Saturday, March 25, 2006

Let Them Be Offended

Anyone who knows me only from this blog might be surprised to discover that I am a fan of South Park. I am. There's no denying it.

Cue Stan Marsh:
Daaaaaad, Tom Cruise won't come out of the closet!

The whole business of whether or not Isaac Hayes actually quit South Park raises a point or two. Most of all in my mind is the question of whether Scientology is a cult. I'm pretty unambiguous on that question but people I respect argue that saying so is offensive.

Here's my spin on it: If people in your religion feel comfortable speaking on your behalf in a legal or business matter as though they had power of attorney, then you belong to a cult. Now it's not clear who is speaking for Isaac Hayes, but I'll bet ten dollars to a doughnut that that's what happened.

It occurred to me right after discussing this with a friend that the more publicly prominent sects of the religion of the alleged prophet Mohammed are also cults. Is that offensive? Well, then let them be offended.

Sunday, March 12, 2006

Gumbo of the Mind

I like gumbo, really I do, especially Emeril Lagasse's recipe which I have cooked up dozens of times. It's mmm mmm good. Gumbo involves throwing a lot of differnt things in a pot and applying heat. This is a great way to make food, it is a terrible way bring enlightenment to consumers of news.

This article: Negative Perception Of Islam Increasing is intellectual gumbo, and it doesn't have any of the charm of the other kind. [Cascading hat tips to Charles at Little Green Footballs and his referrer.] In fact it is as bland as store-brand mac and cheese.

Seriously, given what's been in the news for the last 4 years, who expected that the American people's impression of islam was going to become more positive? Should we expect it to inspire respect that islam's adherents commit barbaric acts around the world on a daily basis? The hate-America crowd has proffered the thesis that if Americans understood islam they would be more accepting of it. I have found the opposite to be true, except among leftists and muslims.

Polling data is so easily manipulated that this number is effectively arbitrary, and by throwing "all" Americans [well... the pollsters proprietary sample] in to the pot you get a sort of bland mix whose movements are difficult to interpret.

Here are some polls I suggest be taken and which I argue would give a more revealing portrait:
  • Poll Sikhs in America and report what percentage agree that islam is a "religion of peace".

  • Poll Hindus in America and report what percentage agree that islam is a "religion of peace".

  • Poll Thai and Indian Buddhists in America and report what percentage agree that islam is a "religion of peace".

  • Poll Zoroastrians in America and report what percentage agree that islam is a "religion of peace".

  • Poll Pakistani Christians in America and report what percentage agree that islam is a "religion of peace".

  • Poll Assyrian, Maronite and Coptic Christians in America and report what percentage agree that islam is a "religion of peace".

  • Poll Armenian Christians in America and report what percentage agree that islam is a "religion of peace".

  • Poll Nigerian and Sudanese Christians in America and report what percentage agree that islam is a "religion of peace".

  • Poll first-generation Serb and Croat Christians in America and report what percentage agree that islam is a "religion of peace".

  • Poll first-generation Greek Orthodox Christians in America and report what percentage agree that islam is a "religion of peace".

  • Poll American Jews with immediate family members living in Isreal and report what percentage agree that islam is a "religion of peace".

  • Since the US Department of State has recently condemned deeds "offensive to the beliefs of Muslims", poll female employees of the US Department of State and report the percentage who have started wearing burkas to work in order to avoid offending muslims.

For info on world opinion I would also:
  • Poll Christians on Mindanao and report what percentage agree that islam is a "religion of peace".

  • Poll London tube riders and report what percentage agree that islam is a "religion of peace".

  • Poll Madrid commuter-train riders and report what percentage agree that islam is a "religion of peace".

  • Poll non-muslim Australian beachgoers and report what percentage agree that islam is a "religion of peace".

  • Poll Balinese night-club employees [well ok, ex-employees] and report what percentage agree that islam is a "religion of peace".

  • Poll Timorese Christians and report what percentage agree that islam is a "religion of peace".

  • Poll Dutch MP's and report what percentage agree that islam is a "religion of peace".

  • Poll French auto-insurance claims adjusters and report what percentage agree that islam is a "religion of peace".

  • Poll Danish cartoonists and report what percentage agree that islam is a "religion of peace".

When I see some of these reported I might start believing that the news-show-business is interested in understanding the issue.

Saturday, March 11, 2006

John Zogby Can Go Sit on a Poll

The pollster who gave us the '04 presidental polling data from the planet Bizzaro recently released a poll which the news show biz was quick to trumpet as some sort of evidence that American soldiers are demoralised and need to be withdrawn immediately.

If you've not read my previous entries on the subject of opinion polling, now is the time for some background reading: Part I, Part II and Part III.

