The hazards of Gambling

[Stephen Unger]

At a time when we are still in deep economic trouble, with huge
numbers of people unemployed, or under-employed, and with no serious
effort to get corporations and the super-rich to pay more taxes, many
states and municipalities are turning to gambling as a "painless"
revenue source. We are seeing more state lotteries, licensing of
casinos, etc., and the use of the internet to make it easier for
people to gamble. Is this a good thing, or is there a serious down
side? Place your bets as to what my position is on this. My analysis
is accessible at:
http://www1.cs.columbia.edu/~unger/articles/gambling.html

Steve
............

Stephen H. Unger
Professor Emeritus
Computer Science and Electrical Engineering
Columbia University
............

[Leon Smith]

Yeah,  I pretty much exactly agree with your conclusions,  and have for years.   (Though I was raised in a very anti-gambling family,  so when I was a teen I was pretty much against all gambling.   It still holds little appeal for me personally,  though I'd kinda like to learn how to play poker well.   I do know enough about probability that games that involve no skill don't appeal to me in the slightest.)

I'm certainly no expert,  but in extreme cases gambling addiction can be just as bad as alcoholism.   Ever hear of Adam Resnick?  Dateline NBC did a pretty good segment on him,  and he's written a book and had numerous news articles written about him.   Admittedly it's anecdotal,  but it's one hell of a story.

Best,

Leon

[Warren D Smith]

It's a utility-conflict issue.

Utility is not the same thing as money. Consequently, people are willing
to make bets which on average lose them money, if they think it gains
them utility.
So for example, buying a lottery ticket will lose you money in expectation but
perhaps you think the expected utility gain will be positive.
Expected utility gain will
NOT be positive for you for any fair or losing bet with
utility=AnyConcaveDownFunction(Wealth), in particular Shannon's
suggestion
utulity=log(Wealth), but utility can depend ConcaveUp on wealth in
some regions,
such as "getting enough money to live" has way more utility than not, sort
of a cliff-edge part of the graph. If you are in such a situation,
gambling makes
sense. If you are not in such a situation, your brain may have
defects which make it act as though you are.

Related is, say, insurance. Buying insurance is a bet which on
average loses you money. But again, utility is not money and you may
think the expected utility gain is positive, because if some uninsured
disaster happens, you will have low utility.

For example, say you have $1000000 and there is a 0.001 small risk of something
happening which brings you down to $1. To protect against this you buy $2000
of insurance. Note this causes the insurance company to get richer by
$998 expected. If for simplicity utility=log10(wealth) then
ExpectedUtility[Uninsured] = log10(1000000)*0.999 + log10(1)*0.001 =
= 6*0.999+0 = 5.994

ExpectedUtility[Insured] = log10(1000000-2000) = 5.99913.

Here you are better off being insured.

In fact, buying a lottery ticket to get a small chance of a big payoff
is kind of the
"time-reverse" of being uninsured to get a small chance of a big loss.

The only reason insurance & gambling happen is utility does not equal money.

Now, there is another question, which is SOCIAL rather than individual
utility.
Society-wide, gambling is an evil because it wastes time and tends to
redistribute
money from poorer to richer people. It thus hurts society as a whole, although
benefitting some.

What about insurance? Well, the exact same arguments I just made
could be used to
claim insurance is an evil? Also risk-distorting insurance, such as
govt-subsidized
flood insurance offered at artificially low rates, would seem to be a
clear additional evil because it encourages people to live in flood
areas, then recompenses them, thus hurting society overall.

If however, insurance were to be a nonprofit enterprise (thus NOT
about redistributing money from many poorer to few richer people as
overall effect) and NOT about
consuming maximum time and with maximum deception, then it would be
more of a good, and less of an evil.

Now actually, insurance IS gambling. For example, if I make a "bet"
with you that the
Yankees will win a game, that is "gambling." If I buy an "insurance
policy" from you
to "protect" myself against Yankees losing next game, that is... different?

So... if you want to contend one is evil and the other not, you
better think deeper thoughts.