Are Entrepreneurs Pirates? 4min Test for HN by Behavioral Economist Dan Ariely | Hacker News Reader

Are Entrepreneurs Pirates? 4min Test for HN by Behavioral Economist Dan Ariely

danariely.qualtrics.com

55 pointssiglesias69 comments

I'm sure that my behavior does not reflect my answers on this test. Given the numbers, I calculate expected value and answer rationally. In real life, a) the values and percentages are much less clear and b) even if they weren't I know I don't follow them as rationally.

This is the point that I'm trying to get at. For Arrington to assert that entrepreneurs are "different" because they like risk, adventure, he is claiming that they knowingly choose bad deals. He is saying that these bad deals are compensated for not by money, reward, but the thrill of doing a startup. Hmm.

We get a thrill from gambling sometimes even though we know that the odds are stacked against us. Arrington is drawing the same analogy for entrepreneurs, and that's a bad analogy. It's a romantic notion, but I don't think the monetary valuation of an entrpreneurial venture is negative in the minds of entrepreneurs.

"Would you rather have a $100k steady job, or a 5% chance to make $1m in a year with a startup?" -> Definitely startup, even if the expected utility is half.

Would you pay 50k to join a startup and work for free as a consultant?

What is this, an anchoring exercise? :P Are you asking if I would pay to work in a startup? No, I can start my own for free.

And just so everyone knows what I mean by this, 50k is the opportunity cost of the other job that's being forfeited. This means that the "thrill" of working in a startup must be worth at least 50k.

But part of the thrill would be the possibility of the million.

IMO, you're right about the romanticism, but it'll be hard to tell anything from this survey because of the tendency for nerds to overthink!

Tell you what: I tried to go with my gut feeling on every question. I am very risk-averse, but the idea of gaining money doesn't mean that much to me. Winning $100 vs $1000 isn't that emotional for me. Oh, sure, I'd like $1000 but I don't even care about $100. I thought about scaling the numbers up til they mattered to me, and decided I'd almost always take the sure thing because I could turn that money into a higher amount thru investing it in expanding my product.

But I would gladly pay $250 instead of even a slight risk of losing $1000. (Or whatever the numbers were I've forgotten already!) IRL, I've bought a new car because I'd rather pay a couple grand more than have the uncertainty of potential repair bills, etc. Clearly I hate negative unknowns.

That risk aversion has clearly influenced the type of "startup life" I've chosen. My products are "boring," but quite profitable. No funding, no 80 hour weeks, no sleeping under the desk, and no potential to be gobbled up by Google, either.

Nevertheless, I find this far less risky than having a job, because nobody can fire me from my own company! Plus, I'm not miserable because people are always screwing with my work.

That said, I think part of the special startup mindset is similar to the people who will bet more on dice they've rolled than dice rolled by someone else... whether a startup is acquired feels like something they can control, however irrationally.

Maybe you could ask what people believe is the average chance of acquisition for a startup in general, and what they believe their chance is?

Same here. The only "interesting" choice was when the safe/risky options had the same expected value.

The percentages are much less clear, as are the payoffs. I was attracted less by the monetary payoff of the entrepreneurial life than the other, less tangible benefits: responsibility, control, independence, learning opportunities, challenge, etc. I'm not attracted to safety, I'm afraid of boredom.

Well, I think that's true for pretty much all surveys. If you are reasonably intelligent, it is actually hard to not figure out what is really asked and give answers based on how you would react instead of how you should react.

I've never seen a survey that wasn't easily manipulatable. So I guess we have to rely on people to try to be honest...

Completely agree, this felt more like a test of basic math skills than a test of behaviour. In real life it tends to be a lot harder to be able to calculate expected value, with some exceptions especially with card games.

Behaving on expected value is not rational.

Exactly.

Choosing between

a) 100% chance of $10 vs 10% chance of $100

is very different question than choosing between

b) 100% chance of $100K vs 10% chance of $1M.

c) 100% chance of $1M vs 10% chance of $10M.

I think many people will choose 10%/$100 combo of a) just for the thrill.

The latter two are of course dependent your financial situation, but hit an interesting sweet spot for YC crowd.

What would be your choices?

