Why It’s So Freaking Hard To Make A Good COVID-19 Model

This was written by a guy who predicts the election results, sometimes well ... other times not.  He predicted Hillary would be President.  She's not.

Glenn Reynolds:

TO MAKE A GOOD MODEL, YOU NEED RELIABLE DATA AND A SOLID UNDERSTANDING OF THE UNDERLYING PROCESSES: Why It’s So Freaking Hard To Make A Good COVID-19 Model. Otherwise you’re just playing with numbers and fancy graphics.

It doesn't help when supposed experts, like Drs. Fauci and Birx refer to models to predict how many people will be infected and die. Because when the are forced to explain why their projections are wrong they are also forced to admit that even statistics like the number of deaths attributed to the Wuhan virus are not accurate because everyone who dies with the virus is assumed to have died because of the virus..

Scare-tistics

 It's clickbait.

“Just one sunburn increases your chances of getting cancer by 100%!”

“People who don’t eat [food] have a 20% greater risk of colon cancer!”

“You have all the risk factors for skin cancer.”

“People with a family history of [malady] have a 40% greater chance of getting [malady] than the rest of the population does.”

Sigh. Just what my nerves need to hear. But what is my risk? What are the odds of my having [malady]? You notice that the news articles never really give you that information?

Scare-tistics. That’s what I’m starting to call the numbers tossed around by media outlets, often concerning health topics but also environment, health insurance, and the like. The goal is to get you to stop and watch/listen/read. Scare-tistics are not really helpful.

For example, and I’ll pull numbers out of the air, because the news piece came out a few years ago, but there was breathless reportage that “breast cancer diagnoses are up over thirty percent from forty years ago!” and implied that some dreadful thing in the environment was causing breast cancer. To which the person in my family who works in a related medical field snorted and said something that rhymes with “pulpit.” What has happened is that 1) mammography has become cheap and common, and 2) because of that and better technology and imaging, more and more tumors are found that would not have been caught before because of their tiny size. Many of those are very slow-growing, most are not malignant at the time of discovery and are tracked but not removed.

Of those diagnosed with cancer, back before mammograms, sometimes the woman would have died of something else first without anyone guessing that they had breast cancer, like my great-grandparent. She/He died of heart failure and his/her autopsy revealed un-diagnosed [organ]* cancer. Her/His heart got him/her, not the cancer. She/He was in his/her 80s. Today, there are women in their 90s who are diagnosed with breast cancer, but who are suffering from other maladies that will do them in long before the cancer would.

Right Wing Extremism vs. Islamic Extremism in the United States: A Look at the Numbers

There's a lie being perpetrated that Right Wing Extremism is more deadly than Islamic exremism.  Here's the reason it's a statistical lie.

Again, as I noted in the introduction, this is not to dismiss the threat of right wing terrorism and its very deadly consequences for some, but only to add perspective to the claims being made. Right wing terrorism is more deadly for Americans only if you add a number of very limiting parameters (e.g. excluding the victims of 9/11, ignoring “lone wolf” attacks without solid connections to groups like al-Qaeda and their affiliates, etc…). But if you lift those limitations, and apply equal standards, then the raw and unfiltered numbers of deaths of Americans due to Islamic extremism in the United States over the last fifteen years dwarf the numbers attributable to right wing extremism by a ratio of over 62 to 1. Even if you leave out 9/11 victims and just focus on the ideological statements and goals of the attackers, then the deaths of Americans due to Islamic extremism still outnumber the deaths attributable to right wingers (which reveals an even greater disparity when compared with population groups). If we move beyond America’s borders, then the disparity becomes far greater, with somewhere around 90% of the world’s terrorism related deaths attributable to Islamic extremism, and only a fraction of 1% attributable to right wing extremism.

Read the whole thing.

More Guns, Less Murder

Ann Coulter on those Immigrant crime studies

Ann Coulter: Every Pro-Immigration Claim is a Lie

In the wake of Kate Steinle's murder at the hands of an illegal immigrant, mass immigration advocates have begun a campaign of lies in defense of their cheap labor. "Studies show," they say, that immigrants commit LESS crime than the native population.

Inasmuch as the vast majority of post-1970 immigrants -- legal immigrants -- are poor, non-white and come from countries with far worse crime rates than our own, that's at least counterintuitive.

The main evidence cited in support of the claim that immigrants commit less crime than Americans is a moronic point about cities with a lot of immigrants seeming to have low crime rates. Check and mate, Mr. Trump! ...

Reason magazine boasts, for example, that El Paso, Texas, has a large Hispanic population and yet El Paso "is among the safest big cities in America."

In fact, however, El Paso's "safe city" ranking is based on an outdated FBI crime index that includes only eight crime categories, excluding such crimes as drunk driving, narcotics offenses and weapons violations. When the FBI's more complete crime index is used, El Paso has a higher crime rate than the national average.

The reason crime has plummeted around the nation in the last few decades is aggressive policing, increased prison sentences and the expansion of concealed carry permits. (All policies currently being jettisoned by liberals.)

