Fuzzy Reality

The number-crunchers won last night (and so, by and large, did the Dems).

Nate Silver, the grand poobah of the number-crunchers, and therefore the chief focus of vitriol from those who did not like the trends the numbers since about mid-October had been showing, turned out to be right, and his detractors wrong.

As I pointed out about six weeks ago, Nate and his number-crunching ilk seek to divine — through number crunching — what all the political polls, taken together, actually mean, and to lay out that meaning on a daily basis during the months leading up to the election.

To do that, the number-crunchers have to smartly aggregate the polls which, because they seek to measure, via sampling, something huge and complex, tend to be noisy and somewhat chaotic. But, the thinking goes, by combining those noisy and chaotic polls in smart ways, a smart number-cruncher should be able to separate the wheat from the chaff — to separate the signal from the noise, as Mr. Silver would put it.

Last night, the theory, in practice, proved quite powerful; the signal, as the number-crunchers divined it ahead of time, aligned with the reality, as it unfurled. So much for punditry intuition, and so powerful the excellent numbers-based prediction.

*  *  *

Many of the numbers populating our financial lives follow this same pattern: they are both signal and noise.

The most obvious example is the account balance for any one of your investing accounts. Think about the balance of such an account that contains stocks and the like — investments that are priced second-to-second throughout the period each weekday during which the stock market is open — and think about how, at most, there are really only two numbers that an unusually fastidious, eye-always-on-the-ball sort of investor really needs to think about, i.e., the account balance as of the opening of the market day and the account balance as of the closing of the market day.

And some would say that that is one — maybe two — numbers too many at that . . .

Now someone who manages money for a living ought to pay more attention than that. And anyone considering selling or buying something for an investment account is well advised to pay more attention as well.

But, in general, the more attention you pay to all those numbers, the more noise through which you must sift and, unless you’re good at distinguishing the signal from the noise, the more confused, and therefore the less informed, you’re apt to be.

And that’s why most financial advisors advise their charges to pay attention to the numbs on, say, a monthly or quarterly basis, and, otherwise, to go about their daily business attending to things that really matter, such as whether that traffic light they’re looking at is green or red.

*  *  *

Many things in nature tend to be rather fuzzy when studied closely. So while there are amazingly precise parts of nature:


there are also those that are inherently uncertain and fuzzy (so much so that pictures comprised of pixels cannot do them justice, and links must suffice).

But then there are times when something that’s chaotic and fuzzy and indeterminate transitions to something that is gaining in orderliness, and then, finally, condenses down to an absolute certainty.

We’ve seen one of those transitions over the course of the last several months, as the probabilities set out in Nate Silver’s projection started out at about 60% odds that Obama would win to, the day before the election, 91% that he would win, with a big odds-course correction starting right around the time of the first debate and lasting for about ten days, after which Obama’s odds started going up again, all the way to 91%.

Here is that pic:

Silver's final odds pic for 2012



Unfortunately, the screen scrape left the scaling info behind, so here is what you need to know to get the scale of the pic: the double line centered vertically (remember: you have vertebrae in your backbone and your backbone goes up and down, and therefore “centered vertically” means “centered up and down”) (also, the horizon goes left/right, and therefore “horizontal” goes left/right) is 50% and each line above and below represents a 12.5% increment, so that the overall up/down scale is 0% to 100%, with Obama’s odds of victory shown in blue and Romney’s in red, and the left/right time scale being May 31 through November 6, and each vertical line being the beginning of a month).

And then, yesterday, the pace at which reality unfurled really picked up steam, via tens of millions of decisions and actions that we all made and took, so that, as of last night at about 8:30 West Coast time, the odds went up to, let’s call it, a 99.9999% probability that Obama won the electoral college (that’s a one in a million chance that he did not).

And, presumably, on December 17, 2012 (the first Monday after the second Wednesday in December), when the electors actually cast their votes, the odds will go to 100.0000%.

Reality will be totally unfurled at that time, probabilities having fully condensed down to a single, immutable, unmistakable reality.

*  *  *

So remember that account balance I mentioned above — the account balance showing at any given time for some investment account of yours?

My advice to you is that you think of that number as a collection of fuzzy probabilities that apply to each of the holdings in that account — as a collection of possible prices, some of which are more likely at any given moment than others and some of which are quite unlikely — and that you also understand that, once you go ahead and decide to change something in that account by, say, selling one of the holdings, then, and only then do you truly know how much that holding is worth.

Until then, please do think of the information value embedded within that account balance as more wispy cloud than Fibonacci-geometried sunflower.

975 words (about a ten minute read, sans links)


Leave a Comment



  1. Posted by Mary Donovan on Wednesday, November 7, 2012 at 1pm
    Outstanding piece, John. and I loved the graphic.