Hugh Hewitt had John-of-a-hundred-evasions on his program when that story was in the headlines. Here is the transcript and audio. When you read that you should be thinking "proprietary data reduction" and "who is the customer?" When the customer is an America-must-lose individual or organisation and Zogby is the keeper of the secret data reduction method we have every reason to be very very very skeptical. Hugh tries to shine light on this, but doesn't get past square one. JZ's evasiveness only compounds my suspicion that this is yet another of his Bizzaro-world data sets.

Pollsters have to please their customers. That much I believe we can confidently say Zogby understands. I suggest that he also approaches this sort of fraud with enthusiasm.

Wednesday, March 08, 2006

The Gitmo Project

I would like to thank Captain Ed for the opportunity to be involved in his blogswarm. I read one section of these documents and learned a number of things about the detainees as well as about al-Quaeda and the Taliban. If my sample can be taken as representative then:
1. Everyone in Gitmo is apparently a driver or cook. There are no actual Taliban nor al-Quaeda fighters there. So clearly we let all the ones with the guns and the hostile intent get away.
2. No one ever joins the Taliban or al-Quaeda voluntarily. It operates entirely by impressment of innocent drivers and cooks.
3. Lots and lots of people from arab countries travel to the northern border of Pakistan with no intent whatever to ever cross in to Afghanistan. That only ever occurs by accident and besides no one really knows where the border is up there.
4. Muslim brotherhood only extends to other racial groups in times of peace. The rest of the time muslims of one race will enthusiastically deliver muslims of other races over to the local authorities who will, in turn, enthusiastically torture same and force confessions out of them.

The big problem with these documents is that they are mostly the detainee's statements. Overall the experience I would compare to reading only the defense arguments in a court case. If you only read defense arguments you are likely to decide that most defendants are innocent. Only some of those detainees are foolish or ignorant enough to incriminate themselves. The remainder at least seem to know what to deny.

Nevertheless, 9 out of 10 of the cases I studied were of a sort that I am confident in saying that the US was wholly within its rights to detain them as enemy combatants at the time of capture. Whether they continue to represent a danger is a more difficult subject and I would not like to have to give an assessment of that without seeing the equivalent of the case for the prosecution.

We mustn't forget is that these are not domestic civilian criminal cases. Detention of enemy combatants is not criminal justice. The rules are different and the purpose is different. The standards for releasing them have to be different as a consequence.

One last thought: there is a recurring theme in these cases that could go unnoticed, but it is along the lines of "whatever you do don't turn me back over to the Pakistani/Iranian/Algerian authorities. You don't understand what they're like." What do you think that could mean?

Friday, March 03, 2006

Corporate Cost Savings

To: Board of Directors
The Big-Stupid Corporation
Chicago, Illinois

From: A. Engineer
Systems Engineer
The Big-Stupid Corporation
Government Widgets Division
Phoenix, Arizona

Re: Enterprise-wide Cost Savings Proposal, "Rule #1 Training"

Dear Sirs,

I would like to propose a consolidation of our Enterprise mandatory training programs. There are currently 31 such courses with titles such as "Laptop Security", "Safe, Smart and Secure", "Industrial Security Refresher", "Computing Security", "Acceptance of Business Courtesies, Anti-Kickback", "Truth in Negotiations", "Procurement Integrity", "Trade Secrets and Proprietary Information", "Offering Gratuities", "Compliance with the Foreign Corrupt Practices Act (FCPA)", "Foreign Office Export & Import Compliance Guidelines", "U.S. Import Compliance Awareness", "Export Awareness Overview Training", "Antitrust Compliance", "Insider Trading", "Fair Disclosure Awareness", "Lobbying Awareness and Reporting", "Recruiting and Hiring Government Employees - Conflict of Interest (COI)', "Privacy Directions Awareness", "Ethics Code of Conduct", "Ethics Challenge", "Ethics Recommitment", "Ethics and our Business", "Ethics Orientation" and so on. The structure of these courses is very much the same. Employees are shown the guidlines, and then some hypothetical scenarios and then asked what the correct behavior is to deal with that scenario in a manner consistent with the policies of The Big-Stupid Corporation.

Each employee must spend 1-2 hours in each of these courses which is purely an overhead charge to corporate budgets. I would like to suggest to you a means of consolidating these courses which can be enacted across the entire enterprise at a savings of at least 20 hours of labor per employee. Given our labor costs and the number of employees in this corporation I estimate the savings at roughly 100 million to 200 million dollars per year for this proposal.

Here is the point, the content of all these courses is roughly the same. There is a single priciple which, if applied diligently can answer all the quandries presented in these courses. I call this principle "Rule #1" and I suggest that we replace most or all of our existing training courses with a single, enterprise-wide course. What is Rule #1? Simply stated Rule #1 is "Don't be a dumbass". If we implement "Enterprise Don't Be a Dumbass" training we will not only save huge sums cash but we will, I suggest, enter a golden age in this corporation as the smog falls away from the eyes of employees.

