Moderate Muslims, everybody.

Well, I gave him a chance to answer the question himself. If he has a different answer, let's see it.

I figure we round 'em all up and divide them into two teams. The heads of Team A get shoved up the asses of Team B. Then the heads of Team B get shoved up the asses of Team A. Then we roll 'em down a beach and into an ocean of pig shit.

I guess you'd like me to care about your opinion on this topic. Do ya think I'm gonna?

The Kurdish approach seems to be to turn the militants into corpses and then dispose of them with front end loaders. It seems pretty effective. Daily Mail link

The Administration approach is to try and shame the militants with hashtags.

Hey, I figured out what Ricky Retardo's reward from the Islamists will be. They'll name a shallow mass grave in his honor.

That's actually quite the feather in one's cap, I'm sure. Color me peanut butter and jealous.

Ooooh, the Islamophiles really didn't like that one!

No, next comes defeating historians saying he was a decent president, defeating his presidential library, defeating his aircraft carrier, defeating him being in public school history books without "Kenyan pretender", next to his name, and defeating his grave being pissless.

There are radical muslims and moderate muslims: A radical muslim wants to behead you, a moderate muslim wants a radical muslim to behead you.

HA! Sorry kids but the truth is a hard pill to swallow.

Pat Condell once again knocks it out of the park and then breaks the bat over his knee:

If VisionCastle likes that he should watch the videos where he destroys the American conservative movement, tea baggers, and especially the religious right. Oh, he also has crusified Fox News for being brain dead propaganda watched by nothing but drooling idiots.

Hint: He is not on your side. I do think he was spot on with that video though.

I don't care whether he's on "my side" or not. What somebody is correct about, they're correct about. See, one important skill people develop as they mature is the ability to separate the claim from the claimant. Seems to be a skill that's sorely lacking from some posters at Teh Wrodfrodge, however.

http://www.newsweek.com/hundreds-turn-out-funeral-copenhagen-gunman-308840

This is the sort of shit that really pissed me off. 1000 people, just about all young muslim men, showed up in Denmark to celibate the life and actions of the piece of shit terrorist who went on a rampage murdering people in Denmark.

They should have detained every single mother fucking last one of them for suspecious of terrorism, gotten their names and finger prints, and even you can't officially arrest them you can at least get their names, finger prints, and put them on terrorism watch lists. You never know, you might get lucky and find some illegal aliens, people with outstanding warrants, or at least put fear into these terrorism supporters.

Oh, and over 130 muslimsfrom Denmark have gone to fight for ISIS and the majority of them came from the Mosque this guy attended and where they celebrated this shitstain's terrorism.

Those numbers are pretty bad. For every 100,000 Muslim immigrants, you get:

78,000 who are offended by images of Mohammed.
27,000 who have sympathy for the motives of the Charlie Hebdo attackers.
20,000 who think the West and Islam are not compatible.
11,000 who have sympathy for those who fight against the West.
7,000 who don't think they should obey British laws.
5,000 who feel no loyalty to Britain.

To be offended isn't a problem. The point of freedom isn't that nothing can offend me, but that it is allowed to offend me.

(Also, gturner doesn't understand how statistics work, but that's no surprise.)

So you can't extrapolate from a small sample size to a larger one?

You can if you know how and when to do it. Do you?

Only very, very carefully and with extremely selective sampling. The more you want to extrapolate, the greater number of independent variables you need in the sample you're extrapolating from. Even then, confounding variables multiply exponentially the further you extrapolate from your original sample size. And if you're starting from anything other than a random sample to begin with, forget it.

True. And pertinent!

So when a poll shows that 78 percent of people in a group feel a certain way, you can't say that 78 percent of people in a certain group feel that way, as in 78 out of 100, or 780 out of a thousand, or...?

No, I'm saying that if you want to extrapolate from 78 in 100 to 780 in 1,000, the 100 and 1000 sample populations must match identically. They have to be identical in every respect. Age, gender selection, income, occupational diversity or similarity, political leanings, geographical area, and everything else. With that last one, geographical area, do you see where extrapolation becomes an issue? The larger sample size has to match the original in every conceivable way in order for the statistics to even slightly hold to the original sample's -- and the bigger you blow up the resulting extrapolation, the more and more unlikely you are to satisfy that condition.

Side note. Here's the fun thing about writing lazy college students' homework papers for them (including papers on Statistics and Probability): Not only do you get college education you can't pay for on your own -- you get paid to get a college education.

Some of the main problems with the "extrapolation" above include:

1) The original sample wasn't random but weighted. The extrapolation ignores this.

2) The extrapolation focuses on the small minorities, but ignores standards of significance for those numbers. There's a reason why a statistical study chooses to say "95% said A" and not "5% didn't say A".

3) The original sample referred to "British Muslims", the extrapolation changes this to "Muslim immigrants". These aren't just not identical, they are completely disjunct groups.

Besides the purely mathematical problems, there's a lot of further equivocation. For one obvious instance, 7% didn't say they shouldn't obey any British laws, but a weighted 7% didn't agree with the idea that they should always follow all British laws. Compared to the 30% among the general population that use illegal drugs (source) and 15% who engage in illegal downloads (source), that would make British Muslims unusually law-abiding.

Right. Nonprobabilistic samples can't be extrapolated from. Period.

Not always true, but where true, this is a serious compromise of the extrapolation. It's a little cynical to assume that any and every extrapolation is going to focus on the outliers rather than on the results within one standard deviation -- but it's an article of faith, and articles of faith do not belong in science.

Well, not completely disjunct -- "British Muslims", considering the tidal wave of Muslim immigration to the U.K., legal and illegal, are bound to be comprised at least in part by "Muslims immigrants.". The problem is in untangling one variable from the other. Hard enough when you're dealing with the original sample, and when it comes to an extrapolation -- again -- forget it.

And that speaks to a confounding variable: How many British Muslims would refuse to obey British laws because they are Muslims, and how many refuse to obey British laws because they find any and all laws objectionable, regardless of religious or ethnic identification?

Alrighty then, Britain has absolutely no problem with their Muslim population, 27 percent of whom sympathize with the reasons for the Hebdo massacre of French cartoonists, because 27 percent of 2.78 million Muslims is only 0.75 Muslims, not a whole one, and it takes a whole Muslim to carry out such an attack.

I'm not saying it's not a problem. I'm saying that if you're going to explain what the problem is, you're going to have to do it from an angle other than statistical prevalence.