« Previous Day | Main | Next Day »

3 April, 2003

Off to the Desert

This is my last post for a few days. We'll be spending the weekend in Borrego Springs. With any luck, I'll return in a few days with some great photos of desert wildflowers.

Posted on 3 April, 2003

Microsoft Sees Google as a Competitor

Interesting article: Microsoft eyeing Google's territory. Bob Visse, Microsoft's marketing director for MSN said:

"We do view Google more and more as a competitor. We believe that we can provide consumers with a better product and a better user experience. That's something that we're actively looking at doing." 

They must have something up their corporate sleeve. Remember this amusing quote made by Microsoft's Jim Allchin is a few months ago?

"Google's a very nice system, but compared to my vision, it's pathetic."

I can't wait to see Jim's vision materialize, and it will be interesting to see how Microsoft defines a "better user experience." I suspect the search screen will be filled with JavaScript-powered menus, pop-up results screens, and lots of flashing ads. And, of course, users will need to sign in with Passport before they can do anything.

But wait a minute! Microsoft already claims that More people use MSN Search than any other search service. If that's really true, then why are they so concerned about a pathetic search site like Google?

Posted on 3 April, 2003

Windows Media Player at the Movies

Landmark Theatres and Microsoft announced that they are equipping 177 screens in all 53 Landmark Theatres across the U.S. with digital cinema playback systems based on Windows Media 9 Series. Details here.

Great! That give me an opportunity to recycle this image.

Posted on 3 April, 2003

Glowing Ice Cubes

I've never seen these before: Litecubes.

Litecubes are realistically designed one-inch clear plastic blocks that resemble frozen ice cubes. What makes the product unique is a battery-operated light inside the Litecube that illuminates the cube (and whatever it is submerged in) from the inside out.

Posted on 3 April, 2003

Blues Lyrics and Hoodoo

If you're a blues music fan, you must go here.

Blues fans who listen closely to song lyrics often wonder what a "mojo" is or why one would carry a "John the Conqueroo." Most people know these terms have something to do with folk-magic, and they may even have run across interviews or songs in which a musician uses the word "hoodoo," but many incorrectly assume that the terms come from the Haitian religion known as Voodoo or Vodoun.

(via Bifurcated Rivets)

Posted on 3 April, 2003

The BSA and Statistical Fallacies

The Business Software Alliance (BSA) released a new study titled Expanding Global Economies: The Benefits of Reducing Software Piracy. The study was authored by IDC. You can read the 25-page report here (click the "white paper" link to view the PDF file).

As I explain, this study is seriously flawed. The authors make an assumption that pretty much invalidates all of their conclusions. Even a beginning statistics student would know better.

The study examined software piracy rates in 57 countries, and also looked at economic aspects of the IT industry in each of those countries (spending, tax revenue, employment, etc).

The basic premise is as follows:

  1. IT spending is a good thing because it stimulates economies.
  2. Countries that spend a lot on IT tend to have lower software piracy rates.
  3. Therefore, if piracy rates are reduced, IT spending will go up.
  4. Increased IT spending will cause good things to happen throughout the world.

The report contains several charts, and one of them is reproduced here. This chart plots the 57 countries using two variables: IT Taxes as a share of GDP (vertical axis) and Annual Piracy Rate (horizontal axis). The data presented are six-year averages.

As you can see there's an inverse correlation between these two variables. Countries that collect more IT tax revenue tend to have lower software piracy rates. This is an interesting finding but, if you think about it, it's not at all unexpected.

Here's the fatal flaw: The authors fall for the old "correlation implies causality" fallacy. In other words, the study assumes that lowering the software piracy rate in a country will magically cause growth in that country's IT sector -- which will in turn result in other economic benefits for the country. A quote:

A country's software piracy rate is a key differentiator between countries that enjoy vast IT economic benefits and those that have yet to unleash them.

That statement may certainly be true, but there is absolutely no foundation to take this giant leap and assume that lower rates of software piracy is what causes the IT economic benefit. It's very likely that both of these variables are influenced by other factors (for example, general economic stability). Or, maybe they got it backwards. Is it possible that growth in the IT sector may be responsible for a decline in piracy rates?

And another quote:

This new data reveals that the economic impact of even a slight reduction in the amount of pirated software can be significant - helping to accelerate IT sector growth. It demonstrates how even a modest and achievable 10-point reduction in software piracy rates can be a powerful tool for delivering jobs, tax revenues and economic opportunity.

Anyone who has received higher than a "D" in an introductory statistics class knows better. A high correlation between two variables definitely does not imply that one causes the other.

The researchers fed a bunch of numbers into their economic model, and then came up with news-worth conclusions such as:

According to the IDC analysis, a 10-point drop in the worldwide piracy rate from 40 percent to 30 percent (2.5 point decrease a year from 2002 - 2006) would create an additional 1.5 million jobs, generate an additional $64 billion in tax revenues, and contribute another $400 billion to economic growth.

What about the data? The study uses BSA's piracy data, which is estimated based on many assumptions . In addition the uses economic data collected from 57 countries. Can you imagine the difficulties in trying to get consistent (and comparable) data from 57 different countries? The study, however, does address this problem:

Here's a key assumption in the study:

The study calculated different effects from piracy on software, services, channel spending, employment, and tax revenues. In the case of software, we used a linear relationship between a lowering piracy rate and growing software spending. (E.g., if a country has a 50 percent piracy rate and $100 million software spending, lowering the rate to 0 percent would create a theoretical $200 million in software spending.).

