DO the math, DON'T overpay. We make high quality, low-cost math resources a reality.

Thursday, August 1, 2019

4 Mathematical Paradoxes (and How to Solve Them!)

When people talk about math, usually what comes up is a rigidity and order to how things are supposed to go. However, mathematicians have for years uncovered and written about situations that don't make logical sense, and tried to figure out how they can be resolved. While there are thousands of paradoxes from A (Abelson's Paradox) to Z (Zeno's Paradox), this will only cover 4 paradoxes and their solutions. (for more paradoxes, explore this list on Wikipedia: https://en.wikipedia.org/wiki/List_of_paradoxes)

Without further ado, here they are!

1. Achilles and the Tortoise


















The great warrior Achilles is challenged to a race 100 meter long by a Tortoise. Knowing of Achilles' speed, the Tortoise asks for a 10 meter head start. Achilles tells the Tortoise that the head start will be nothing, as he can run 10 meters in a second. The Tortoise responds that, in the time it takes Achilles to run 10 meters, it will have run 1 meter. When Achilles runs that meter, the Tortoise will have run .1 meters. When Achilles runs .1 meters, the Tortoise will have run .01 meters. Thus, Achilles will seemingly never be able to catch the Tortoise despite being much faster.

Achilles, will of course, be able to catch the Tortoise. The paradox can be resolved by thinking about it one of two ways:
- The distance between them gets smaller by a factor of 10. The distance between the two will be decreased by a factor of 10 infinitely many times in theory, so the distance between them will eventually equal zero. At this point, Achilles and the tortoise will be next to each other, and Achilles will pass the Tortoise.
- Let Achilles' speed be 10 and the Tortoise's speed be 1. At time t, Achilles' position will be 10t and the Tortoise's will be t+10. These positions will be equal when 9t=10. After this time, Achilles will be ahead of the Tortoise. What will their position be? 11.11111...

2. The Boy or Girl Paradox

Image result for boy or girl paradox

You're visiting a couple of statisticians who have two children. You know neither of the children's genders. After you knock on the door, a man arrives at the door, with his son poking his head around the corner. What is the probability that the other child is a boy as well, assuming that the probability of a child being a boy is 50% and being a girl is 50%?

It might seem fairly straightforward, given that there's a 50% probability of a child being a boy. However, we have to remember that there are 2 children total. View the problem from the perspective of knowing neither of the children's genders. There are 4 possibilities for the genders of the kids, with the oldest child having their gender listed first: MM, MF, FM, and FF. These probabilities are all equal. Having seen that one child is a boy, we can only eliminate the final option as a possibility. Thus, the three possibilities for the children are MM, MF, and FM. These all have equal probabilities, so the probability that the other child is a boy is equal to one third.


3. The Potato Paradox

Image result for potato paradox

Let's say you have a 100 pound bag of potatoes. 99% of the weight of potatoes is water. You decide to let your  bag of potatoes dehydrate until the weight of the bag is only 98% water. After doing this, you're astonished that your bag of potatoes now weighs 50 pounds!

How does making water a slightly smaller percentage of the weight change the weight so dramatically? We know that the original bag was 99% weight by water, so this means that 1 pound of the weight came from the 'solid' part of the potatoes. Thus, after dehydration, 1 pound of solid accounts for 2% of the weight of the bag. We can thus solve for x, where x is the weight of the bag, for 1 = .02x. Thus, x = 50 pounds.

4. The St. Petersburg Paradox

Image result for st petersburg paradox

While you're walking one night, a friend approaches you with a fair coin and asks to play a game. They tell you that they will give you $2 if it lands on heads once, $4 if it lands on heads twice in a row, $8 if it lands on heads three times in a row, $16 if it lands on heads four times in a row, and so on and so forth. The game will end as soon as the coin lands on tails. After explaining the rules, your friend asks you if you want to wager $100 to play the game. Being a mathematician, you decide to calculate whether or not this would be reasonable. After doing the math, you quickly pull out a $100 bill and ask to play.

