Thursday, September 07, 2006

Twice as bad

In a piece today on Comment is Free, Polly Toynbee writes:

As for these polls, just remember that the same voters who say they don't want Blair to go yet are the ones who give him satisfaction ratings twice as bad as when Mrs Thatcher was toppled

It's not the same voters 16 years on, of course, but more interesting is the "twice as bad" bit.  Thatcher left on 22 November 1990, and MORI polls show that in a poll carried out between 15 and 19 November 1990 she had 25% satisfaction and 71% dissatisfaction (source).

Twice as bad as 71% disapproval?  What's that, then, 142% disapproval?  Actually, the latest MORI figures (same page as above) give Blair a 23% satisfaction rating, and 67% dissatisfaction.


sean said...

It's your maths that are at fault. A satisfaction rating that is "twice as bad" would require you to half the number of people that state they are satisfied - not double the number who are disatisfied. Duh!

FactcheckingPollyanna said...

Sean --

You could make the argument both ways, of course -- and the original statement is not a helpful way of expressing it. FOllowing your logic, a 1% approval rating would be "twice as bad" as a 2% approval rating, which doesn't feel particularly helpful to me.

Even on your definition, though (and bear in mind that this is not maths in the sense of arithmetic but rather the interpretation of words), I don't accept that 23% is half of 25%.

Nice attempt to be patronising, though. You just keep working on that.

sean said...

Maybe this is more about smenatics than maths. While 1 is 50% worse than 2, 2 is 100% better than 1.

You are correct of course that Polly was wrong to suggest 23% is twice as bad as 25% - perhaps she was being rhetorical, rather than factual? But you were pretty petty in suggesting that she had implied that 142% of the population - an impossibility - was disatisfied.

So: why so smug?

FactcheckingPollyanna said...

Sean --

I would certainly agree that Polly was not being factual when she said "twice as bad".

The reason why I semi-facetiously asked about the 142% is that both interpretations of the "twice as bad" (i.e. dissatisfaction twice as high or satisfaction half as high) are -- I think -- valid, and so both needed to be shown to be untrue for me to show that Polly was not being factual.

Of course 142% dissatisfaction is an impossibility -- I was hoping people would realise that without it needing to be spelled out. I was relying on reductio ad absurdum to make my point. You might find it petty, but it is a logical tool with quite a fine pedigree.

As for my lightly patronising tone, think of it as an allergic reaction to having my maths questioned when, as you concede, the argument isn't really about maths, and to the use of the word "Duh!", which I found to be quite a poor attempt to be patronising.

sean said...

Hey - you're the one who set yourself up as a "factchecker". You then twisted a rhetorical comment about a "satisfaction" rating by doubling the "disatisfaction" rating. This resulted in you distorting the facts far more than the subject of your once unhealthy obsession.

FactcheckingPollyanna said...

Sean --

I do not see this as a distortion at all. "Twice as bad" could legitimately mean halving the satisfaction rate or doubling the dissatisfaction rate -- I think both are legitimate interpretations, though of course quite different.

I showed that both were wrong.

The fact that you view one possible interpretation as a "distortion" is interesting, but given that it is a possible and legitimate interpretation it was necessary to debunk it.

You could dismiss "twice as bad" as a rhetorical comment, but it is clearly not factual under either interpretation, is it?

sean said...

Hmmm. Are your pedantic witterings on this topic half as good as you think or twice as bad as I think? As you would no doubt say - it can be argued both ways...

FactcheckingPollyanna said...

Sean --

You tried being patronising, and cried foul when I patronised you back.

When I try and argue using logic you accuse me of pedantic wittering.

You accuse my maths of being faulty, then concede it is about semantics, and when I point out that I was merely exploring both legitimate interpretations of the comment and showing both to be wrong, you dismiss the comment as being rhetorical rather than factual.

I suspect that you are not actually trying to engage in a debate but rather twisting your argument every which way you can to try to catch me out. Like I say, doubtless entertaining for you, but not very insightful or helpful for me.

sean said...

Hey babes, please calm down.

When your facts are challenged your response is: it can be argued both ways... That is only if you are unwilling to accept any responsibility for what you have to say, or are too cowardly to hold an honest opinion.

At least you have admitted that your version of facts are actually "semi-facetious" absurdities. It is easy to carp at social commentators like PT, much more difficult to win hearts and minds.

You're a clever pedant - granted (though a pedant should know that a double negative is always unacceptable - re your 142%).

Sadly you inhabit a soulless world. It must be a lonely place.

Ciao for now!

FactcheckingPollyanna said...

Sean --

What facts do you think you've challenged exactly?

I didn't say "my" facts are semi-facetious. In fact, you'll note I said I was semi-facetious when I asked a question. The two things are quite different.

And it's always calm in a soulless world. Partly due to the lack of ad hominem abuse.

Anonymous said...


Anonymous said...

Do either of you clowns have the actual statistics to hand that Polly George Orwell was referring to?

The Happy Warrior said...
This comment has been removed by the author.
The Happy Warrior said...

I've got to congratulate Sean on starting a profoundly stupid argument on a nothing point and then managing to drag out his self-justification for quite as long as he did before finally buggering off. It was indeed an impressive feat of pedantry.

P.S. I hope no one notices I ended a sentence with a preposition.

Anonymous said...

Old I know but I didn't notice the date till I was about to post, so dear nobody...

Old I know but...
It seems fairly obvious that "satisfaction ratings twice as bad" means that the satisfaction ratings for Tony Blair were "twice as bad" as the satisfaction ratings for Thatcher, not that the dissatisfaction ratings were twice as high.
"twice as bad" may be a poor way of expressing it, but in a poll about satisfaction we can probably all probably agree that 0% satisfaction is bad and that 100% satisfaction is good.
"twice as bad" on such a measure of satisfaction would seem to be half of the satisfaction rating.
Toynbee expressed it poorly, Factchecking Pollyanna willfully interpreted it incorrectly to point out a mistake that didn't exist.
Factchecking Pollyanna should have pointed out that "I don't accept that 23% is half of 25%"

FactcheckingPollyanna said...

Anonymous --

I am not sure I quite understand what the "mistake that didn't exist" is -- there is no interpretation of the data in the original post which shows that the satisfaction ratings were "twice as bad". Not one. That is a mistake which did exist.

And I am not quite sure why you point out that I should have said something which I, er, did in fact say.