Ask HN: How do you detect liars?

Well there's liars on both ends. On the one end you've got the fancy vocab know-it-all smooth talker. On the other end we've got the I can learn quantum finance on-the-fly guy (perhaps quantum finance is not too hard, but you get the point).

Honestly though - I don't think anybody will give you a very useful tip as most honest people will probably admit that their "gut" is still a very important factor. But I do understand you're simply looking for inspiration so not that it's pointless to ask.

Personally I think that most poor and average-level liars tend to play it safe. They talk a lot, but most of what they say is stuff that's very agreeable, they don't have the balls to make up their own complex-sounding insight. Of course a top class liar will pass that. And sometimes, you know, a liar is so top-class that he might even actually be able to do the job :-)

It's a game. And by setting it rules you're simply determining who's capable of beating it. Someone will always be able to. I'd rather focus on making it easy to fire them.

The problem is sometimes people believe lies they've been told -- they might have read an article somewhere that x can do y, when they really misinterpreted the entire thing, and hold that (mis)belief in their head. They might even sound confident when they talk as well, because to them it is true.

That being said, if someone has knowledge in the expertise you are looking for, they should be able to break things down. So if you ask them a question about Z, and they say "well of course I can do Z", you ask them to break down what the steps they would go through in order to do Z. If they know their stuff, it should be trivial.

On a side note, the people who know their limitations are, like you said, the ones you want to talk to. Socrates was considered the wisest man in Greece, it was said, because "[he] knew that [he] knew nothing"(paraphrased)

Not everybody who makes incorrect statements is a liar. People may not know better or may skip details ("all birds can fly", "three times three is ten"), either because they have learnt to be not too nerdy to people they do not know, or because they know leaving out details sells better. I meet many more of either of those than outright liars. So, the primary rule could be: in general, people do not lie.

Also, you should never rule out that your 'pretty good in depth knowledge' is not broad enough and/or not pretty good enough, relative to some people you meet.

There are three possible reasons for incorrect statements:

1) Stupid but honest - doesn't know the truth and doesn't realize he is wrong.

2) Ignorant and dishonest - realizes he doesn't know and tries to cover it up.

3) Liar - knows the truth but lies anyway.

4) simplification - if a weatherman says "it will rain tomorrow" (s)he will leave out lots of 'given', 'statistically', etc. Similarly, one can find fault in any categorical statement ("1+1 equals 2" -> not in Z/2Z). If you meet a person for the first time, establishing conventions for the level of simplification can be difficult. That may lead one to think somebody does not know what he is talking about.

There also are variations on #3 that I would not really call lying:

5) "Santa flies from the North Pole". Technically #3 above, but feels different to me.

6) Politeness - as when answering a "do I look fat…" question.

Actually, 1 + 1 is 2 in Z/2Z because 0 = 2, no?

This is incredibly tough. I've found that if they can't go at least a bit more in depth about whatever it is, or if they can't point to relevant examples of whatever they're talking about, that can be an indicator of them being full of it -- however, this isn't enough to make me suspicious unless it happens fairly often, and if it was from an area I had absolutely zero domain knowledge of it would probably be fairly easy to fake as well, if a bit more rare for people to do so.

Other than having people that you know are competent in X, or studying up on X to the point where you can start to smell bullshit, I'm not aware of any sort of really reliable techniques.

People who lie a lot tend to suss themselves out over time, but that doesn't mean they haven't cost everyone a lot of time and money by the time they do.

Maybe some of the body language indicators could serve as an additional "hmm seems like this person might be full of it" -- blinking more, looking up and away, etc. These are also not reliable as a steadfast rule though, especially in fields that are filled with introverted left-brainers, ha.

It'll be interesting to see if anyone has any decent techniques for this.

Haha have to cite a study here [Kruger and Dunning 1999, Unskilled and unaware of it ...]

(recollection from CFA Level 3 material)

So the authors of the study asked people to rate how they have performed on a logical reasoning test. What they found is that people in the lower half of the test results scored less than 40% correct, yet they thought that their scored in the 60% range. This was particularly bad for the lowest 25% who got 10% correct.

The irony is that people who did very well (top quartile) actually thought they performed worse (55% vs 85% performance)

tl;dr: the skills needed to correctly answer questions are exactly the skills needed to evaluate performance. The best bet is to have some sort of objective measure. Learn a little bit about the field you are interested in evaluating others in. Think about fizz-buzz: a non-programmer can ask the question and get people to sweat

Saying "I don't know" seems like a display of honesty to me... I'd rather have someone reply with "I don't know" than give me a load of bull...

So I went to a course from http://www.humintell.com/ in San Diego a few weeks ago. Now I realize that this is not specifically what you were asking, it is a way to work on your people reading skills. You said you were going to be hiring people, and if you interview face to face - this type of expertise could be helpful.

I learned to "trust my eyes, not my ears" so to speak -- ie what people actually do in preference to what they claim about themselves. If there is a big disparity between the two, they are untrustworthy, even if they aren't intentionally deceiving you. I find it useful to remember that people who are lying to themselves about something cannot tell me the actual truth, not because they are intentionally deceiving me but because stating the truth to me would make it impossible to keep lying to themselves. I also look for patterns rather than specific instances. I mean, if someone makes a miss-statement, he may not be lying. He may be tired, distracted, mis-informed, etc. So I don't read too much into a single incident. I am big on giving people the benefit of the doubt. I used to think I am "forgiving", but I have concluded that is not entirely accurate. It is more accurate to say I am slow to judge, but once I have arrived at a definitely negative view of someone, I am a very hard sell.

I concluded years ago that "Forgiveness is a gift. Trust is earned." Abusive, untrustworthy people are very, very big on claiming you should "forgive and forget" and not hold some transgression against them. Such people should not be trusted any farther than you can pick them up and throw them. Decent human beings who made an honest mistake typically are fine with making it right and don't expect you to simply give them a pass. Some folks are simply young and have not yet made a definitive choice as to which camp they belong to. I generally give such individuals a chance to prove themselves if I can, even sometimes in the face of behavior I wouldn't likely tolerate from someone older.

What exactly is the technological domain? I'm sure someone on HN knows enough about it to give you some good starting points for asking questions

I play L.A. Noire to hone my lie detecting skills.