The Brilliance of Al-Reuters

bash · Post by **bash** » Wed May 19, 2004 4:14 pm

http://story.news.yahoo.com/news?tmpl=s ... ricanes_dc

NOAA Expects Above Normal 2004 Hurricane Season

NEW YORK (Reuters) - There is a 50 percent probability that the 2004 Atlantic hurricane season will have above-normal activity, the National Oceanic and Atmosphere Administration (NOAA) said in its outlook.

Tetrad · Post by **Tetrad** » Wed May 19, 2004 4:33 pm

Oh now you're just fishing at the bottom of the barrel for insults. That line makes perfect mathematical sense, given that the 'normal' probably isn't a discrete number, but a given range of tropical storms. Therefore something that is abnormal is probably unlikely to happen. Saying 50% likely to have above-normal activity actually means something, unless you have no idea how statistics work.

bash · Post by **bash** » Wed May 19, 2004 4:44 pm

Something that has a probability of 50% of going up also has a 50% probability of going down.

Tetrad · Post by **Tetrad** » Wed May 19, 2004 4:45 pm

That's not what it said at all. It said there was a 50% chance of being outside (edit: and above) the normal range.

bash · Post by **bash** » Wed May 19, 2004 4:46 pm

Explain the headline then. Doesn't that connote more hurricanes whereas it could mean less? The headline is misleading to all but statiticians.

Tetrad · Post by **Tetrad** » Wed May 19, 2004 4:50 pm

What, are you trying to blame the weather on a vast liberal conspiracy now?

bash · Post by **bash** » Wed May 19, 2004 4:51 pm

It's called sloppy reporting.

Tetrad · Post by **Tetrad** » Wed May 19, 2004 4:53 pm

There's nothing misleading about it.

NOAA: 50% chance of having more hurricans than normal
Headline: NOAA expects above normal hurrican season.

Makes sense to me.

bash · Post by **bash** » Wed May 19, 2004 4:55 pm

You just agreed (before you deleted your post) that there also exists a 50% probability of less or the same number of hurricanes, did you not? Isn't 50% no probabilty at all since it could go either way? In other words, it's a coin toss. Either it will be normal or it will not.

Tetrad · Post by **Tetrad** » Wed May 19, 2004 5:01 pm

bash wrote:You just agreed (before you deleted your post) that there also exists the probability of 50% less hurricanes, did you not? Isn't 50% no probabilty at all since it could go either way?

You don't know how probability works, do you? Either that or are trolling expecting people who don't to agree with your underlying point.

You have a range of hurricans which is considered "normal". The reason it is considered normal is that it is very likely (well over 50%) that for that season there are going to be that many hurricans in that range. The NOAA came out and said "the chances of hurricans being above that normal range is 50%". So, there is a good chance that there will be more hurricans than normal.

The opposide 50% is not "There will be fewer hurricans than normal", but rather "there will be a normal amount, or a small possibility of fewer".

bash · Post by **bash** » Wed May 19, 2004 5:02 pm

Heh, why are you belaboring the point? Either there could be more, or the same or less. True? So, that makes this a non-story with a misleading headline since MOST people will read that headline and assume it means there will be more hurricanes.

Tetrad · Post by **Tetrad** » Wed May 19, 2004 5:06 pm

bash wrote:MOST people will read that headline and assume it means there will be more hurricanes.

There's a very good chance there will be.

bash · Post by **bash** » Wed May 19, 2004 5:08 pm

There's also a 50% probability you're over-reacting to a trivial post.

Make that 100%.

Tetrad · Post by **Tetrad** » Wed May 19, 2004 5:10 pm

Hey, at least I know I'm right.

bash · Post by **bash** » Wed May 19, 2004 5:11 pm

No, I'm right, you're left.

Testiculese · Post by **Testiculese** » Wed May 19, 2004 5:14 pm

I read it as there being a chance for more hurricanes.

aside: You two are retarded.

Lothar · Post by **Lothar** » Wed May 19, 2004 5:36 pm

50% chance of "above normal" is actually a meaningful statistic if "normal" refers to a range of activity, rather than a single point.

For example, if "normal" means 3-6 hurricanes, then a 50% chance of "above normal" means there's a good chance of 7 or more hurricanes this year, which is abnormally high. (I'd imagine the chance of there being fewer than normal hurricanes is something like 2%, which is abnormally low.) So Reuters is actually correct to phrase it like they did, despite what Mr. Taranto says (statistics have never been his strong point.) Reuters is often wrong, but in this case they're not.

Anyway... how about we all just pretend this never happened? It's not particularly relevant to E&C, anyway...