Are the questions really randomized

[BassPlayingSugarplum]

Even if the unrandomized tree isn't by difficulty order, the point is that the scrambling still isn't random.

Why bother saying it is? Why even do it?

[Bob78164]

Even if the unrandomized tree isn't by difficulty order, the point is that the scrambling still isn't random.

Why bother saying it is? Why even do it?

Random doesn't have to mean all outcomes are equally weighted. A biased coin can still be random. --Bob

[BassPlayingSugarplum]

From Dictionary.reference.com:

ran·dom [ran-duhm]
adjective


2.
Statistics. of or characterizing a process of selection in which each item of a set has an equal probability of being chosen.

[Bob78164]

From Dictionary.reference.com:

ran·dom [ran-duhm]
adjective


2.
Statistics. of or characterizing a process of selection in which each item of a set has an equal probability of being chosen.

That's not what a mathematician (such as I once was) means by "random." --Bob

[Bob Juch]

From Dictionary.reference.com:

ran·dom [ran-duhm]
adjective


2.
Statistics. of or characterizing a process of selection in which each item of a set has an equal probability of being chosen.

That's not what a mathematician (such as I once was) means by "random." --Bob

I'm glad you're not a mathematician anymore. What's your definition?

[TheConfessor]

From Dictionary.reference.com:

ran·dom [ran-duhm]
adjective


2.
Statistics. of or characterizing a process of selection in which each item of a set has an equal probability of being chosen.

That's not what a mathematician (such as I once was) means by "random." --Bob

I'm glad you're not a mathematician anymore. What's your definition?

As Bob#### stated earlier, if there is bias in the system, an event can be random without producing equally likely outcomes. For example if you played Russian roulette with a single bullet in six chambers of a revolver, you can give the cylinder a random spin and pull the trigger. When it's over, random chance will determine whether you are alive or dead, but one outcome is more likely than the other.

[Bob Juch]

That's not what a mathematician (such as I once was) means by "random." --Bob

I'm glad you're not a mathematician anymore. What's your definition?

As Bob#### stated earlier, if there is bias in the system, an event can be random without producing equally likely outcomes. For example if you played Russian roulette with a single bullet in six chambers of a revolver, you can give the cylinder a random spin and pull the trigger. When it's over, random chance will determine whether you are alive or dead, but one outcome is more likely than the other.

But if there's bias then it's not random. Flipping a penny is not random because it's has a 51% chance of landing on the same face it was launched.

[BassPlayingSugarplum]

But Russian roulette is not a good example because while one OUTCOME is more likely than another, there is an equal chance on the spin landing on any one of the six chambers. So it's random.

[Bob78164]

"Of or relating to a type of circumstance or event that is described by a probability distribution." Note that the probability distribution need not be uniform. A uniform probability distribution is one in which all events are equally likely. --Bob

[Bob Juch]

"Of or relating to a type of circumstance or event that is described by a probability distribution." Note that the probability distribution need not be uniform. A uniform probability distribution is one in which all events are equally likely. --Bob

Okay, but the probability distribution must be measurable.

[Bob78164]

"Of or relating to a type of circumstance or event that is described by a probability distribution." Note that the probability distribution need not be uniform. A uniform probability distribution is one in which all events are equally likely. --Bob

Okay, but the probability distribution must be measurable.

That doesn't mean what I think you think it means. It refers to the concept of Lebesgue measure.

It takes a fair amount of work to construct a set that isn't measurable, and such sets are truly weird. For starters, it's a theorem that any such set must require the Axiom of Choice to be defined. Immeasurable sets are the basis for constructions that, say, divide a cube into four regions and reassemble the regions into a larger volume. --Bob

[BassPlayingSugarplum]

Bob,

I take your point, but I think to suggest than "random," as spoken by Cedric the Entertainer, means anything other than even distribution, would be a bit sneaky and unfair.

[Bob78164]

Bob,

I take your point, but I think to suggest than "random," as spoken by Cedric the Entertainer, means anything other than even distribution, would be a bit sneaky and unfair.

Sneaky, perhaps, though I think ambiguous or misleading would be better words. But unfair? It's not as though there's any way to plan game play based on how the categories might or might not be shuffled. --Bob

[TheConfessor]

It's all irrelevant once you get past the first ten questions, which is a requirement for anyone who actually wants to be a millionaire and tries to make it happen. All the shuffling and randomization is just a distraction from the goal.

[Bob78164]

It's all irrelevant once you get past the first ten questions, which is a requirement for anyone who actually wants to be a millionaire and tries to make it happen. All the shuffling and randomization is just a distraction from the goal.

Not necessarily. It may help choose when to use ATA. --Bob

[BassPlayingSugarplum]

It would absolutely change the strategy if you knew the audience had a reliable success rate on levels 1-5 and that statistically you were likelier to get those questions early.

For instance, in my game I would have tried to answer question 3 (the Gatorade question) or asked the audience, and jumped the question that sent me home. I was pretty sure of the Gatorade question, but not wanting to leave so soon after what it cost me to come out from L.A. And it's hard to gauge the difficulty of the question when you're standing there, if you just get a randomized stack. So if the stack isn't randomized, it tells you use your ATA early and save your jumps for later in the round. If it turns out you don't need them, well, hooray, you have them for round 2.

[TheConfessor]

When I watch the show at home, I often try to predict how well the audience will do when ATA is used, and whether using it was a good move by the contestant. I think I have a pretty good feel for when it will be helpful, mainly based on the subject matter, not on where the question happened to be placed in the question stack. On the other hand, when I was in the hot seat, I used ATA twice, and both were failures. But it was a different game then, with different lifelines and different decisions to make.

I agree with your point about the bad grammar in the question you missed possibly affecting the outcome to some extent. And I appreciate your comments and insights about game show strategy. This message board hasn't had much focus on that in recent years.

[MarleysGh0st]

It's all irrelevant once you get past the first ten questions, which is a requirement for anyone who actually wants to be a millionaire and tries to make it happen. All the shuffling and randomization is just a distraction from the goal.

Not necessarily. It may help choose when to use ATA. --Bob

And also when to use JTQs. They can ask some very obscure questions with "obvious" answers. If you knew the question was what used to be a first tier question, you could go with that obvious choice without much risk. If there's a chance it's one of the higher level WWOQs, that choice might be a distractor.