The bolded is an assumption that you're pulling out of thin air and that's getting you into trouble. There is no intrinsic reason for that assumption to be true (of course you can assume it, but it's not a given with how the problem was phrased), in fact it directly contradicts the previous sentence (you have no information => you can't say what the probabilities for your methods are). In fact the next sentence after the bold is also incorrect and requires even more assumptions. The probability distribution can be as asymmetric as you like - here's an example - I only have red cards.

If you still think it's 0.5 - I encourage you to post this in SMP, as I feel like you're not listening to me and maybe you'll listen to someone else (or maybe someone else can please say smth here).

We disagree. Without any information of a bias the probability of a bias in any method is also symmetric. Its 0.5. Its tax time and I am busy but I also hope someone will chime in here. I do not have the time or motivation to start a thread in another forum.

Your example just supports my point. Sure you might have all red cards but with no info of a bias there is a symmetric opportunity for you to have all blue cards.

There has to be other people lurking on this. Where is Sklansky when you need him?

While I "like" the 50% idea, I believe it's controversial. However, in this case I don't know that it applies because I think there are an infinitude of "possible methods" and I don't know how you would graph such a thing.

Having opportunities != having same probabilities
No info about bias != there is no bias = there IS info about bias
etc.

*sigh*

I won't be here for the next 2 days, hopefully someone will chime in and this issue will be cleared up by then.

The assumption here is that given no information about bias, there's no more reason to believe there is bias in one direction than there is to believe that there is bias in the other direction. Since we have no information about which is more likely, to us they are equally likely. It might be more accurate to say "to us their likelihoods are undefined," I guess that's the question. Sometimes I think it can be useful to assume 50%, especially when you have no clue but must take some kind of action based on a probability. At other time I don't like even giving a value because it can be confusing and contradictory (I don't think we can plug 50% into the Drake equation for unknown values, for example).

fyp (edit: I guess you got the fix yourself - I think you basically got it, except that it's never useful to plug in numbers without good reasons)

I have two cards in my hand - each card can be either red or blue. You select one - do you think the probability that you'll get a red one = probability that you'll get a blue one? No (I can easily just have two red ones in my hand). Do you think the outcome is unknown? Yes. Would you be indifferent between choosing the two? Yes - this doesn't mean the probabilities are the same - you're just indifferent because you have nothing to guide your choice (but NOT because you have the probability of 0 that guides you to be indifferent).

Thanks for chiming in. My point is you don't need to know exactly how to make the graph. All you need is that with no bias information, each time you put a point on the red side of the graph you must put an identical point on the blue side. To not do that requires bias information.

Sheesh, just flip a coin to decide which color to choose. Of course you'd pay $0.10 to play this game.

Edit/ I might be misunderstanding the game due to skimming the thread.

To do that ALSO requires bias information.

For reference:

The game is - I have a card in my hand that can be red or blue. I'll pay you $1 if it's red and 0 if it's blue. That is ALL of the information available - you have no clue how I decide what cards to have (I could always have red one, or I could always have blue one, or I could have them with equal probabilities, and so forth).

How much would you pay to play this game? Or equivalently what's the EV of this game?

Gotcha, I thought we got to choose red or blue, my mistake.

Ah, this could be our problem. I thought the game was:

You have a card that could be red or blue. I get to guess. If I guess right I get $1 and $0 if I guess wrong.

As you stated the game you pay $1 on red and nothing on blue. Your method of choosing is unknown. But you have a bias. In any method when you can pick you have an incentive to secure a blue card. Thus the EV now is unknown. It depends how much $1 means to you.

edit:

For the record, this was the first statement of the red/blue issue.

I'm missing the difference between the two formulations...?

For the game to match the OP isn't there only one color that pays off?

Oh, I see, so you want to have a payoff if you pick one color, and no payoff otherwise? Sure, then you always choose the one with the payoff - but how is that in ANY WAY relevant to the two gods examples, lol? In the two gods example you get opposite payoffs from the two choices.

P.S. The whole EV thing is still utterly wrong, but (at zero cost) you'll choose one vs the other because for one EV is >=0 and for the other it is strictly = 0. And you still can't decide whether to play that game or not (for any finite cost), because the EV is still unknown.

P.P.S. I'm in the middle of packing up, so I'm a bit rushing (so the very first line in this post is silly - I'm not going to correct it for consistency) - obviously if the problem is which color to choose - it's the one with the payoff, but the EV of that choice is incalculable. It's not 0.5 or 0 or 1 or anything else. It's incalculable - unknown. Hopefully you'll have this resolved when I'm back - it's been impeding the conversation for too long now.

Talking to you is a lot of work. I have to go back to keep reminding you what we were discussing.

I said:

Thus in the situation where God does not guide the choices then we have neutral ev.

You said:

Thus I can pick either atheism or theism but have no idea which god exists or the probability of the two. Thus the EV of a choice is unknown while I maintained it was zero.

I said:


Then you said:


Then I tried a poker analogy and then moved to cards. The scenario I drew was consistent with your gods example because there were 2 unknown probabilities and a choice between them with rewards. Thats how we got here.

This was too much work. Believe what you want Eddi. Its not worth it to me.

My last deed before I leave:

http://forumserver.twoplustwo.com/47...ple-ev-760405/

The question is copy-pasted from your post. Enjoy

okay. i reread OP. and it has EVERYTHING to do with what you want to be true. you want life to have meaning, and you want your choices to have consequences.

you dont argue at all for why these things are actually best, you simply assume it.

at best, OP argues that theism is the most comforting. but you are kidding yourself if you think its the most logical.

That's pretty semantic.

You could as easily say that once the deck is shuffled, the probability of a spade falling on the turn is either 1 or 0. But that's impractical - it's much more practical to say that the perceived probability (assuming that nobody has indicated having or not having spades) is 9/45, and you need 4 to 1 odds to call. The fact that the top card either is or isn't a spade (rather than being in some 20% spade superposition) is irrelevant to the subjective probabilities, which are really all we mean when we discuss probability in the real world.

Like I wrote earlier, none of this matters unless God will accept that you assume he exists even if you don't particularly believe that he does. There is a good reason to think that God does indeed feel this way and I'm going to post about it later. But some Christians say such a hypothesis contradicts the bible.

I just do not see this. My point seems to be independent of what I want.

1. I am going to die some day - whether I want it or not.

2. After I am dead there will either be an afterlife or there will not - whether I want it or not.

3. If there is no afterlife, my state after I am dead will not be changed by any of my actions in life - whether I want that or not.

4. If there is an afterlife, my state after I am dead may be changed by my actions in life - whether I want it or not.

Where did my desires influence these points?

If it is somewhere else, pull the phrase out and show me because I cannot find it.

I don't know if I made this clear but I basically agree with you. Although I am curious about the good reason you allude to. Also, if you have time could you weigh in on the red card blue card issue. Eddi put it in SMP under simple EV question.

maybe im wrong here. i thought you were trying to show that the meaninglessness that results from a materialistic universe makes the theistic assumption better because it provides meaning. but you are not doing exactly this.

rather it seems you are doing pascals wager in a sort of cost-benefit analysis style...where any choices made in a materialistic universe end up being 0 (at some point in the distant future,) and hence any possible gains from a theistic assumption will be better.

is this a fair assessment?

Everything start out as 50-50 until you have more information. However the mere fact that someone asks the question or even the mere fact that you have decided to think about the question is itself information. Information that could substantially change the price.

There will be no charge for that. I waive my fee when resolving disputes between two Phds.

I didn't even think of that aspect. Thank you.

Allow me to offer you one free chemical engineering problem gratis as quid pro quo.