Tuesday, February 28, 2017

“Lies, Damned Lies, and Statistics”

"Lies, Damned Lies and Statistics" was popularized by Mark Twain.

I took a statistics class in college (actually it’s statistics and probability). I realized all of it was just opinion, not fact, not science.

You take a small sample and try to extrapolate it to tens of millions of people. Example: I have no idea who was polled, but it was not the Heartland, when it came to Trump. He was given a 2% to 3% chance of winning.

I laughed.

I had done my own polling in my hometown, which had been Democrat since forever. All I heard was “Trump, Trump, Trump.”

My hometown was a middle-class steel mill town with the steel mill now out of business. I extrapolated that to rest of the U.S. and realized Trump was going to destroy Hillary Clinton.

I was right; that is exactly what he did. I knew it was going to happen, beyond all doubt.

The only part of the country that matters politically is the Heartland. Flyover Land. Los Angeles and New York City don’t count at all.

Perhaps the next Presidential election the pollsters will grow some brains and ask questions in Missouri or Tennessee. Because those states matter a lot more than New York or California.


Glen Filthie said...

Stats and science are so useful tools Bob - but ya need half a brain to interpret them. I gotta admit that when you started calling it for Trump way back - I thought you might be full of beans. Just as the media got stuck in their own echo chamber, I've seen us more conservative guys do it too. I would never've thunk Obutthole would pick up a second term - but he did! (Did you call that one correctly too, Bob?)

Stats are like anything else: GARBAGE IN=GARBAGE OUT.

One Fat Oz Guy said...

Where do you stand on California leaving the USA?
Personally I'd say let them go and give them their share of the national debt as a percentage of total actual population (include the undocumented immigrants) but I'm not an American citizen, so I don't really have a say.

Bob Wallace said...

"Where do you stand on California leaving the USA?"

Break it up into three states.

Carnivore said...

Last October, I drove up from Chicago to the UP. After Milwaukee, I got off the interstate and took back roads, which pass through small towns. The entire way, I saw 1 or 2 Hillary signs; the rest were all Trump signs. Realized the 'Hillary will win by a landslide" the MSM was spouting was BS.

Bob Wallace said...

My hometown is as Democrat as can but I was seeing Trump signs everywhere. I knew it was over for the Hildebeast.

The Damage Dealing Meat Shield said...


My hometown experienced much what you witnessed with your hometown. In my hometown, it too was for the longest time Blue/Democrat. But once Trump took center stage, businesses were renting $10,000/month billboards showing their business support of Trump. Moreover, other businesses were posting MAGA in their store windows. And anyone in business knows that one shouldn't express their political support for it can be detrimental for their livelihoods, but many threw caution to the wind. Which luckily for them it didn't backfire. Additionally, people who I've known all my life were died in the wool Democrats, quickly switched parties and voted Trump. And what I described, is what happened all over Pennsylvania. Counties that like my home town were blue every election year went red.

James said...

The mathematics behind Statistics and Probability is highly accurate. However, it, like climate change, can be fudged to achieve whatever the goal of whoever uses them. And since the media only reports the percentages and not the test used, sample size, variables, probability levels set at the onset of the study, etc, all the herd sees is what their masters want them to see.

Bob Wallace said...

"The mathematics behind Statistics and Probability is highly accurate."

They were completely wrong about Trump/Clinton.

James said...

No, the "numbers" weren't wrong. What was sampled was wrong. Deliberately. Statistics doesn't give you accurate results IF you do not do an appropriate sampling. If, for example, you only sampled democrats in Chicago, you would have gotten an invalid result. If you took a truly RANDOM sample from all 50 States, the results would have reflected the eventual results. Face it, the media is not interested in reporting the truth. They are interested in influencing outcomes.

Bob Wallace said...

I called the election correctly by sampling six people and looking at yard signs.

Statistics is horseshit. I know how it works. I had a class and that's why I know it's Statistics and PROBABILITY. Probability is just a guess, nothing else.

James said...

"I called the election correctly by sampling six people and looking at yard signs."

So, you do use Statistics. No, probability is a percentage. And Statistics is not horseshit. How it is used is what determines whether it is horseshit or not.

James said...

You do realize that successful poker or blackjack players use Probability, don't you? They can never say what a result will be before the final outcome. But, what makes them successful is that they compute the probability of a certain outcome. They will not be right all the time. They will just make a living playing a card game.