Could A Science Professor Really...
davids
a) Not know what Godel's theorem was
b) When he found out, call it an "obscure result" when it is actually regarded as one of the most important results of the 20th centry
c) Not know what a Turing award was?
In addition, doesn't the fact that the said POSER simultaneously claims to be a church goer lower the probability a further 400%? What is up with the said POSER?
Got something to say?
Start your own discussion
97 comments
Latest
Some guys might prefer watching shows about flostery over boxing. But I think guys should try it both ways. I am guessing that most guys are going to prefer boxing. Not that there is anything wrong with flostery. Masculine, straight guys can like that too. I guess.
You can feel free to spend nothing if you want. I want lapdances. They are fun. They cost money. I don't care if the strippers consider me a customer, because that's what I am. If you didn't have fun when you spent money then it's probably good you stopped, but don't make the mistake of thinking your personal preference represents some universal truth or revelation.
It's a win/win situation for the customer. Customers gets to save money AND have more fun.
PS: How long did it take you to pay off your max'ed out credit cards? Was that fun for you?
Going to strip clubs and not spending money is more fun than when you do spend money. But I guess some quack who just spouts off "theories" w/o ever actually trying it won't know.
davids, Here I was happily ignoring you, tired of the whole thing, and now I just have to find out your big supersecret proof. Whitch will be interesting, because I'm not lying.
P.S. Did you ever try to just have a good time at a strip club? Try that experiment, then maybe you'll understand that the reason we all reject your theories is because we go to stripclubs to have fun, not for some sort of self validation by pickeing up strippers and not spending money.
Quandry? Here, I'll help you out:
I am more opened minded b/c I go out and test things out in reality to see what works and what doesn't.
I don't dismiss other peoples ideas like the rest of you do, just b/c they came from someone else, or someone unpopular (RL, for instance) or b/c they don't agree with my intuition.
I respect the ideas of those that don't agree with me or my intuition until I have the chance to go and try it out. That's the scientific method at play.
But after I observe what reality is I have no mercy:
If people will then tell me that things will work better one way just based on their intuition or their experience when they haven't tried the alternative and their ideas do not work better in practice then I have no use for these people.
"you'll alienate yourself from anyone who doesn't kiss your ass and tell you what you want to hear "
Projecting your problems with women/general life problems onto me are you?
"If you accepted that an educated man and scientist actually disagreed with you, well where does that leave you? Since you can't be wrong and a scientist can't be wrong, QED, I'm not a scientist. "
This statement is patently absurd. If I knew two scientists to disagree about anything then one of them would have to be wrong and hence I could not believe that both of them are scientsts.
But Oppenheimer was a Socialist and Teller hated Socialism. Yet I accept that they are both scientists. Yet I know that one of them, perhaps even both would have disagreed vehemently with me.
Spin that one through your "higher than first order logic" (btw I was tempted to throughly demolish your latest claims about Godel's theorem in this thread, but I figured:
What the fuck if it took me 20 posts to show him that he couldn't calculate a limit how long to show him that his latest claims about Godel's theorem were wrong. Plus I have already demonstrated you can't do math even at a first year level, of course you won't be able to do graduate level stuff.)
The reasons I think you are not a scientist are as follows:
a) you lack the intuition
b) the raw knowledge, and
c) intelligence that a scientist would posses.
This thread made it clear that you can't even do 1st year math. And you're a researcher/physicist/professor: Yeah, right!
(Oh, there is also a secret reason that lends overwhelming evidence to the fact that you are not a scientist, but I am keeping that one to myself, b/c I want to know when you are tripping up and lying again in the future: so you better just caught it out with the lying AN b/c I am going to know it)
Back to the bar for another drink for you, old man!
I know my whole post is BS, as was your "proof", first of all because random sampling doesn't apply to posters on this board and any scientist knows that. I just thought it'd be fun to apply your standard back to you. You said "proof" and "exact", therefore any approximation and limits were either willful lies or idiotic mistakes! Any mistake means you have no clue and are obviously lying!. That is the standard you use. You claimed to use Bayes' formula but you modified it! Therefore is wasn't Bayes' Formula!! Liar!
