California, Santa Cruz and the Arecibo Observatory.

Of course Giorgini and his colleagues also realized, upon processing all this wonderful data, that 1950 DA had the greatest chance of hitting Earth that had ever been calculated.

The odds are 1-in-300 that the asteroid will hit our planet on March 16, 2880, assuming it is spinning in a certain way and its pole points in a particular suspected direction. However, the pole might be pointing somewhere else. It's hard to tell, the scientists say. If the pole points elsewhere, then the rock will be propelled onto a different path and the odds of an impact are nearly zero.

The newly calculated impact risk represents a stroke of luck not likely to be repeated soon.

"Although other currently unknown asteroids may hit before 2880, the situation with 1950 DA is likely to be unique for all future detection announcements," Giorgini says. "This quality of information for a hazard case just won't happen again for a long time."

Which is to say that asteroid scares will likely be with us for some time to come.

All the while, the precisely calculated chance of doom on the eve of Saint Patrick's Day in 2880 has the potential to hang over the heads of 35 generations of human beings.

So, should we or our descendents worry for the next 878 years about a gargantuan space rock named 1950 AD?

Not according to Giorgini. It could take decades, but eventually it's quite possible that fresh observations will reveal 1950 AD to be headed somewhere else. And regardless, he said, there is plenty of time to figure out a way to deflect the asteroid if need be.

"The most likely thing is that Saint Patrick's Day parades in the year 2880 may be a little more festive than usual as this object recedes into the distance, having passed Earth by," he said.