Math Intollerance at The New York Times and The Atlantic.

To explain the rather brief post below, Matthew Yglesias has some typically ruthless snark "Surely the NYT has it within its powers to be aware of the results of its own polls and get its writers to characterize the trends accurately."

He is commenting on a totally dishonest op-ed by William Kristol. I quote all of it (fair use be damned).

Some Editing

Bill Kristol's column from Monday:

In a New York Times/CBS News poll in late February, Obama was defeating John McCain 50 to 38. Two months later, the Times/CBS poll had McCain and Obama tied. The poll that came out yesterday showed Obama reopening a lead over McCain — but clearly over this period a vulnerability for Obama was exposed.

As Noam Scheiber notes it's a bit curious of Kristol to have left out the precise numbers from the new poll. But what it says is that Obama hasa lead of 51 to 40 which is identical to Obama's previous lead. I'm hardly shocked to see Kristol playing some funny games, but shouldn't there be some kind of editing of the Times columnists? Surely the NYT has it within its powers to be aware of the results of its own polls and get its writers to characterize the trends accurately.

Brilliant except that, uh, 50-38=12>11=51-40. The post would be just as snarkily good but also, you know, arithmetically accurate if Yglesias had written "almost identical" or "within rounding error."

It would be far less snarkily good but, at least, accurate and full of his characteristic homonyms, if he had written "given the fact that results are rounded to the nearest percent, for all we no, Obama's lead in the latest CBS/New York Tiems pole could be slightly larger than his lead in the late February CBS/New York Times pole two witch crystal refers."

"Some Editing" poster edit thyself.
I actually suspect that Yglesias didn't decide that 1% is essentially 0 and then claim that it was 0, but rather he considered that we don't know which lead was larger before rounding and decided (wisely) not to bore readers with arithmetic. However, English is well equipped with words: "almost", "essentially", "basically" and Yglesias can think of dozens more I'm sure, which are well suited for avoiding the risk of boring his readers with arithmetic without writing something which is false.

However, my complaint about math intollerance at The Atlantic is due to the fact that I tried to post "11<12" as a comment and got an error message. Furthermore when I previewed my post to see the problem it had become "11." The Atlantic web server has stolen my < sign and I want it back. It's not just for computers, people sometimes use it too.

OK OK I admit the benefits of html and the web and stuff are > the cost of this minor irritation, but why did blogger accept my post when The Atlantic didn't accept it as a comment ?