We flip a coin. If you lose, you pay $100; if you win, you receive $X. How large does X have to be before you would voluntarily take the bet?

Edit: Just one flip.

Please go read my hypothesis after voting: viewtopic.php?p=89772#p89772

I wrote that and then I thought, why not just test it?

_________________
That which can be destroyed by the truth should be.
--
My (sadly neglected, but not forgotten) project: http://dailyjoseki.com

As long as I am convinced that I will receive my win and that the coin is fair I accept any bid above €100.

_________________
Occupy Babel!

How many times are we flipping the coin? Once? An infinite amount of times? Some finite amount of times?

_________________
My plan to become an SDK is here.

Just one flip.

_________________
That which can be destroyed by the truth should be.
--
My (sadly neglected, but not forgotten) project: http://dailyjoseki.com

_________________
be immersed

See, Kirby, being in slightly less favorable circumstances than yours, I would happily take the million-dollar bet... they'd have no way to collect if I lost!

I don't have a million yet, either. I hadn't considered that. I wonder if filing for bankruptcy is allowed in this game...

_________________
be immersed

I voted for less than $110 but I'm heavily indoctrinated by poker literature and a math degree.

I voted for 125-149$.

Anything less than 12.5$ is not worth my time

Seems like there's an issue with finding samples from the below 10k category.

I think a better way to phrase this bet is that I must bet $100 in order to win $100+x. The mechanics would be that I put $100 into the kitty, and you put $100+x into the kitty, then flip the coin, then the winner takes all.

Statistically, anytime x>0 it's a good bet, but emotionally, its pretty unattractive, particularly coming out of the cold. Perhaps circumstances matter - if I was off to a casino with the aim of doing some gambling, then i would be emotionally ready to lose some money, and the terms of this bet would be fantastic if x>0.

But if you came to my house and offered me this bet, I would need it be big (I chose 180).

I would consider taking off your math hats for a moment and rejecting the premise entirely. Why gamble at all? The risk isn't 100$.

Reminds me of a joke:

A repairman comes to a farm to fix a broken power line, and after he's done, he asks the farmer if he can use the toilet. The farmer points him to the outhouse, but warns him that it's a two-seater and says that he has to go too. So the two men go to the outhouse together. While pulling down his pants, a quarter falls out of the farmer's pocket and lands in the latrine. "Darn it!" says the farmer, and pulls out his wallet, takes out a ten dollar bill and tosses it into the latrine. "What the heck did you do that for?" asks the repairman. The farmer looks the repairman in the eye and says. "I darn well ain't going down there for a quarter."

_________________
Patience, grasshopper.

Mathematics misses the point. "We" never flips coins. Usually the person who offers the bet will. When he is willing to offer an obviously bad bet from his perspective, it likely is a fraud. That is what I would expect and I would act on this premise, although it could be a psychologist (studying reactions to surprising offers) or a go player (trying to convince me that I have loss aversion) instead. In fact the more obviously the bet is bad for the person offering it the more convinced I would be of that.

I chose 200. My math degree won't help increase my income... So let's try to at least get more from 100$

_________________
Geek of all trades, master of none: the motto for my blog mostlymaths.net

i voted for $125 - $149. if it was lower bet or if we would play longer at the same stakes, i would go for anything above $100 as says math and my gamesmanship, but i am a poor student and if i am to risk $100, i want it to be for something more profitable

_________________
Spilling gasoline feels good.

I might be wrong, but probably not.

Alright, so 26 players above 10k responded so far, and by my count 18 are either really bad at visualizing their reactions or have way less than normal loss aversion. The other 8 are much closer to the number which is supposed to be average. It's not a big sample size but the effect size seems enormous, much larger than I expected. Unfortunately I think I picked too low of a strength cut-off, so it will remain untested as to whether loss aversion correlates with strength. But this does seem like evidence in favor of my hypothesis that go players have reduced loss aversion. But are you guys a representative sample of go players at large? Who knows, probably not! So it's not great evidence.

tapir: That is a totally valid point, and certainly one that I would go by in real life. For this exercise, though, assume that no fraud will be involved. Like you're playing the stock market or something. There's no fraud there ever.

quantumf: I used (my memory of) the wording in the book, even though I personally prefer yours. Your points, tapir's, and more are mentioned in the book.

Anyway, thanks everyone for participating. I do highly recommend the book ("Thinking, Fast and Slow" by Daniel Kahneman).

_________________
That which can be destroyed by the truth should be.
--
My (sadly neglected, but not forgotten) project: http://dailyjoseki.com

And there are other considerations, too. What will your spouse say to you when you get home if you lose (will you tell her)? Will you have to walk to an ATM to get the cash if you lose (how far is it, what's the fee)? Can you really expect to collect from the person who offered the bet (how much time, effort, and burnt social capital is involved)? And on and on.

This makes it quite difficult to interpret the results of survey-based research into psychological questions like this. You can tell people to abstract from those concerns before they answer, but they rarely do, and if you follow up and ask them to explain their choice they will often give answers that they were specifically asked not to consider by the researcher. You can do experiments with real rewards, but these usually involve trifling sums of money or irrelevant tchotchkes (like mugs and pencils). There have been experimenters who took their research budgets to the third world, where they could offer people bets that were a substantial part of their annual income, but for ethical reasons they were never able to do loss aversion experiments like this...

I typed this yesterday, but it didn't submit : I would take a poll and look for arbitrage opportunities .

But assuming that losing wouldn't cause me to starve, and that transaction costs are negligible, I take the flip starting a few cents above $100, because there is a positive expected value. If it was my last $100, and I would starve without it, I wouldn't take the flip.

On the other hand, if I owed someone $200, and anything less would get me killed, AND this was my last opportunity to make the $200, I would take the flip for a penny.

People always fight these hypotheticals.

One problem with these things is that psychologists (apparently) are not professional gamblers. If they were, they would know that the size of your bankroll matters. Moi, I would not make $100 bets without a bankroll of at least $10,000. I do not not have such a bankroll nor would I, for the sake of argument, contemplate shifting that much of my net worth to a dedicated bank account to serve as a bankroll. I think that a lot of people are in the same boat. Why bet at all?

OTOH, if I enjoy gambling, and set aside a certain amount of money for that hobby, an entirely different set of considerations apply.

_________________
The Adkins Principle:
At some point, doesn't thinking have to go on?
— Winona Adkins

Visualize whirled peas.

Everything with love. Stay safe.