Fav quotes

[ArtemisVale]

https://www.facebook.com/photo.php?fbid=233392193451366&set=a.173798099410776.6491.172932662830653&type=1&theater&notif_t=photo_comment

[Travis Retriever]

[yt]Dy_LRK2Pkig[/yt]
So first half through and I was surprised how close my answer was with the breast cancer example.  I guessed 7.2%.  The reason being it seemed circular--at least the way it was phrased--"What's the chance she really does have breast cancer?" "Oh, by the way, here's the chance she's going to have breast cancer." (yes, I know that was technically the prior probability--at least I do now, but you gotta admit, that's how it came off.)  So I just multiplied 0.8% by 0.9 (probability of true positive)

[MrBogosity]

[yt]Dy_LRK2Pkig[/yt]
So first half through and I was surprised how close my answer was with the breast cancer example.  I guessed 8%.  The reason being it seemed circular--at least the way it was phrased--"What's the chance she really does have breast cancer?" "Oh, by the way, here's the chance she's going to have breast cancer." (yes, I know that was technically the prior probability--at least I do now, but you gotta admit, that's how it came off.)

Hmmm...right for the wrong reason. Interesting. I wonder how many of the doctors in the survey thought that way and accidentally got close? That number might be higher than 90%!

[Travis Retriever]

Hmmm...right for the wrong reason. Interesting. I wonder how many of the doctors in the survey thought that way and accidentally got close? That number might be higher than 90%!

*edit* I actually guessed 7.2%.  Blah. I had a day from hell today. @_@  So I'm surprised I was still as close as I was.
Anyways, very fascinating video.  Hell, I'm about 15 minutes in and it's been just awesome.  And yeah, I swear my probability education was roughly on par with your algebra education in High School.  I've taken two stats classes and not once did we get into this stuff (though it does seem a bit advanced, mathwise--given the huge Bayesian equation--for beginners, but still!) or even statistical significance.

And yes, one of your best videos yet.  If not *the* best.  And this is coming from a guy who's seen pretty much all of your YouTube videos.

It's such a shame that probability theory isn't emphasized more in schools over, say, Algebra and Calculus.  Hell, even the AP Statistics courses are usually so basic and easy, most students think of it as a slack class.

[dallen68]

It seems to me that you would have to have a good deal of prior knowledge of whatever the topic happened to be to use it effectively. I mean, I guess you could eliminate the ridiculous, but it would be hard to reach a conclusion without further data.

Let IDK=I have no clue.

P1=D(idk)/H ~P0=D(idk)/H

[Travis Retriever]

[yt]Dy_LRK2Pkig[/yt]

Well, actually, regarding e^x = 0.  Try this:  x = negative infinity.  Or at least limit(e^x) as x approaches negative infinity) = 0.
Even my TI 98 Titanium agrees when in exact mode for both cases.
Though ln(x) = -infinity doesn't work when x = 0, I simply take the limit as x approaches zero from the right and it works.
A minor point, but figured it was worth a mention.

Finally, the bit of you saying you never really got the hang of algebra would explain your confusion with Newton's Second Law with MasterGhostKnight when I brought up the equation from Wikipedia.  Guess it's just my own bias as my two year degree with in mechanical engineering, so I've used that equation (sum of the forces = 0 and sum of the forces = ma) to find the forces in static systems a LOT of times.  Not once did we talk about it having anything to do with temperature or whatever.  We did talk about internal forces, but saying the stuff you did reminds me of my mom talking about how the theory of relativity by Einstein is really just like when people traveling fast in an airplane feel like time is traveling differently because they are stressed.  Same with you confusing exponential and 1/r^2 relationships while accusing the people calling you out on it pedantic; while bitching about various other pedantic points in the comments of other videos mentioned above. :\

I've always been one of, if not the, best at physics and math I know.  So I'd like to think I know my shit; even if not as good as I could be.   Even if it doesn't exactly agree with the weird conceptual stuff you were taught like the coordinate substitution showing centrifugal force to not be a fictional force where you put the acceleration = zero where the object being spun is instead of the center.
And considering what I throw out there is the stuff we're taught in physics, math and engineering, here's hoping it's at least right for those purposes, that it what you corrected me with gets taught after the second year, or I suggest you hold your breath every time you cross a bridge, ride your car or board a plane or boat, etc.  If the latter, well, that's just one more problem with the university/college system in addition to the cost...It's bad enough with economics, sociology and psychology (as me and nilecroc) have already covered in various other threads. /rant

[Travis Retriever]

[yt]Dy_LRK2Pkig[/yt]

