I'm teaching a course on quantitative reasoning for an audience of nearly two hundred math-averse students, and one of the ways I'm hoping to "sell the product" to them is to pitch the course as a kind of self-defense art that helps you not get ripped off by used car dealers, cell-phone companies, credit-card companies, banks, etc. For instance, I'll talk about the scam wherein the scammer sends free investment advice to 1024 people (half of whom get one piece of advice and half of whom get a conflicting piece of advice), then sends free advice to the 512 who got good advice on the first round (again splitting his advice half-and-half), then sends free advice to 256 people, and so on, and finally starts to ask a small number of people to pay a lot of money, hoping that they'll reason that someone who's been right so often must be onto something. I'll also show the class the grifting scene in "Paper Moon". Can any of you think of other good mathematical cons, or good resources for finding out about them? ("Ricky Jay's Big Book of Math Hustles" would be perfect if such a book existed!) Jim
1) The "Freakonomics" book has some good examples I think. [Sumo wrestling fixing, real estate agent incentives (they tend to keep their own houses on the market much longer than they do their client's houses) ] 2) Ponzi schemes 3) How much would you pay your neighbor for his $125 million winning Powerball ticket? 4) Money laundering, check kiting, fringe banking, electronic check deposit rules and rights of recission, etc, all of these have interesting stories behind them of scams that people tried at one point or another that bankers responded to with some sort of combined mathematical/rule making response. When large sums have to be paid businesses will go extreme lengths to collect even one day of additional interest (dropping off checks at precisely closing time to obscure offices--"hey, we paid according the contract! you just didn't cash the check!" ) stories like these might motivate people to actually listen to the drone about paying attention to interest rates. Thane Plambeck http://www.plambeck.org/ehome.htm James Propp wrote:
I'm teaching a course on quantitative reasoning for an audience of nearly two hundred math-averse students, and one of the ways I'm hoping to "sell the product" to them is to pitch the course as a kind of self-defense art that helps you not get ripped off by used car dealers, cell-phone companies, credit-card companies, banks, etc.
For instance, I'll talk about the scam wherein the scammer sends free investment advice to 1024 people (half of whom get one piece of advice and half of whom get a conflicting piece of advice), then sends free advice to the 512 who got good advice on the first round (again splitting his advice half-and-half), then sends free advice to 256 people, and so on, and finally starts to ask a small number of people to pay a lot of money, hoping that they'll reason that someone who's been right so often must be onto something.
I'll also show the class the grifting scene in "Paper Moon".
Can any of you think of other good mathematical cons, or good resources for finding out about them?
("Ricky Jay's Big Book of Math Hustles" would be perfect if such a book existed!)
Jim
_______________________________________________ math-fun mailing list math-fun@mailman.xmission.com http://mailman.xmission.com/cgi-bin/mailman/listinfo/math-fun
Here is another topic: The (main) differences between an IRA (pay taxes later) and a ROTH (pay taxes now). In particular, the amount of money you end up with after, say, 6 years is: IRA: $ * i1 * i2 * i3 * i4 * i5 * i6 * t6 ROTH: $ * t0 * i1 * i2 * i3 * i4 * i5 * i6 $ = amount you set aside tk = (1 - marginal tax rate in year k) ik = (1 + interest rate in year k) Since multiplication is commutative, the choices don't look that different unless t0 != t6. However, there are a number of rule differences. In particular, with a ROTH IRA, your limit (is it $3000) bound $, while in a ROTH, you initially deposit the after tax money ($ * t0). So your initial investment $ is typically higher than $3000. I.e., you are allowed to invest more in a ROTH than a regular IRA. If anyone has listened to the money shows, they all seem to pretend this is much harder than it is. David
The mathematics of lotteries and casino games comes to mind. LL ----- Original Message ----- From: "James Propp" <propp@math.wisc.edu> To: <math-fun@mailman.xmission.com> Sent: Thursday, January 19, 2006 10:33 PM Subject: [math-fun] mathematical cons
I'm teaching a course on quantitative reasoning for an audience of nearly two hundred math-averse students, and one of the ways I'm hoping to "sell the product" to them is to pitch the course as a kind of self-defense art that helps you not get ripped off by used car dealers, cell-phone companies, credit-card companies, banks, etc.