As I said in another comment, i tested this exact question out on mechanical turk: http://joshua.schachter.org/2008/09/amateur-economist.html

Behaving based on expected value is the only rational choice.

There was no beta attached to the time-value of money - so it was no way to measure the risk.

I'm glad to see that I'm not the only one who calculated EVs.

Based on all these "rational actor" responses, I take it none of you purchase insurance?

I certainly don't on things I can easily afford to replace.

I have car insurance because it is legal mandated. I don't have health or life insurance.

Only when it's legally/contractually required (car/financed-house), or comes with negotiated discounts (health), or handles a catastrophic/risk-of-ruin situation (chronic disease/business liability).

Insuring against small losses is a net financial loser, or else the insurance company wouldn't offer it.

Actual rational actors understand that extreme loss has a disproportionate effect. (In other words, losing $100k is more than 1k as expensive as losing $100 for most of us.)

Naive rationality may be easy to analyse but that just means that the GIGO transfer doesn't cost much.

Then why didn't you answer according to what you think you would actually do? The idea of the test was to gain insight into your personality with regard to risk and reward. If you can't honestly see yourself accepting a (.4 * $1000) loss over a $50 loss, then you answered incorrectly according to what the question was actually asking.

Personally I had the same dilemma, but I found it revelatory to really look inside myself and see if I could do what expected utilities would dictate. It was especially difficult on the loss questions, as I have very little capital to my name.

I think the problem is that risk always has to be calculated against reward, and in this test, it was always reward-against-reward or risk-against-risk.

For example, let's take one of the loss scenarios and pretend that it's some wacky hosting cost.

* $50 guaranteed loss - paying $50 for 100GB of bandwidth

* 0.04% of chance of $1,000 loss - paying $2.50 for 100GB of bandwidth, but an extra $997.50 for unlimited bandwidth if you go over 100GB.

If I don't expect to do much traffic, I'll take the 4% chance; I project I might do 20GB of traffic during the term of the contract, so I will save 20x as much money. My product has a 4% chance of going viral. If I lose the contract lottery, it's because I've won the product lottery, and that 4% means I've got a rapidly-growing product that is going to make me a lot of money. The risk of that extra $1000 is well worth knowing that I'm not going to have to move my product between hosting providers when the blogosphere in in a mad frenzy about my internet tortilla delivery service.

That's completely different than a scenario wherein the hosting provider says "This is a variable contract; you owe us $2.50 for 50 GB of bandwidth, but each month we pick one of our 25 customers and they pay $1,000." There's no way in hell I'd sign on for that. There is no indicator of success correlated to the risk there.

Without context, I'm not sure that these questions are all that useful. You point this out yourself - the risk questions will be interpreted variably depending on the individual's resources and perspective. I can absorb a $50 loss trivially, so if the reward is equivalent, I'll choose that.

When comparing to the rational choice, I did what I do in real life: added a safety margin. If anybody tells me there is a 10% chance for a project to succeed, I will always subtract a couple of percentage points for unforeseen circumstances.

You treated it like a school test, looking for the "right" answer. To paraphrase Nicholas Nassim Taleb, if a coin falls head 100 times straight, the chances it will fall tails the 101 time are 50% only in academia - in the real life the coin is fixed.

Results!

http://news.ycombinator.com/item?id=1922412

I just went through it, choosing almost exclusively certainties. I'm definitely an entrepreneur, although currently a very struggling one.

I fear the survey is going to be incredibly influenced by income level and current financial security. If you had changed the magnitudes of the numbers, my answers would have been completely different. I can do the math, but I also know that my immediate financial needs are unmet and that my ability to sustain losses is very limited.

Yeah, my budget is pretty tight as well. I can handle losing $50, but if I lose $1500, I can't make my mortgage this month.

Without a doubt. For instance, I will always take a $50 loss over any chance of a $1000 loss, because $50 one way or the other barely has in impact on my day, let alone my life. $1000, on the other hand is significant.

If however, the questions had been about $500 vs. $10,000, I'd have to think a bit more carefully.

For instance, I will always take a $50 loss over any chance of a $1000 loss

That's almost certainly not true. 0.1%?

It seems to me that they should first ask how much of a fine would you be able to pay without having to check on your checking account (for instance). That way they could scale the questions a little.