Which leads me to believe the prior probability for the state's legitimacy, morality and efficacy is probably a bit low.  It's had like, what? 6000 cockshitting years to work?  Hell, evolution's kicked ass and taken names with only 150.  P(State) might not be zero, but goddamn is it close.  Like...10^(-100^100) or some other freakishly low number.
With the bar being set below the cave that was mined under what used to be the bottom of the barrel, statists will have to produce evidence that goes beyond extraordinary to get me to become a believe. Specifically, they'd have to do the following AT LEAST:
1) Solve Mises Economic Calculation Problem for the State. AND prove it better for a free market (the burden of proof is on them.  Why go with violence if the default of no violence is equal or better economically?)
2) Prove that only good people will ever get into their state.
They must do both.  Period.
If they fail 1, they condemn billions to die of starvation and famine, or other horrors caused by economic miscalculation.  Or at least have lower standards of living.
If they fail 2, they condemn billions to die by the brutal hands of evil people twisting state violence for their own agendas.
I'm not sure if these both are mutually exclusive.

[dallen68]

I would call them mutually inclusive. That being said, I suppose it is possible to have a state that *isn't* based on violence. It's never, to my knowledge, happened before... but it's conceivable. For the second part, they don't necessarily have to show evil people will never be in the state. They have to show that evil people will never hold power in the state; and good people will not become evil once in power in the state. Which leads us to: If the premise for the state is violence, then evil people will take power and use the state for their own ends, to varying degrees. If the premise for the state is not violence, then evil people will not take power and use the state for their own ends, to varying degrees.

[MrBogosity]

*edit* I actually guessed 7.2%.  Blah. I had a day from hell today. @_@  So I'm surprised I was still as close as I was.
Anyways, very fascinating video.  Hell, I'm about 15 minutes in and it's been just awesome.  And yeah, I swear my probability education was roughly on par with your algebra education in High School.  I've taken two stats classes and not once did we get into this stuff (though it does seem a bit advanced, mathwise--given the huge Bayesian equation--for beginners, but still!) or even statistical significance.

Keep watching. There's a MUCH easier version of the equation coming. I just had to lay some groundwork for it first.

[MrBogosity]

Well, actually, regarding e^x = 0.  Try this:  x = negative infinity.  Or at least limit(e^x) as x approaches negative infinity) = 0.

Infinity for p=1 and negative infinity for p=0 are not solutions. As you pointed out, you have to use calculus to get there, and you see the limit as infinity. Which, mathematically speaking, means that they can never be reached. No matter how much the evidence for piles up and up and up, it will never reach infinity, hence p never reaches 1, and the same the other way: no matter how much evidence there is against something, it can never actually reach negative infinity, and hence, p likewise never reaches 0.

It's actually not a minor point; it completely reinforces what I said!

[Travis Retriever]

Infinity for p=1 and negative infinity for p=0 are not solutions. As you pointed out, you have to use calculus to get there, and you see the limit as infinity. Which, mathematically speaking, means that they can never be reached. No matter how much the evidence for piles up and up and up, it will never reach infinity, hence p never reaches 1, and the same the other way: no matter how much evidence there is against something, it can never actually reach negative infinity, and hence, p likewise never reaches 0.

It's actually not a minor point; it completely reinforces what I said!

Which was a thought I had, given that we're talking about p being a function of the amount of evidence thinking, "yes, but can that ever be reached or something?"  Like I said, I was really friggin frazzled last night. >.<*
But yeah, I swear, my brain wasn't built to handle probability and statistics.

[Travis Retriever]

[yt]kr5Z-6ub_2k[/yt]
Wonderful. I'm about 4 minutes in.  I notice the additional tests part reminds me of statistical significance and sample size.  Like how larger and larger sample sizes tend to be more reliable, having a larger number of tests makes the result more reliable.  It also reminds me of how at court cases (in your I Thought I Was A Feminist video) how the defense attorney tries to find holes in even the offense's character/reliability etc.  It's like they're trying to lower that prior probability of the person they're defending being guilty/found guilty.

[dallen68]

like they're trying to lower that prior probability of the person they're defending being guilty/found guilty.

OF COURSE THEY ARE!!! THAT'S THEIR FUCKING JOB!!! It's the definition of being a defense attorney.

That being said, If I take a A.I.D.S. test, I expect it to be 100% accurate, because either I have the H.I.V. virus, or I do not.

[MrBogosity]

That being said, If I take a A.I.D.S. test, I expect it to be 100% accurate, because either I have the H.I.V. virus, or I do not.

Then you need to find a different universe to live in. NOTHING is 100% accurate.

[Lord T Hawkeye]

Moral of the story: ALWAYS get a second opinion when it comes to medical tests like that.