Okay, once again, I’ve been bested. Last week I guessed this Sunday would be 72 and partly sunny and on Thursday, my weather app predicted 83 and mostly sunny. It turns out that it was 88 and sunny today, so the weather app has won for the second week in a row.

Now in my defense, I’ve guessed seven days out for two weeks in a row and the weather app had two and three days respectively for its predictions. I say they lucked out, but I’ll change the rules again, because I’m the one writing this and likely the only one paying attention to this competition. Next Sunday, according to my feelings, the weather for New Hampshire’s seacoast will be partly cloudy and 68. Checking the weather app on my phone, it’s saying: 67 and partly sunny. Ok. Truth be told I guessed before I looked at the app and it looks like we’re both guessing pretty close. But that’s the way these things work.

So we’ll see how we did. This could be a Price is Right kind of competition this week, where if it’s lower than 67, the app wins and if it’s over 68, I win. Here goes . . .

Sunday's game isn't about football

Well I’ve noted my weather app’s prediction for Sunday and it’s pretty different from what I guessed last week. To recap: I’m pitting my sheer guesses against my phone’s weather app to see which is more accurate. On Thursday, my phone told me that Sunday would be 83 and mostly sunny. Last week I guessed it would be 72 and partly sunny. Now all we have to do is wait.


It takes a big man to admit he’s wrong. In the challenge of me versus iPhone, it’s weather app 1; me 0. Ah well. The weather for Sunday is 77 and partly sunny. On Friday, the weather app correctly predicted 77 and partly sunny and that’s what it is. I had guessed 72 and partly cloudy (isn’t that basically partly sunny?) a week ago without any help from computers or satellites, etc. I was off by five degrees.

But I still think I’m onto something. I’m going to even things out just a hair. Let’s put the weather app back a day. Let’s see how accurate it is three days out. I’ll still stick with a week, just to show that I’m completely guessing. So next Sunday, I’m predicting, will be partly sunny and 72. I’ll check the ol’ weather app on Thursday and take a note of its prediction and we’ll go for another round . . .

Quiet everyone! The bet's on.

Last week I griped about the inaccuracy of my phone’s weather app and I predicted what the weather for this Sunday would be without the aid of any computers, satellites or the like. (Yes, I’ve used several different apps, but they all seem to be equally bad at predictions.) Yesterday I checked my app and it said tomorrow (Sunday) would be partly sunny and 77. My prediction last week was partly sunny and 72.

Now we play the waiting game. Let’s see who was more right. This is getting exciting, right?

Weather app, schmeather app

I’ve had a theory on weather predictions for the past decade or so. It goes like this: Sometime, long ago, an editor or manager of some news service told his meteorologist that it would be wonderful for circulation if he could predict the weather two days ahead instead of just one. The meteorologist, knowing his craft, said that to do so was folly. With the innumerable things that can affect weather, he said, accuracy wouldn’t be better than 50-50. One day, he relayed, was pushing it; two was a crapshoot. To hell with it, the editor replied. Give us two and people will think it’s better than nothing.

So the meteorologist started the two-day forecast and it was a hit. Despite its inaccuracy, the public figured this news outlet must know more than the others, something to do with modern technology.

Then the editor at a competing news organization asked his meteorologist why he wasn’t giving them two days, to which the man replied the same as the first: It’s nothing more than an educated guess. Well, dammit! We look like Johnny-Come-Latelys. Pump it up to three days! And so the second meteorologist did his best to predict three days out.

Then we had five-day forecasts, and then seven and now 10. How accurate is 10? They have trouble with hurricanes that far out, and those tend to be really big, so I’m guessing a 10-day prediction has some correlation to a broken clock being right twice a day.

Faithfully each morning I open the weather app on my phone and check the latest predictions for the coming days, and sure enough it’s just as apt to be wrong as right (often even on the same day) and I wondered how my random guesses would compare. So I’m going to try predicting the weather without any satellites, models or computers. It’s Sunday, so I’m going to guess the weather for Seacoast New Hampshire for next Sunday. That’s seven days out. Then I’ll note my phone’s prediction on Friday and keep up the guesses for the weeks or months after and see how complete guesses stack up.

Next Sunday, Sept. 17, in my area, will be partly cloudy and 72.

The game’s afoot!
Blog Navigation