There's obviously an emotional point at which the questions stop being a mathematical puzzle, and make one reflect on one's appetite for risk : that's really what they mean to be investigating.

That just means that your utility curve isn't flat. In other words, you value very large sums more than smaller ones.

And people who honestly can't afford the certain $50 penalty might be forced to take a chance of a $1000 loss even when it doesn't make sense (assuming a flat utility curve), etc.

You make a good point about the magnitude of numbers they chose to use. I'm not wealthy by any means but I can afford to lose $1500. I would definitely choose to put up this amount if the odds were really 13%. This is assuming that the 'game' (i.e. startup idea) is in my area of specialty, and I felt it was not just a dice throw.

I would never gamble this amount, and I would not put up $15,000 or $150,000 unless I were part of the team. I wonder how my own magnitude levels will change as my personal income increases. I remember as a student saving a few dollars for the weekend for a beer, and now I'll grab a taxi for those few dollars every day without even thinking about it.

I didn't like this test very much. The financial questions are highly dependent. I guess I was relating to a hypothetical situation rather than the one I'm in now, but I don't think the choices here have anything to do with entrepreneurship; they're all circumstantial.

As all entrepreneurs know, things can fluctuate quickly; there are times when $50 today, little as it is, is more important than the potential of $5000 in the distant future, there are times when you need all the fumes you can get to keep things running. Then there are times when there is a lot of cushion available and $50 today is almost so worthless it's not worth considering -- in that case, of course it's better to take the gamble.

This test has very little to do with entrepreneurship. I ruined it for everyone and just chose random things.

Just to ruin your day, choosing random answers doesn't do any damage :)

This is all about statistics, and also about comparing with a control group. I assume the authors have (or can get) baselines for all the questions here. They look like the kind of questions Dan Ariely would work a lot with.

So it doesn't matter that the questions are context dependent. If HNers massively answer one way while the college student population doesn't, you have a difference you can work with. Of course, it's not a definitive answer but it's still valuable information.

As for the results, I'd guess entrepreneurs are more risk averse then average, but also more inclined to chose rewards in the future.

I'm really not sure it means much to say that people are "risk-adverse" or not; humans (and monkeys! see http://www.ted.com/talks/laurie_santos.html) are BOTH: they prefer to play it safe when they expect a gain, and prefer risk when they expect a loss.

For example, people (in general) prefer $500 now than a 50% chance of $1000, but the same people prefer to take a 50% chance of losing $1000 than losing $500 now with certainty.

Behaviorists call this "irrational" because the two cases are the same: whatever you prefer, you should prefer the same, that is:

- if you're "risk adverse", you should prefer the certain outcome in both cases

- if you have a preference for risk, you should take the chance every time.

But it's not what happens; to repeat, the same individual usually prefers the outcome that is certain when the expected outcome is a gain, and usually prefers to take a chance when the expected outcome is a loss, even when those options are mathematically identical.

So, if one wants to show that entrepreneurs are different from ordinary people as regards to risk, they should study at least those two dimensions: gain/loss. And since most people prefer risk when confronted with a possible loss, then entrepreneurs would have to either

- prefer risk always when confronted with a possible loss

- prefer risk more often when confronted with a possible gain.

- - -

If there actually is a difference between "entrepreneurs" and other people, it may be that entrepreneurs have a higher confidence in their own capacities, so that they think that they can influence the future in a good way.

That's a rather important point. Loss and gain are not symmetrical in most cases. I actually wonder if there's a study that looks at that in relation to actual wealth - it's easier to risk a loss if it doesn't materially affect you.

HN, anybody know of something like that?

I'm sure my choices here have been completely spoiled by all the pop-science books I read, in that my first thought upon each type of question was "Ah, that's from this and this research, where they found out that..." They need an "are you familiar with these types of questions?" question, because both the statistically expected answer and the rational answer were already known to me. I assume this is the case for more entrepreneurs, so I doubt this questionnaire will display a large discrepancy between entrepreneurs and non-entrepreneurs.

All I can say is, Pop-science books are fun, everyone should read more of them.