Let me give you some examples from our existing training courses to illustrate the efficacy of Rule #1 training:

Example #1
You are an employee involved in a contract proposal-writing effort and you receive and anonymous letter which proports to give you proprietary information about your competitor's proposal. How do you respond? By applying Rule #1 you immediately realise, "Don't be a dumbass, that's probabaly unlawful." and take the packet directly to the legal department with an explanation of how you obtained it.

Example #2
You are an employee with access to personnel files. You are going on a trip tomorrow and you are concerned that there will be nothing for you to do while waiting at the airport. You consider downloading some of the files to your laptop to bring along to the airport. However, Rule #1 reminds you, "Don't be a dumbass, downloading personnel files to your laptop's hard drive is probably a policy violation".

Example #3
You are a program director and you spend a lot of time with your Government contract monitor. Eventually she offers to sleep with you and asks that you give her daughter a job. If you have had Rule #1 training you realise "Dont' be a dumbass, sleeping with the customer whether in exchange for contract consideration or not is probably a criminal offense."

If my proposal is of some interest to you I can offer you a more detailed course outline and arrange to teach a pilot class for demonstration purposes. Please let me know your decision.

Sincerely,
A. Engineer

Thursday, August 25, 2005

Understanding BMI: Doublethink From the CDC

Compare the following statements from CDC's FAQ about the statistic known as "Body Mass Index" or "BMI".

Calculating the BMI is one of the best methods for population assessment of overweight and obesity.

versus
Most studies have examined the relationship between BMI and risk of disease. Therefore we do not know whether two people with the same BMI but different amounts of fat have different risks for disease.

So while they overstate the value of this statistic in the first statement they hedge by the time they get to question 3.

Dare I ask how both of these statments can be true?

I'm working on a table to show how silly this is, but here is an example:

A man is consdered overweight with a BMI higher than 24.9.

If he is 6'0" tall that means he can weigh no more than 183.3 lbs. without being classed as "overweight" by the CDC. He is "obese" at or above 220 lbs.

Yet to weigh as little as 136.9 is considered "normal". Normal?!? that's death-camp survivor range!

A man I know is a shade over 6 feet tall and his last reliable independent body composition measurement said he had 176 lbs of lean mass. A truly healthy body-fat proportion would be 10%. That would put his net weight at roughly 195.5. Thats well in to the CDC's "overweight" range. To be at the upper borderline of CDC "normal" he'd have to be at 4% body fat. Sorry folks, that really would be unhealthy.

That was three years ago. The subject has been working out since and has put on some muscle mass, what then? This isn't a pro athlete and this isn't a guy who pumps iron for 2 or more hours a day. This is an office-worker who does 10-20 minutes of daily calesthenics and can't stand to do anything physical outdoors from April through September [in central Arizona].

The answer is that the CDC and other BMI advocates are just crazy, and they're able to hide their insanity behind a calculation for which they know they dare not publish tables.

Wednesday, August 24, 2005

BMI: The Lie That Keeps on Giving.

Junk science fads come and go. This particular piece of trash is linked over at Hugh Hewitt's blog.

It is a pretty snazzy presentation purporting to document an "obesity epidemic" based solely on the rediculous statistic known as "Body Mass Index" or BMI. According to Mark Twain there are three kinds of lies: Lies, Damned Lies and Statistics. BMI is the latter two.

BMI is a ratio of Mass to the square of Height done in [of course] SI "units" kg/meter^2. That's all it is, Mass over Area. What does that tell us about obesity? Nothing at all. Really, it makes no discrimination between fat mass and non-fat mass. Someone please explain to me how a statistic that contains zero information about the subject's quantity of body fat can be used to deduce obesity? [Don't bother, that's rhetorical.]

Let me give you an example: when Arnold won Mr. Universe he was 6'0" tall, weighed in at roughly 225 lbs. and had approximately 4% body fat. That gives him a BMI of 30.5 at the time. Was he obese? According to the proponents of BMI, emphatically YES.

See the disconnect? Ever seen the Odessa/Permian football team? Guess what? Every strapping young man in their entire line count in the alleged "obesity epidemic". I call shenanigans!

I forgive Hugh for not being a scientist, but enough with this horse dung already. It doesn't take a professional medical researcher to see what a crock it is as an obesity statistic.

The obesity crisis has become the bandwagon for people who expect the government research grant spigot to be open wide over their mouths. Worst of all there are some who use this debate to advocate that the government impose fad pseudo-science "healthy eating" uniformly across our society. Such authoritarianism goes hand in glove with socialised medicine.

Such advocates are happy to blur the reality with misleading statistics because if they didn't they couldn't credibly claim any kind of emergency (national or otherwise) exists.