Feds Question Big Media's Piracy Claims 261
Posted
by
kdawson
from the show-me-the-losses dept.
from the show-me-the-losses dept.
WrongSizeGlass writes "CNET is reporting that the GAO's study of big media's piracy claims has raised some questions. (Here are the study's summary, highlights [PDF], and full report [PDF].) 'After spending a year studying how piracy and illegal counterfeiting affects the United States, the Government Accountability Office says it still doesn't know for sure.... The GAO said that most of the published information, anecdotal evidence, and records show that piracy is a drag on the US economy, tax revenue, and in some cases potentially threatens national security and public health. But the problem is, according to the GAO, the data used to quantify piracy isn't reliable.'"
Re:Not reliable? (Score:5, Informative)
More likely the RIAA/MPAA/BSA/TLA took a sampling of total users connected to a popular torrent and compared that to the total people downloading pirated material from BitTorrent and then extrapolated that to the entire US population to show that everyone's a pirate.
Re:Thinking about the popularity of D&D (Score:3, Informative)
It's a nerd-friendly example about questioning assumptions which GP is using to illustrate his point... about questioning assumptions, which is what the GOA report seems to be doing.
Re:The article (Score:5, Informative)
In what appears to be a setback for Hollywood and the recording industry, the government said that it sees problems with the methodology used in studies those sectors have long relied on to support claims that piracy was destructive to their businesses. The accountability office even noted the existence of data that shows piracy may benefit consumers in some cases.
Re:Self interest (Score:3, Informative)
TFR in TFA didn't come to the same conclusions as the RIAA. They mostly said "there aren't enough data to kinow" and pointed out a lot of stuff you've read at slashdot that points out that piracy may actually be helpful to the media industry.
They were, however, down on knockoffs, such as fake airplane parts and the like.
Re:Not reliable? (Score:5, Informative)
you have obviously never driven with a passenger who was whacked out of their gourd.
The criteria for the study was a BAC of >=0.05. I don't know what your friends are like when they drink, but mine have to drink considerably more than that to get "whacked out of their gourd".
It's about shoes, not music (Score:3, Informative)
Read the actual report. The big "piracy" problem is fake copies of shoes and handbags. That isn't even a copyright issue; that's a trademark issue.
You can legally copy garments; the only legal protection is for logos. So it's not even about the design.
Re:Not reliable? (Score:3, Informative)
It's usually a matter of tolerance. I asked a police officer what the highest BAC he had ever seen was. He told a story about pulling someone over for a broken taillight, she wasn't driving badly and he had no indication that she was intoxicated. When he reached the car he smelled booze and went through the process. She passed all of the field sobriety checks but failed the breath test -- with a 0.61!
A BAC of 0.4 is the LD50 for alcohol. Most people will pass out at BACs exceeding 0.2. Yet this woman was "sober" enough at 0.61 to pass the sobriety checks and operate a motor vehicle. Turns out she had a number of previous DWIs and a bit of a drinking problem. She had consumed alcohol for so long that her body built up a near super-human tolerance for it.
Re:It's about shoes, not music (Score:3, Informative)
Er no, designs are copyrighted.
You can't copyright a "useful article". That's the domain of design patents.
It's hard to get an enforceable design patent on apparel. Someone will go into the fashion library (a giant closet of famous garments) at the Fashion Institute of Technology and demonstrate that Coco Chanel did the same thing in 1931. Fashion is cyclical, not original. This is what resulted in the emphasis on exclusive "logos".
Re:Not reliable? (Score:3, Informative)
At least the Census numbers are close. They jive pretty well with the H&HS's survey, performed over the phone at regular intervals. 40 mil's closer than the other side's 15 mil at least.
http://www.politifact.com/truth-o-meter/statements/2009/aug/18/barack-obama/number-those-without-health-insurance-about-46-mil/ [politifact.com]
http://www.politifact.com/truth-o-meter/statements/2009/aug/21/orrin-hatch/who-are-uninsured-hatchs-take/ [politifact.com]
You know, I don't think this is the first time you've trotted out how wholly inaccurate that 42 mil number is either.
Oh, and the best part?
I don't trust these GAO piracy numbers. They are biased, nonscientific, and inflated. But sadly they'll problably be quoted by politicians during the election, in order to justify why the ACTA Treaty needs to be passed.
The GAO was studying the studies, and in an even bigger revelation, GAO agrees with you!
FTFA: "Three widely cited U.S. government estimates of economic losses resulting from counterfeiting cannot be substantiated due to the absence of underlying studies," the GAO said. "Each method (of measuring) has limitations, and most experts observed that it is difficult, if not impossible, to quantify the economy-wide impacts."
None of those studies put out were from the GAO, instead they were cited as coming from the FBI, CBP and FTC. GAO actually called them and asked for the report. All three came back saying, "what report?"
So we have a government entity saying these piracy reports are wildly inaccurate. They did something right. Is your mind blown yet?
Re:Not reliable? (Score:3, Informative)
I'm not sure what criteria they used. The standard rule of thumb that I learned says your body will eliminate approximately 0.02 per hour. For most males that corresponds neatly to the amount you wind up with from a single drink -- a drink being 12oz of beer (5%), 5oz of wine (12%) or 1.5oz of spirits (40%). Females will usually get a higher BAC increase out of each drink, owing to less body mass and a greater percentage of body fat.
Re:God. some common sense from an us. govt agency (Score:3, Informative)
My impression of the GAO (as an outsider) has been that it doesn't put up with much in the way of statistical bullshitting from anybody, and that as a government agency it does its job pretty consistently and well.
Re:Not reliable? (Score:3, Informative)
From the NHTSA notes in a FAQ found at the National Center for Statistics and Analysis:
"In order to perform complete-data analysis of FARS data with respect to alcohol involvement, the missing BACs need to be simulated (imputation!)"
And to explain Multiple Imputation (MI):
"Multiple Imputation is the state-of-the-art technique to impute missing values. Each missing BAC value is replaced by ten simulated values of BAC using rigorous statistical techniques that consider the interaction of all the characteristics of the case. MI allows for the computation of Standard Errors and Confidence Intervals."
Sorry, but it sounds like they just make up the missing data using really, really good techniques. But not any actual data, of course. Just other data that they think would indicate what the missing data wouldhave been, assumimg their asssumptions are correct.
That would be funny if it weren't serious. I think they just stated that to perform a complete data analysis, they have to 'make up' any missing data. How they do it doesn't impress me. If they use correlating data to make what they consider to be valid assumptions, then why bother to input the 'missing' data at all, indeed, just don't bother to use any of that value - just use the correlations and let the assumptions drive your analysis.
Which is, of course, untenable. That's no longer a valid study. It's guesswork.
But given the ferocity of the jihad against drunk driving, this is not entirely unexpected.
Alas. Bad things done in the name of noble purposes. Surely no harm is done, right?
Re:More like inflated data (Score:2, Informative)
I think they're talking about counterfeit items here. Substandard counterfeit parts in things like airplanes, or even chips with espionage functionality built in to them, or counterfeit medicines, can and do cost lives or compromise national security.
Re:Not reliable? (Score:3, Informative)