For instance, I'll talk about the scam wherein the scammer sends free investment advice to 1024 people (half of whom get one piece of advice and half of whom get a conflicting piece of advice), then sends free advice to the 512 who got good advice on the first round (again splitting his advice half-and-half), then sends free advice to 256 people, and so on, and finally starts to ask a small number of people to pay a lot of money, hoping that they'll reason that someone who's been right so often must be onto something.
I'll also show the class the grifting scene in "Paper Moon".
Can any of you think of other good mathematical cons, or good resources for finding out about them?
("Ricky Jay's Big Book of Math Hustles" would be perfect if such a book existed!)
Jim
_______________________________________________ math-fun mailing list math-fun@mailman.xmission.com http://mailman.xmission.com/cgi-bin/mailman/listinfo/math-fun
There's the famous "Truth or Consequences" problem. You have three doors. Behind one of them is a big prize, behind the other two are booby prizes. You get to pick one door. The host reveals a booby prize behind one of the other doors, and you get a chance to switch. Should you? Most people think it doesn't matter but they suspect the host is trying to trick them into switching away from the big prize, so they won't switch. Also, a problem I remember from grade school. The teacher gave two scenarios about borrowing [say] $100 for a year with $5 interest to be paid: (1) You prepay the interest, or (2) you pay the interest when you pay off the loan after a year. He had the students stand in two lines, one who said the two scenarios were equal, and the other who said one was more advantageous. I seem to remember being the only one on the latter line, and being disappointed that there was no opportunity to explain. And here's an interesting anecdote from real life. A friend was talking to her stockbroker near the end of last year because she had some long term capital gains she didn't want to pay taxes on (even though the current tax rate for them is only 20% (15% fed, 5% state)). So she wanted to take some offsetting losses. He told her he would sell some stuff, wait 31 days (to avoid the wash rule that prevents you from taking a loss if you buy it back too soon), and buy it back. She worried that she might pay more in commissions than she would save in taxes, but he assured her that it wasn't a problem, because his commission was only 1% while the tax rate was 15%. [Here I invoke the advice "Never ascribe to malice what can be adequately explained by incompetence."] As the readers on this list will instantly recognize, the 1% is on the value of the asset being sold, not the amount of the loss, and it has to be doubled to 2% if the asset is being rebought. [Problem left to the reader: what percentage loss leads to break-even between commission and tax savings.] --ms James Propp wrote:
I'm teaching a course on quantitative reasoning for an audience of nearly two hundred math-averse students, and one of the ways I'm hoping to "sell the product" to them is to pitch the course as a kind of self-defense art that helps you not get ripped off by used car dealers, cell-phone companies, credit-card companies, banks, etc.
For instance, I'll talk about the scam wherein the scammer sends free investment advice to 1024 people (half of whom get one piece of advice and half of whom get a conflicting piece of advice), then sends free advice to the 512 who got good advice on the first round (again splitting his advice half-and-half), then sends free advice to 256 people, and so on, and finally starts to ask a small number of people to pay a lot of money, hoping that they'll reason that someone who's been right so often must be onto something.
I'll also show the class the grifting scene in "Paper Moon".
Can any of you think of other good mathematical cons, or good resources for finding out about them?
("Ricky Jay's Big Book of Math Hustles" would be perfect if such a book existed!)