Why would you do this? The formula for expected value suggests that this is a smart idea. The formula for expected value is:
E = P(A)*A + P(B)*B +P(C)*C + ....
We plug in the probabilities of the coins landing on consecutive heads and the value you'll be given to find:
E= 0.5*2 + 0.25*4 + .125*8 + ... = 1 + 1 + 1 + 1 + ....
This value will be infinite. While at face value there doesn't appear to be a solution to resolve the paradox, economists and mathematicians have worked on an equation that gives a finite result to this equation - making it more in line with the thinking that one wouldn't bet everything on a game where earning everything back would be very unlikely!

8 comments:

  1. I rather prefer Xeno's Paradox to be a 1 mile race, with Achilles giving the tortoise a 0.9 mile head start. The end result is the demonstration that 0.999... exactly equals 1.

    I'm sure you'll get yells about your (correct) answer to #2. I like this analogy: You have a bowl with 8 balls, 4 each of maroon and fuschia. Each ball is connected with a long string to one other ball, such that the pairs are MM, MF, FM, and FF. You grab one ball at random and it is a maroon one. What are the chances, once you follow that ball's string to its partner, that the partner is fuschia?

    Note the difference if you say that the OLDER child is known to be a male. In my analogy, one of each pair of the balls is larger, such that the pairs are Mm, Mf, Fm, Ff. Grabbing one at random, you see you have an M (not an m) ball. Now the odds are 50%.

    However, your answer to #4 is wrong, because the friend certainly doesn't have infinite dollars. Since there's a cap to the payout, you need a cap to what you pay to play.

    ReplyDelete
    Replies
    1. OK. I take it back. Your answer to #2 is not correct. You've given the odds assuming you asked the parent "Is either of your children male?" and were given a positive response. That is, you have eliminated one of the combinations WITHOUT re-weighting the others. Because you have actually seen one of the children, you HAVE re-weighted the combinations. Consider my analogy of 8 balls:
      By a Baysian analysis, you are equally as likely to have grabbed the MM combination as you are to have grabbed the M member of an MF combination. Note the difference in the analogy if you closed your eyes, grabbed one ball and pulled out it and its partner, and then asked someone to tell you whether or not the pair you are holding includes a maroon ball. This is the equivalent of asking the parent whether either of his children is male, but you might actually be holding a fuschia ball.

      Delete
    2. So the correct answer is still 1/2. You might be seeing the older child of an MM combination, the younger child of an MM combination, the older child of an MF combination, or the younger child of an FM combination. In half of these, the other child is female.

      Compare to my analogy with the balls: Clearly, you could be holding any one of the maroon balls, with equal chance of each. Two of the maroon balls connect to another maroon, and two of them connect to a fuschia ball.

      Delete
    3. Still not convinced? This will do it.
      Let's say that rather than two children, you know that your friend has six children. You also know that he has 5 of one gender and only 1 of the other, but you forget which way it is. So, starting out, you have equal chance of MMMMMF (in any order) and FFFFFM (in any order). However, you spot one child and he is male. By your logic above, you still consider both possibilities to be equal probability. However, Reverend Bayes and I consider the first one to be much more likely.

      Delete
    4. Not to beat a dead horse, but here's one more way to look at it.

      Let's say that you are at a convention of "Families with two children" and you know that the four possible combinations are equally represented. You come across a lad who is lost and you help him find his family. What are the chances that his sibling is male?

      It is 1/2. Half of the boys in that convention have female siblings.

      Delete
    5. The answer to 2 is 1/2 simply because MF=FM so there are initially truly 3 answers MF (or FM) MM and FF and since FF was eliminated at the beginning, only MF and FM remain and the probability of eachoccuring is 1/2

      Delete
  2. Here is the puzzle written so that it does produce the 2/3 chance that you are espousing:

    You and your spouse have just heard about some adorable puppies at a nearby farmer's house, and you both decide to get one. Your spouse wants a male, so you call up the farmer and have the following conversation.

    You: Do you still have any of those adorable puppies left?
    Farmer: Yes, there are two still unclaimed
    You: Is there at least one male available?
    Farmer: Yes.
    You: Thanks! We'll be over in 15 minutes! Bye.

    So you get in the car with your spouse, who, being a fickle sort says, "I've changed my mind. I want to get a female puppy."

    Knowing what you know, what are the chances that the farmer has a male puppy available?

    ReplyDelete
    Replies
    1. Sorry, the last question is supposed to be "Knowing what you know, what are the chances that the farmer has a female puppy available?"

      Delete