Hack to the Future

As I must have mentioned here a few times, I'm fascinated by the art and skill (those are not quite the same thing) of weather forecasting, especially the part that involves distilling all those computer models and isobar analyses down to a few brief sentences which will help millions figure out how to dress for tomorrow and whether to take an umbrella. At one point in another time and place, I briefly flirted with an offer to become a TV weatherman before reminding myself that was never what I wanted to do with my life. The brief flirtation, I might add, lasted about twenty seconds.

Now, I follow the forecasts not just to know how hot or cold it might be next Monday but as a fan/student of those who do it for a living. I believe today's forecasters are very good when it comes to telling you the weather for the next 48 hours. When people bemoan that weatherfolks are always wrong, it has usually been because the complainers have taken longer-range forecasts too seriously, which is to say at face value. In the weather forecasting business, there's a tremendous pressure on the meteorologists to issue 7-to-15 day projections…or even longer.

It would make all our lives simpler if we could know if it'll rain two weeks from now…and there are periods where that can be done with a fair amount of confidence. The problem is that when you do 15-day projections, you have to do them all the time. You have to put out that 15-day projection when it's easy to see that far ahead and you have to put it out when it's impossible to do that.

We in Los Angeles have a couple of storms heading our way now. As recently as last Monday, the official word was that we'd have a quarter-inch on Friday, a quarter-inch on Saturday and perhaps as much as 1.5 inches on Sunday, bleeding into Monday. But really if they could, the forecasters would probably have said, "We don't know yet." On Monday, the elements that will comprise this weekend's weather simply had too many options, too many directions in which they might drift.

Around Tuesday, the jet stream did an unanticipated shift and that two inches of precipitation is now heading mainly into Central California and proving to be not quite as damp. Where I am in L.A., we're now looking at one decent shot of rain and one slim chance of a few drops more. Between Friday night and Saturday morning, we'll get a quarter- to a half-inch of H2O. North of us, they should get twice as much, and of course the mountains around us usually get double whatever we get.

They're calling it a 60% chance but it looks to me more like a 100% chance of some rain and a 60% chance of achieving the stated amount. Then on Sunday, there's a very strong chance of some light rain in Central California and a tiny chance of it reaching us. The main computer model, the GFS, says the storm will fall apart totally before it gets to us. The other computer models say there could be sprinkles. As I write this, the National Weather Service hasn't hung a percentage on this one but if they had to, they'd probably put it at 20% just to cover asses.

I often think of weather forecasters when I read political pollsters. They too are under a competitive pressure to peer farther into the future than they should. Like the guys in the weather biz, they have to gather data on now, then throw out educated guesses for then, ignoring the many possible scenarios that could occur before the date in question. They publish them, then they try to refine the predictions as circumstances change and we get closer to the actual event.

I have a feeling that Obama will do quite well but I can also think of a hundred things that could make it, as Vin Scully likes to say, a Brand New Ball Game. Any poll today minimizes game-changers like Romney actually getting the nomination, his choice of running mate, battles at the conventions, what's said in debates, what kind of scandals get unearthed, how the stock market performs, unemployment numbers, possible terrorist attacks, effective TV commercials and the inevitable stupid gaffes we may hear. Look how just one or two clumsy statements or embarrassing revelations changed the presidential prospects of Rick Perry and Herman Cain.

The CBS poll finds Obama and Romney in a dead-even tie at the moment. That's significant because I think that could sew up the nomination for Romney, if not in South Carolina then soon after. A lot of Republican primary voters are likely to vote for The Guy Most Likely To Beat Obama and if the polls say that's Romney, so they'll go. Ron Paul is a close second but Ron Paul is not going to get the nomination because he's Ron Paul and there's not much he can do about that now. Romney's true rival for the nomination is Gingrich and this poll gives Obama an 11 point advantage over Gingrich. That's enough to make a lot of Republicans who'd prefer Gingrich to Romney pick Romney.

In the meantime, the Pew Research poll and PPP have Obama five points ahead of Romney while the CNN and ABC/Washington Post Polls have Romney one or two points ahead. At some point in the future when the Fox News and Rasmussen Polls show the G.O.P. contender with more strength than other polls, we'll be hearing about pollster bias and there may be a smidgen of validity on either side. But right now, the traditionally Conservative polls show Obama doing better than the traditionally Liberal polls…though it's all within the margin of error.

The intriguing difference between predicting politics and predicting the weather is that the former seers actually change the storyline they're trying to project. What the National Weather Service says on Monday doesn't change what the weather is on Friday. But candidates drop out or achieve frontrunner status because of polls. They get or lose campaign donations and even change strategies and platforms because they're down or up.

In both cases, the predictors are expected to say what's going to happen and they're not allowed to shrug and admit that any forecast is premature and meaningless. They've gotta say something so they do. It helps me to remember that they aren't quite the same job even though in both professions, it's possible for a prediction to be too early to mean a damn thing. It's always possible for the jet stream to shift to the north.