Yeah, I'm afraid I also did the "I'm a rational economic actor" answers,

That's funny--I had no trouble figuring out what I'd actually prefer, and even though I'm familiar with the various stories about risk-aversion and such, I think my answers make sense in the context of my income and finances.

I just got a significant check from a client today -- I am sure that affects my money now vs. later answer.

The questions are flawed because many entrepreneurs have already used up all their risk seeking "slack" and might quite rationally prefer to lose $50 over a 1% chance of losing $1500, since a $50 loss could be dealt with more easily in the context of a ramen budget... while a $1500 loss might destroy the company entirely.

In other words, just because you are in a casino and put all your money on double zero doesn't mean that you should also play russian roulette with a pistol while the roulette wheel is spinning.

These tests are flawed because the absolute value of the money we're dealing with is too much of a variable between subjects. $1500 is not much for subject A to lose or gain but it is for subject B.

A test should account for that if we want results to be comparable across subjects.

Mr. Ariely's book, Predictably Irrational, is a great read for those who aren't familiar with Behavioral Economics. It contains a lot of solid insights that entrepreneurs should understand, particularly relating to how consumers think.

A lot more people will complete it if you give them some idea how long it is. You should use Wufoo.

Thanks for the advice, Paul. I've updated the title and first page with more information about length and structure.

I also cannot show the entire survey on one page out of concern that its structure might influence the earlier responses.

Lawrence et al (2008), Nature 456, p168-169

http://dx.doi.org/10.1038/456168a

Some snippets from the above:

"...when subjects were introduced to the hot components of the task, differences were observed. We found that entrepreneurs behaved in a significantly riskier way, betting a greater percentage of their accrued points (63%) than their managerial counterparts (51%)."

"… risk-taking performance in the entrepreneurs was accompanied by elevated scores on personality impulsiveness measures and superior cognitive-flexibility performance. We conclude that entrepreneurs and managers do equally well when asked to perform cold decision-making tasks, but differences emerge in the context of risky or emotional decisions"

"... the entrepreneurs (mean age 51) are comparable to the young adults aged 17–27, whereas the managers (mean age 50.5) resemble their age-appropriate group" [with respect to risk tolerance]

"There are courses that teach the 'know-how' of entrepreneurship and, within this, teach how to mitigate risk through extensive business planning and market research. Yet given this new evidence, courses teaching the opposite — risk-tolerance in both behaviour and personality — may also be desirable"

Very interesting. I think loss aversion will turn out to be just as prevalent for self-professed entrepreneurs as for the general population, if not more so.

I'm working on a paper about entrepreneurial traits and such mdoells. I'm still not sure if an entrepreneur si risk avers. Some scholars argue, he likes risks, because he is quiting his job to get an uncertain "payment". Others argue, entrepreneurs don'tlike risks. If they go for an opportunity, they know the risk quite well and soforth have already evaluated the risk and decided, it is doable. Is their any new thought on this?

i have to agree with the others, I just calculated the value and chose the higher one. Interesting could be the outcome of the last questions where no probability, or intrest rate was given.

Additionally, I think, even if all of us are entrepreneurs (there's still now agreed definition of the term see http://bit.ly/amULry ) we decide different. We decide different, depending on our situation or the environment (see Gartner 1989ab 1989b, 1990) we are in. I'm a student my self, perhaps getting $150 now could get me through the month, but one million in two month isn't. So I would take the 150 bucks.

The experimental setup isn't great: the HN population almost certainly has an elevated level of education in areas like math and statistics. I don't think you should ever compare subjects who can compute "expected value" with ones who can't. Personally, I don't believe I can give great answers for this, because I'm conscious there's an experiment going on.

Here's a question: how does this assertion sync with Arrington's claim that entrepreneurs would choose low-to-negative expected value projects simply for the thrill of gambling, as it were? What's going on here?

Hi Samuel,

I think entrepreneurs are not so enthused over gambling as you would believe (gambling as in casinos & slots, or an online questionnaire).

I think that what matters more is the intrinsic value of building something, but I'm not sure about how to measure it..

I abhor gambling - risk assumed for Nothing - except entertainment. Shudder.

I'm a serial lifetime Entrepreneur.

The magnitude of the numbers matter -- significantly!

I actually tested this out, too: http://joshua.schachter.org/2008/09/amateur-economist.html