(Untitled)

[squirrelloid]

I haven't posted anything in like forever. I don't expect this will actually change that. I just wanted to rant about people stupidity on TV gameshows

Comments

[bobbzman]

I'm pretty sure that for Deal or No Deal, they screen for people who are not good at probability and game theory. Otherwise the show would be boring, with contestants quickly and callously estimating whether the deals are worth it or not and setting realistic standards of when to quit, rather than jumping all around shouting at the audience and family members. The "game" aspect of the show is simply not that interesting, especially to the average American.

I also note that many recent game shows have tended towards being easy (Deal or No Deal, Who Wants to Be a Millionaire?, Are You Smarter than a 5th Grader?) so that the viewers at home feel smart when they know more than the contestants. Jeopardy is a relic at this point.

[nemene]

"Who Wants to be a Millionaire" is a great example of the new philosophy of game show design. (it was actually worth watching when it was new) The idea is the questions are easy/pop culture so we can laugh at the contestant when they get it wrong, and all feel like we could when if we were on.

Knowing the answer is no longer the hard part, the hard part is recalling it when you need it and remaining confident. The entire game is structured to make you second guess yourself and get rattled.

The interesting part of the game was watching people react to stress, some people fell apart and every once in a while there was someone like one of the first full winners who was completel immune to the "Is that your final answer?
" technique. He only used one life-line he called a friend, and said "I don't need your help I just wanted to tell you I am about to win a million dollars." He kicked ass

[nemene]

Actually the behind door #1 does not apply here at all. IN the "lets make a deal" scenario there is information given in that the hosts gaurantees they will open a non-winning door. IF the game is changed so that there is chance the host will open a winning door then there stops being an adavantage to switching. Simillarly here there is no information given to change the cases, statistically the three remaining cases are identical and there for the swap is worthless, simply to cause anxiety at the end.

[squirrelloid]

Ok, there are two cases left, one of which you chose much earlier. That case has P = n/26 of being the right one, where n is the number of cases equal to or greater in value than the highest value case left. (Since we're considering relatively high value cases, n is small). As its the highest value case left, you obviously haven't revealed it yet - ie, all the cases you chose were by definition 'wrong', because they weren't that case. So you've basically defined all your prior picks as wrong

Optimal Strategy

[nemene]

Optimal strategy for Deal is actually kind of interesting because average return is not necessarily the best strategy. Average return works when you can repeat the same activity again and again, not the case here so risk needs to be given weight by most people. Also while more money is always good (for point of discussion), 400,000 does not necessairly provide twice the benefit of 200,000. For instance with me 200,000 would pay off my mortgage and all other debts plus some. The add 200,000 would be nice but if given the choice between 400,000 50% chance and 200,000 100% chance, I would pick the 200,000 for certain. (I am not risk averse typically, but without a chance to repeat I would hat to face the option of zero).

The above indicates why you would want to take an offer that is below average expectation. In the case where the offer exceeds average, duh take it! I have never seen that happen in the few times I have wated.

Re: Optimal Strategy

[floccinau]

I'm not totally convinced of any argument for rationality based on maximizing average dollars. Let's say my final two options are $.01 and $10,000. The offer is for $5,100. I'm reasonably likely to go for the chance of $10,000, even though this does not maximize my expected dollar earnings and cannot be explained by risk aversion. I don't believe this is irrational -- $5,100 is not enough to make a significant dent in my current financial goals, so I discount the financial reward and play to have fun. (Did I pick the right one? Let's find out!)

In a long run repeatable scenario, maximizing average dollars is close to correct. In a single choice, though, I believe most utility functions diverge significantly from expected dollars even when adjusted for risk, and this is not necessarily irrational.

Your risk aversion point is a decent one, though -- given a choice between a sure $100,000 and a 25% chance of $500,000 (with the other 3 choices small enough to not really matter), I'm likely to go for the $100,000.

Re: Optimal Strategy

[nemene]

I will seed you a point on your first paragraph. There is a value to "Fun" for you the value outways the risk plus $100 (the difference in average return for the two choices). I wasn't considering the scenario where you are down to just low values, at that point you have already lost, so what was the point.