Ever since I met Michael Martin Nieto nearly a decade ago, I've been fascinated by his work on the Pioneer Anomaly. It's partly because the anomaly is so intriguing: the spacecraft, as they headed out of the solar system, behaved as though there's a little more gravity tugging on them than their ought to be. But part of the fascination is Nieto himself. He's a funny guy, who loves the mystery, and has a very careful attitude toward the whole thing. The hook here is that there are a couple of proposals for new space missions to test the thing. But that's just an excuse. I really just wanted to write about it again. From this morning's Journal:

When a scientist finds new data that conflicts with a well-established theory like Newton's or Einstein's, the first reaction is not likely to be to book tickets to Stockholm for their Nobel Prize ceremony. The more plausible explanation is that they've missed something.

So for the last decade, Nieto, Anderson and others have banged away at the data, trying to make the "Pioneer anomaly" go away.

A software error was one of the first things they tested. They analyzed the data using software developed by two completely different groups, and got the same answer.

With that ruled out, they figured the most likely explanation was what they call a "systematic error"— something about the spacecraft that they've missed, perhaps a gas leak that's sending it off course, or some sort of uneven heating of the craft, or something of that sort.

When people offer Nieto that explanation he's likely to agree. "I say, 'Hell, you're probably right, but I don't do faith-based science.' ''

In other words, it's not enough to say it's probably a systematic error. You have to find it. But try as they might, the Pioneer anomaly team has been unable to do so.