Jim
_______________________________________________ math-fun mailing list math-fun@mailman.xmission.com http://mailman.xmission.com/cgi-bin/mailman/listinfo/math-fun
And of course there's the old chestnut of asking whether you'd prefer an annual raise of $200 to a biennial one of $50. [NB in spite of widespread misuse of `biennial' for `two-yearly', it should mean `bis' (twice -- the French for `encore'!) yearly] With a starting salary of $10000, the former gives halfyearly payments of $5000+5000+5100+5100+5200+5200+... while the latter gives $5000+5050+5100+5150+5120+5250+... On Thu, 19 Jan 2006, James Propp wrote:
I'm teaching a course on quantitative reasoning for an audience of nearly two hundred math-averse students, and one of the ways I'm hoping to "sell the product" to them is to pitch the course as a kind of self-defense art that helps you not get ripped off by used car dealers, cell-phone companies, credit-card companies, banks, etc.
For instance, I'll talk about the scam wherein the scammer sends free investment advice to 1024 people (half of whom get one piece of advice and half of whom get a conflicting piece of advice), then sends free advice to the 512 who got good advice on the first round (again splitting his advice half-and-half), then sends free advice to 256 people, and so on, and finally starts to ask a small number of people to pay a lot of money, hoping that they'll reason that someone who's been right so often must be onto something.
I'll also show the class the grifting scene in "Paper Moon".
Can any of you think of other good mathematical cons, or good resources for finding out about them?
("Ricky Jay's Big Book of Math Hustles" would be perfect if such a book existed!)
Jim
_______________________________________________ math-fun mailing list math-fun@mailman.xmission.com http://mailman.xmission.com/cgi-bin/mailman/listinfo/math-fun
Good idea, but I'm not optimistic about the outcome. There's a good reason that some of the best mathematical minds of the last several centuries have been involved in the analysis of games -- they can be arbitrarily difficult to analyze. Probably the best you can hope for is to instill a respect for the complexity and subtlety required for their analysis, so that these students don't naively think that the answers are obvious. A similarly difficult task is instilling respect for the scientific method. (I'm not trolling here...) The response of the scientific community to "intelligent design" has been anything but intelligent -- instead of utilizing this as an opportunity to explain what science is and how it works, scientists have simply tried to replace one "faith" with another (believe me because I have a Nobel prize -- NOT because a number of experiments have verified my theory). I think that Lewis Carroll found the same problem trying to teach Logic ("What the Tortoise said to Achilles"): "Then Logic would take you by the throat, and force you to do it!" http://www.lewiscarroll.org/achilles.html I think that science fiction writers have discussed this problem at length -- how long will it take human DNA to evolve to the point where rational thinking is the norm? At 08:33 PM 1/19/2006, James Propp wrote:
I'm teaching a course on quantitative reasoning for an audience of nearly two hundred math-averse students, and one of the ways I'm hoping to "sell the product" to them is to pitch the course as a kind of self-defense art that helps you not get ripped off by used car dealers, cell-phone companies, credit-card companies, banks, etc.
For instance, I'll talk about the scam wherein the scammer sends free investment advice to 1024 people (half of whom get one piece of advice and half of whom get a conflicting piece of advice), then sends free advice to the 512 who got good advice on the first round (again splitting his advice half-and-half), then sends free advice to 256 people, and so on, and finally starts to ask a small number of people to pay a lot of money, hoping that they'll reason that someone who's been right so often must be onto something.
I'll also show the class the grifting scene in "Paper Moon".
Can any of you think of other good mathematical cons, or good resources for finding out about them?
("Ricky Jay's Big Book of Math Hustles" would be perfect if such a book existed!)
Jim
Last Spring we (I and my wife) were visiting the Old and New Cathedral in Salamanca. When entering the Old Cathedral from the new one, we had to pay 3 Euros per person. I gave the guy behind a window a 50 Euro banknote. He gave me back some coins (must have been 4 Euros) and a wad of banknotes. In the very moment I started to count the banknotes, he gave me another 10 Euro banknote. It must have been already in his hand. I instantly recalled Scarpia's words in Act I of the opera Tosca, scandalized by Tosca's swearing: In chiesa! BTW, I had the same experience two days before the Salamanca experience, but I do not recall where. Emeric
participants (8)
-
David Wolfe -
Emeric Deutsch -
Henry Baker -
James Propp -
Loren and Liz Larson -
Mike Speciner -
Richard Guy -
Thane Plambeck