According to The New York Times, the drop in crime in New York City during Giuliani's first two years as mayor aloneaccounted for 35 percent of the reduction in the national crime rate.

The second main line of attack on the idea that immigrants are committing prodigious amounts of crime are the apocryphal "studies."

The two researchers whose work is cited over and over again for the proposition that immigrants are less criminal than Americans are Alex Piquero, criminology professor at the University of Texas at Dallas, and Bianca Bersani, sociology professor at the University of Massachusetts, Boston.

Pew cites their studies -- and everyone in the media cites Pew, leading to headlines like these:

"UT Dallas prof finds immigrant kids less likely to commit serious crimes, re-offend" -- The Dallas Morning News

"UMass Boston Prof: Stereotype of 'Criminal Immigrant' Doesn't Hold Up" -- Targeted News Service

"Surprise! Donald Trump is wrong about immigrants and crime" -- The Washington Post

Curiously, we are never shown the actual studies, but simply told -- with some heat -- "studies show!"

I looked up some of these alleged studies this weekend. They're all hidden behind ridiculous Internet paywalls. I was often only the sixth person to read them.

It turns out that neither Piquero nor Bersani compared immigrant crime to "the overall population" -- as the British Guardian recently claimed in an article purporting to prove Donald Trump wrong. Rather, they compare immigrants' crime rate to the crime rate of America's most criminally inclined subgroups.

Thus, for example, once you get past the paywall, you will find that Piquero and Bersani's joint study, "Comparing Patterns and Predictors of Immigrant Offending Among a Sample of Adjudicated Youth," used as its base group "adolescents who were found guilty of a serious offense."

THAT'S NOT A REPRESENTATIVE SAMPLE OF AMERICANS! It's a representative sample of teenagers who are convicted criminals.

Read the whole thing.

That "5% GDP Growth" number is misleading.

The MSM are gleefully reporting that the economy grew by 5% in the third quarter.  This must be the economic boom that Obama has been promising since 2008?

Well, no.

It appears that most of the 5% growth last quarter was achieved by an increase in consumer spending on higher health insurance premiums. In other words, spending more on insurance premiums for ObamaCare is supposed to be good for the economy.

Does that make you feel richer already? No?

Well take a look at how Team Obama managed to create that 5% figure.

It seems that they "revised" (i.e.cooked the books) spending on health care services by taking it out of the 1st quarter and putting it back in later, in this case, the 3rd quarter.

Here's how they took it out:

Here's where they put it back in:




Here's a little secret: the government gets to revise the numbers for months and quarters in the past under the assumption that given more time they get more and better data. I found this out years ago when I hears a talk by a very smart - and witty - Wall Street analyst who made great fun of government economic statistics by quoting numerous revisions for GDP numbers for the same period.  It was eye-opening, and I have never viewed government statistics in quite the same way.

Revising economic data as better information is gathered sounds plausible, but it also gives the government the opportunity for political mischief.   Thanks to the "polar vortex" of early 2014 that slowed the economy, the government statisticians figured that the 1st quarter was so bad that they may as well use it as a "kitchen sink" quarter. Via Zero Hedge:

Back in June, when we were looking at the final Q1 GDP print, we discovered something very surprising: after the BEA had first reported that absent for Obamacare, Q1 GDP would have been negative in its first Q1 GDP report, subsequent GDP prints imploded as a result of what is now believed to be the polar vortex. But the real surprise was that the Obamacare boost was, in the final print, revised massively lower to actually reduce GDP! ... Because the brilliant propaganda minds at the Dept of Commerce figured out something banks also realized with the stub "kitchen sink" quarter in November 2008. Namely, since Q1 is a total loss in GDP terms, let's just remove Obamacare spending as a contributor to Q1 GDP and just shove it in [a later quarter].

.....

In short, two-thirds of the "boost" to final Q3 personal consumption came from, ... the same Obamacare which initially was supposed to boost Q1 GDP until the "polar vortex" crashed the number so badly, the BEA decided to pull it completely and leave this "growth dry powder" for another quarter. That quarter was Q3.

And that, my friends is how the MSM is now able to giddily report that the Obama economy is rolling in high gear.  

H/T Bud Norman at Central Standard Time 

Don't let your daughter be raped in college!

What is the relationship of cheese consumption and death by bedsheet?

There is a nearly 1 to 1 correlation between eating cheese and dying by becoming entangled in your bedsheets.

	2000	2001	2002	2003	2004	2005	2006	2007	2008	2009
Per capita consumption of cheese (US) Pounds (USDA)	29.8	30.1	30.5	30.6	31.3	31.7	32.6	33.1	32.7	32.8
Number of people who died by becoming tangled in their bedsheets Deaths (US) (CDC)	327	456	509	497	596	573	661	741	809	717
Correlation: 0.947091 Which only goes to prove that if you torture the numbers long enough they will tell you anything you want them to say.