RobinZ comments on The Irrationality Game - Less Wrong

38 Post author: Will_Newsome 03 October 2010 02:43AM

You are viewing a comment permalink. View the original post to see all comments and the full post content.

Article Navigation

Comments (910)

Comments (910)

Sort By: Best

Show:

You are viewing a single comment's thread. Show more comments above.

Comment author: RobinZ 11 October 2010 04:55:14PM 5 points [-]

I see in your posting history that you identify as a Christian - but this story contains more details than I would assign a 99% probability to even if they were not unorthodox. Would you be interested in elaborating on your evidence?

Comment author: Vladimir_Nesov 11 October 2010 05:38:07PM * 1 point [-]

We should learn to present this argument correctly, since complexity of hypothesis doesn't imply its improbability. Furthermore, the prior argument drives probability through the floor, making 99% no more surprising than 1%, and is thus an incorrect argument if you wouldn't use it for 1% as well (would you?).

Comment author: RobinZ 11 October 2010 06:01:41PM * 8 points [-]

I don't feel like arguing about priors - good evidence will overwhelm ordinary priors in many circumstances - but in a story like the one he told, each of the following needs to be demonstrated:

God exists.
God created the universe.
God prefers not to violate natural laws.
The stories about people seeing angels are based on real events.
The angels seen during these events were actually just robots.
The angels seen during these events were wielding laser turrets.

Claims 4-6 are historical, and at best it is difficult to establish 99% confidence in that field for anything prior to - I think - the twentieth century. I don't even think people have 99% confidence in the current best-guess location of the podium where the Gettysburg Address was delivered. Even spotting him 1-3 the claim is overconfident, and that was what I meant when I gave my response.

But yes - I'm not good at arguing.

Comment author: Vladimir_Nesov comment score below threshold [+] (6 children)

Comment author: RobinZ 11 October 2010 10:24:03PM * 4 points [-]

Edit: 99.8% assumes independence, which is certainly violated in the proposed case.

Here's the thing: in order for nick012000's stated confidence to be justified, every one of these six points must be justified to a level over 99% - and the geometric average must be over 99.8%. The difference between 99% and 99.8% may not be huge in the grand scheme of things, but for historical events it's far from negligible.

Comment author: RobinZ 11 October 2010 06:14:13PM 1 point [-]

Is my elaboration of the "burdensome detail" argument faulty? How would you advise I revise it?

Comment author: Vladimir_Nesov 11 October 2010 06:18:13PM -1 points [-]

Is my elaboration of the "burdensome detail" argument faulty?

"Burdensome detail" is wholly about priors, and you started the elaboration with "I don't feel like arguing about priors", and going on about updating on evidence. Thus, I don't see how you made any elaboration of burdensome detail argument, you've described a different argument instead.

Comment author: RobinZ 11 October 2010 06:47:02PM * 1 point [-]

I think I might see what you mean.

I don't want to argue about the priors for 1-3 specifically. Such arguments generally devolve into unproductive bickering about the assignment of the burden of proof. However, priors for arguments about specific historical events, such as the location of the podium from which the speeches were delivered at Gettysburg, are known to be of ordinarily-small levels, and most evidence (e.g. written accounts) are of known weak strength in particular predictable ways*. In fact, I mentioned Gettysburg specifically because the best-guess location changed relatively recently due to new analysis of the written evidence and terrain. In pure terms of my own curiosity, therefore, I anticipate more interesting discussion on 4-6 than 1-3, as nick012000's evidence for the latter I expect to be wrong in familiar ways.

* cf. Imaginary Positions - rounding to the nearest cliché is a standard failure mode.

Comment author: Vladimir_Nesov 11 October 2010 06:54:05PM -1 points [-]

Such arguments generally devolve into unproductive bickering about the assignment of the burden of proof.

Russel's teapot seems quite settled, and most religions go the same way for similar reasons. This argument is quite strong. Anyway, this is what I referred to; I don't want to discuss evidence about religion.

Comment author: RobinZ 11 October 2010 08:47:08PM 1 point [-]

I do wish to discuss evidence about religion - at least, I do today. I hope nick will oblige.

RobinZ comments on The Irrationality Game - Less Wrong

Article Navigation

Comments (910)

Nearest Meetups

Virtual Study Room

Recent Comments

Recent Posts

Recent Rationality Quotes

Recent Wiki Edits

Top Contributors, 30 Days

Recent Karma Awards

Comments (910)

Top Contributors, 30 Days

You'll need to login or register to do that

(Don't worry, it only takes a few seconds)

Create

Login