Irrational Fear as Law

I noticed in my normal news feed that Obama will ‘evaluate’ bill to ban online munition sales. We have to recognise that many of our laws are enacted to satisfy irrational fears associated with risk.

Billions of cartridges are manufactured in the U.S.A each year. For example, about 4 billion .22 rimfire cartridges are produced per year.

We must ask, of all the cartridges made, what percentage of those used in violent crimes are purchased online?

The violent crime rate in the U.S. is about 403 per 100,000 people. Then we have to subtract those crimes that do not involve a firearm. Then we have to subtract those in which the ammunition was NOT bought through online sources.

If we can then properly quantify how many violent crimes were committed by someone with a firearm while using ammunition bought online, then we can calculate the percentage of all ammunition manufactured that was sold online and then used in violent crimes.

I can guaranteed that the percentage will have a lot zeroes after the decimal. This would be policy by irrational fear. Fear of a risk that is not worth considering.


Liars, Damn Liars, and Statisticians

While researching an issue surrounding violence to women in a particlular neighbourhood, I came across an article titled, Men who buy sex tend to commit more crimes, study says. This is a typical use of statistics to lie.

Careful examination of the article reveals the sample is not representative. The respondents were only people who read “a local free daily newspaper”. This can create all kinds of biases in the resulting data, both deliberate and not.

Next, one has to consider the medium reporting the “study”. In this case the popular media. Media outlets usually report things that they feel will raise readership and are not above tweeking the content to make it more sensational. The author is described as a ” anti-prostitution activist” so what would you expect the conclusions of the “study” to be? In this case, the study states that men who are open about their use of sex-workers are more likely to admit that they’ve committed crimes. This merely means that this particular group of customers are likely to be more open about their misdeeds. The control group could easily have much less propensity to be open. How can a corelationship be established if the propensity to be open about past misdeeds is not the same?

These types of “studies” are just an attempt to support a pre-existing belief or world view and offer little in the way of hard data.


Zanran for Numerical & Graphical Data

An excellent article about a beta search engine with promise.

Zanran – a new data search engine

… a new data search engine called Zanran – that focuses on finding numerical and graphical data.

Zanran focuses on finding what it calls  ‘semi-structured’ data on the web. This is defined as numerical data presented as graphs, tables and charts – and these could be held in a graph image or table in an HTML file, as part of a PDF report, or in an Excel spreadsheet. This is the key differentiator – essentially, Zanran is not looking for text but for formatted numerical data.”

Disgruntled Employees Outdo Terrorists

Two articles on the Brand Killer Robots blog drew my attention. Not because the data offered anything new, but that Stephen Ryan was able to create a bot to clearly show that insiders, employees, and former employees are the most likely to launch cyber attacks.

Raps BOT : Predicts Insider Cyber Terrorism Threat HIGH

Raps Bot : Sniper Attack Methods – Number 1 Cyber Terrorism Threat

Don’t Believe Everything You Read

Misleading RCMP data undermines counterfeiting claims by Michael Geist

“The RCMP has been the single most prominent source for claims about the impact of counterfeiting in Canada since its 2005 Economic Crime Report pegged the counterfeiting cost at between $10 to 30 billion dollars annually.”

“Responding to an Access to Information Act request for the sources behind the $30 billion claim, Canada’s national police force last week admitted that the figures were based on “open source documents found on the Internet.” In other words, the RCMP did not conduct any independent research on the scope or impact of counterfeiting in Canada, but rather merely searched for news stories on the Internet and then stood silent while lobby groups trumpeted the figure before Parliament.”

Big Numbers – Big Confusion

The nomenclature of big numbers varies depending on who writes them.

In the USA, a billion is 109 (a 1 followed by 9 zeroes) and a trillion is 1012. This seems straight-forward until you realize that Americans spell differently than the British and other Commonwealth countries. Some further thought might be in order.

In the UK and Germany, a billion is 1012 and a trillion is 1018. Now this makes more sense to me.

Bi means twice or 1 million squared for a billion, or 1012, as 1 million is 106. Tri is three times, or 1018, or 1 million cubed. The same follows for quadrillion being 1 million to the 4th power, or 1024, rather than the American 1015.

This confusion means having to get clarification on all large numbers unless all the zeroes are present, or they come from a recent article written by a bona fide Economist or Statistician who will use the American interpretation of these large numbers.

Now I’m sure you’ll never be confused by this again!

May the Farce Be With You

A frivolous email campaign during 2001 to have people record their religion as “Jedi” on the national censuses of Australia, Canada, New Zealand, England, and Whales was surprisingly effective in distorting the census results.

This campaign succeeded in having 70,000 people in Australia record their religion as Jedi, 20,000 in Canada, and 53,000 in New Zealand. In England and Whales 390,000 stated their religion as Jedi, surpassing Sikhism, Judaism, and Buddhism, and making it the fourth largest reported religion.