Tuesday, December 17, 2019

Polygraph test is BULL SHIT

Polygraph tests do not measure deception or lying directly, but rather possible signs that a person could be deceiving the interviewer.

It is also a well known open secret than a person can be taught to lie on a polygraph test.

Accuracy for Polygraph is only 70% while facial recognition is 80% to 90%.




Most people believe that they can tell whether a person is lying by looking them in the eye. In fact, people who attempt that method are less accurate than a coin flip at discovering whether somebody is lying.
Most people also believe that "lie detector" tests can accurately predict whether somebody is lying. In fact, lie detectors are BS, which is why they're no longer admissible as evidence in courts.
Since it's hard to be successful when you're being subjected to bullsh*t, here's an easy way to estimate the probability that a statement is lie, based on a (well-sourced) article in The Cut.
Simply answer the following YES/NO questions to the best of your ability:
1. Does the source stand to gain if I believe the statement?
Explanation: People are more likely to lie if they're getting something in return for it being believed. Obviously.
Example: a salesperson who will say anything to make a sale or a politician who will promise anything to get elected.

2. Does the source stand to lose if I discover I've been lied to?

Explanation: People are less likely to lie if the revelation that they told a lie would damage an ongoing relationship.
Example: A stranger in a bar has no stakes in being truthful because you'll likely never see that person again. By contrast, a salesperson who wants ongoing business from you is far less likely to bullsh*t while selling.

3. Does the source stand to gain if I don't believe it?

Explanation: If the source communicates something that is an unpleasant truth (to the source) and which would be less unpleasant if you didn't believe it, the statement is less likely to be a lie.
Example: a colleague who admits doing something stupid is highly unlikely to be lying about it.

4. Is the source wealthy?

Explanation: Contrary to what you might think, wealthy people lie, cheat, and steal more than people who lack wealth.
Example: Pick any random event in the past two centuries that involves Wall Street bankers. Any statement made by any banking executive is guaranteed to be a lie.

5. Can the source answer "drill down" questions?

Explanation: With lies as which much in life, the proverbial devil is in the details. Most liars only create top-level bullsh*t. When you probe, you find inconsistencies and blank spots.
Example: A company claims that their system uses "Artificial Intelligence." Unless they can answer with precision (and not techno-blab) what they mean by that, the claim is almost undoubtedly just some marketing bullsh*t.

6. Are there independent, credible, collaborating sources?

Explanation: Any piece of information that comes from a single source is more likely to be bullsh*t than one that's widely attested in a variety of usually credible sources. One proviso, though: a source that's simply echoing another source is not independent.
Example: Most conspiracy theories fall into this category.

7. Is the statement too good to be true?

Explanation: any statement that perfectly states what you wish to be true is probably a lie.
Example: "The Fast Food Quick Weight Loss Diet"

8. Is the statement couched in weasel words?

Explanation: Clever liars create plausible deniability by making statements that are technically true but which most hearers will misinterpret.
Example: "4 out of 5 Dentists Recommend Colgate." They were actually recommending that people use toothpaste; they weren't singling out Colgate as being preferable to other brands.

9. Does the statement contain averages?

Explanation: The concept of "average" is only meaningful when comparing things that are similar, like grade point averages. The concept is misleading when comparing things that are widely different, like the net wealth of individuals.
Example: The average net worth in a room containing one billionaire and 999 homeless people is $1 million per person.  

10. Does the statement assume causality?

Explanation: Just because A happens after B doesn't mean that A caused B. Liars often restates coincidences as being causally connected. Note: a person who does this might just be confused rather than lying.
Example: Tim Cook, who rises at 3:45 am, is successful. Therefore, if you rise at 3:45 am, you'll be successful.

SCORING:

Score 1 for each "YES" answer.
  • 0 to 2: Probably not a lie.
  • 3 to 4: Possibly a lie.
  • 5 to 6: Probably a lie.
  • 7 to 10: Almost definitely a lie.
By Geoffrey James

No comments:

RM1 billion takeover of KKMart coming soon.

  Conspiracy, conspiracy. That is the hot issue today. Before I start, can you guess WHO STAND TO GAIN OVER THE SOCK ISSUE? EBIT LIEW LOST H...