Anyone remember that weird math discovery that predicts how long any phenomenon lasts?

[HamdenRice]

It was in the press about 10 years ago. It was a statistical theory that enables you to predict with 90% certainty how long any phenomenon, human or natural, will last.

I probably have the numbers off but it went something like this: You are 90% certain that any ongoing phenomenon or situation will continue to last for between 10% of the amount of time it has been going on and 10 times that amount of time it has been going on.

So, for example, if you want to know how long the US will continue to be a constitutional democracy, you can't predict the future, but because it's been going on for 200 years, you can assume that there is a 90% probability that it will continue to exist for between the next 20 years and the next 2000 years.

If the current economic crisis has been going on for six months (180 days), then there is a 90% probability that it will continue for between 18 days and 1800 days, or about 5 years.

I know it sounds trite, but it was supposed to have had serious applications in things as diverse as stock market predictions and physical sciences. In economics, I think it allowed statisticians to rule out certain unlikely scenarios, like -- in the examples above -- the collapse of the country within 5 years or a 10 year bear market.

Anyone heard of this/remember it?

[BlooInBloo]

1. I have no idea what you're talking about, but....

Birth-death processes have been around for a loooong time.

http://en.wikipedia.org/wiki/Birth-death_process

[Whoa_Nelly]

2. Is it The Long Tail?

http://en.wikipedia.org/wiki/The_Long_Tail

[ChairmanAgnostic]

3. when combined with the special theory of relativity,

the relative being my ex, all I know is that the length of my marriage seemed 10x longer than was healthy for me, and cost 10x as much as it should have, leaving me with 1/10th the assets I should have been able to keep.

[Warpy]

4. That sounds about right, but the cost, when added up,

was still about 0.01% of the cost of psychotherapy had I stayed in the marriage for 50% more time.

[HamdenRice]

7. No, see below, but they are related ideas, based on the Bell Curve

The Copernican Principle is just a fancy way of saying that you are more than likely in the middle section of the Bell Curve rather than on the end.

The Long Tail is a business model that seems to be saying that you can make money on the ends of the Bell Curve because most other businesses are focused on the middle of the Bell Curve.

[On the Road]

8. Never Heard of It Before

but that is absolutely fascinating. It makes a lot of sense as well. Doesn't cover all the bases, of course, but that falls under the remaining 5%.

[Posteritatis]

5. The Copernican Principle

I first came across it a year or so ago in a Jim Holt column.

[HamdenRice]

6. Thanks, that's it! And it's 95% accuracy. It was discovered in 1993.

http://www.nytimes.com/2007/07/17/science/17tier.html?_r=1

A Survival Imperative for Space Colonization

By JOHN TIERNEY
Published: July 17, 2007

In 1993, J. Richard Gott III computed with scientific certainty that humanity would survive at least 5,100 more years. At the time, I took that as reason to relax, but Dr. Gott has now convinced me I was wrong. He has issued a wake-up call: To ensure our long-term survival, we need to get a colony up and running on Mars within 46 years.

...

Dr. Gott has used his technique to successfully forecast the longevity of Broadway plays, newspapers, dogs and, most recently, the tenure in office of hundreds of political leaders around the world. He bases predictions on just one bit of data, how long something has lasted already; and on one assumption, that there is nothing special about the particular moment that you’re observing this phenomenon. This assumption is called the Copernican Principle, after the astronomer who assumed he wasn’t seeing the universe from a special spot in the center.

Suppose you want to forecast the political longevity of the leader of a foreign country, and you know nothing about her country except that she has just finished her 39th week in power. What are the odds that she’ll leave office in her 40th week? According to the Copernican Principle, there’s nothing special about this week, so there’s only a 1-in-40 chance, or 2.5 percent, that she’s now in the final week of her tenure.

It’s equally unlikely that she’s still at the very beginning of her tenure. If she were just completing the first 2.5 percent of her time in power, that would mean her remaining time would be 39 times as long as the period she’s already served — 1,521 more weeks (a little more than 29 years).

So you can now confidently forecast that she will stay in power at least one more week but not as long as 1,521 weeks. The odds of your being wrong are 2.5 percent on the short end and 2.5 percent on the long end — a total of just 5 percent, which means that your forecast has an expected accuracy of 95 percent, the scientific standard for statistical significance.

...

Now, that range is so broad it may not seem terribly useful to you, and Dr. Gott readily concedes that his Copernican formula often is not the ideal method. The best the formula could do regarding Bill Clinton, who had been president for 127 days when the 1993 paper in Nature was published, was predict he would serve at least three more days but not more than 13.6 more years. You could have gotten a narrower range by using other information, like actuarial data from previous presidencies, or factoring in the unlikelihood that the Constitution would be changed so he could serve more than two terms.

But the beauty of the Copernican formula is that it allows you to make predictions when you don’t have any other information, which is how Dr. Gott managed to predict the tenure of virtually every other nation’s leader that day in 1993 — a total of 313 leaders. If none of those still in power stays in office beyond age 100, Dr. Gott’s accuracy rate will turn out to be almost exactly 95 percent.