That, of course, makes no sense. And I see no mention of the impact of open source software. (I discussed BSA's piracy methodology last November). But this time they attempt to rationalize this assumption:

While not every piece of formerly pirated software will be purchased if piracy rates go down - some will be substituted, some not used - at the same time lower piracy rates yield more economic activity that stimulates more software production and purchase. The two countervailing forces seem to cancel each other out. This is the conventional assumption for most previously published piracy studies.

It's a huge assumption -- and one that I don't buy into.

As I've stated in the past, I certainly don't condone software piracy. But I think it's important to get the facts straight and understand what's really going on.

This type of research doesn't come cheap. The BSA paid lots of money to IDC to conduct this study, and I suspect that the conclusions were known in advance. The end result is that their preconceived notions now appear to be based an actual scientific study. High-priced propaganda.

In the final analysis, the study is flawed, and the conclusions are based on some fancy number-crunching, lots of assumptions, some dubious estimates, and a glaring statistical fallacy. The intended audience for this report is the press -- the folks on a tight deadline who will talk about these numbers out as if they are accurate. The same press who probably will never read past the executive summary.

Posted on 3 April, 2003

More Cell Masters

In February, I became an Excel Cell Master, thanks to Colo. Colo is from Japan, and runs an Excel site. Each Cell Master answers a series of questions, and then nominates the next Cell Master.

Since my induction, several others have received this honor. If you hang out in the Excel newsgroups, most of these names will be familiar.

  • Andy Pope: I've never met Andy, but we've been in touch via email on several occasions. And his web site has some incredible downloads. My favorite is his simulated 3D scatter chart.
  • Jon Peltier: Jon is very active in the Excel charting newsgroup, and he was also the technical editor for my Excel Charts book. I had the opportunity to meet him at the most recent MVP Summit conference in Redmond. He's not as nerdy as I thought he'd be. He has lots of great chart examples at his web site. Maybe some day he'll abandon Geocities and get a real web host.
  • Debra Dalgleish: I first met Debra at the MVP Summit in 2001. Once I got to know her, I realized that she has the best sense of humor of all the Excel MVPs. She lives in Canada, but doesn't eat poutine. Her web site has a nice collection of Excel tips.
  • John McGimpsey: John goes by the initials J.E. I met him at the last MVP Summit, but didn't get to spend much time with him. He preferred to hang out with the Mac people. As far as I know, John doesn't have a web site.
  • Monika Weber: I also met Monika at least year's MVP Summit. She's from Switzerland, and is a fellow computer book author. Monika seemed very self-conscious about her English. Frankly, she spoke it well (better than many Americans) and I understood 95% of what she said. Monika is a fun and upbeat person. And she takes lots of digital photos. Her web site is in German.
  • Bob Umlas: I first met Bob two years ago at the MVP Summit. Bob is your typical Excel / accounting guy -- except that he also has a sense of humor (sort of). Bob has an uncanny ability to delve into the hidden nooks and crannies of Excel, and pull out some weird stuff. Bob is responsible for most of the entries at my Excel Oddities page.

Posted on 3 April, 2003

The Skyscraper Page

A web site about... skyscrapers.

Posted on 3 April, 2003

Geisha Photos

Lots of photos of Japanese geisha.

A month ago, I would have passed right by this site. But I recently read Arthur S. Golden's excellent novel, Memoirs of a Geisha, and now these photos are very interesting to me.

(via Speckled Paint -- not via Geisha asobi blog)

Posted on 3 April, 2003

Faked War Photos

On Monday, the L.A. Times published a front-page photo that had been manipulated in Photoshop to make it more dramatic.

You can view the before-and-after photos here.

Posted on 3 April, 2003

The Iraqometer

War stats at a glance: bombs dropped, civilian casualties, WMD sites uncovered, oil wells aflame, coalition fatalities, leaflets dropped, territory control, Iraq soldiers surrendered, cost per taxpayer, and leadership status.

Posted on 3 April, 2003

Still More on Office 2003

Yesterday I cited an article that describes the various product configurations for the upcoming Office 2003 -- including the new pricing for the Student and Teacher Edition.

The Student and Teacher version of Office XP was apparently viewed as a "trial" to see how it would sell. The Office Student and Teacher Edition sells for about $129, which is about $300 less than the equivalent Standard Edition. And what about sales? According to this CNET article, the Student and Teacher Edition...

...continues to outsell the standard version by a huge margin. For boxed productivity software sold at retail, the student and teacher version has 34.4 percent market share versus 4.7 percent for Office Standard, according to NPDTechworld.

The article also includes a quote from Microsoft's Office Product Manager Simon Marks:

"In the trial we had that strange licensing, where only the person licensed to use (the student and teacher version) could use it in the household," Marks said. Market research determined this approach simply wasn't sensible in practice."

C'mon Simon. Does Microsoft really need to pay for market market research to determine that their licensing policies are absurd? I would have told you that for free.

Posted on 3 April, 2003

An Amusing Headline

This is from a story at Slate.

(via dive into mark)

Posted on 3 April, 2003