Most of your problems on the board come about because of this inability to see anything except your interpretation as anything other than a willful lie. It's why you constantly cry strawman. Any deviation from your exact quote, even if logically identical in an analogy, is a strawman because it wasn't exactly what you said. That seems to be more of a tactic however. The real quandry to me is why you claim to be so open minded and advanced in your thinking yet continually treat any expression that doesn't agree with yours as obviously idiotic and wrong. Why? It's not a very scientific way of thinking.
I understand why you need to convince yourself I'm not a scientist. You place an excessive amount of importance on credentials and almost deify science. If you accepted that an educated man and scientist actually disagreed with you, well where does that leave you? Since you can't be wrong and a scientist can't be wrong, QED, I'm not a scientist.
Same thing with our views of strippers. You want to date them and you obsess about what their opinion of you is. You go through elaborate theories and practice and read about how to come across as the guy they want to sleep with. Others claim to just have fun and not care. Well if that were the case you'd be the one who is chasing strippers like a PL, not us. QED, we must be lying or in denial.
It may be very comforting and self re-inforcing, but ultimately the results will be the same as on this board, you'll alienate yourself from anyone who doesn't kiss your ass and tell you what you want to hear (i.e. strippers) or doesn't completely agree with everything you say.
In the end you are here on a board, creating multiple identities so you have someone to agree with you and say "yeah, that davids is a smart guy" assuming that it somehow elevates you or legitimizes your opinions. You aren't speaking truth to power. You're trying to promote your own little quirks and obsessions as some sort of universal truth to somehow legitemize them.
Last time. I dont want to date strippers. I don't want to have sex with them. I don't really give a thought to their opinion of me. Don't tell me you know what I really believe because you don't.
The line to being a PL is not crossed because you spent over a certain amount of cash. That's just a symptom. It's crossed when you start to tell yourself you are different than all those other guys, special, and the strippers will see that. You're the one chasing strippers as if you are something special. The only advance you seem to have made is that (if what you say is true) you won't max out all your credit cards doing it.
This troll stuff makes me feel so ...dirty. davids, how have you kept it up for so long?
Beg? I'm giving you a chance to prove you CAN argue. If it is as obvious as you say it's a slam dunk for a genius like you.
You want to say well he's wrong and have that accepted as fact, yet everyone else has to provide doccumentation and proof at the level YOU deem proper, i.e. no person who does not know how to use first order logic and apply Godel's theorems can be a scientist, that's absurd on it's face. If that is your claim, which it seems to be, you're an idiot. If it isn't then here's your chance to explain why you aren't an idiot.
You claim to have been working for 15 years. Assuming you went to college that means you would be at least 35. Give you the benefit of the doubt and say you are a genius child prodigy and started work at 18. That makes you 33 at best. A 35 yo guy trying to pick up 20 yo girls in a stripclub is the definition of a PL.
Now you are also lying. You claim to work, which 95% of americans who are in the workforce demographic do, so that is reasonable, but then you claim to work with the best theoretical mathemeticians in the world. LOL, you are such a f-ing liar. Take the top 1000 theoretical mathemeticians in the world represent and assume half are in the US. For a workforce of 129,739,000 only
2,932,810 work in computers and mathematics according to the US Census Bureau.
Be generous and count them all as possible even though most are system admins etc. Now assume that the top mathemeticians in the world encompases 1,000 people (a very generous assumption) and that half of them are in the US (also generous). That means that only 2% of the workforce is mathemeticians and of
them .17% are the top mathmeticians in the world. Assume that since they mostly work with other mathmeticians and computer scientists and be generous and assume they each work with 10 other people so we can assume that about 1.7% of mathmeticians work with the top mathmeticians in the world, so we see that by
claiming to work alongside mathmeticians, and the top ones in the world the chances are 0.003% or 1 in 28,722. F-ing liar. Especially since he called a ratio Bayes Theorem and then called an approximationa proof and had to be told that his approximation only worked at the limit -> 0 and that Bayes theorem didn't apply at 0, but he used them in his calculations anyway.
I didn't think you were knew anything about math, but now I'm sure. You are a lying old PL, and I just proved it.
The only explanation is that you drink too much. You are always posting late at night. You're probably drunk. You're an alcoholic 35 year old single loser trying to pick up strippers and pretending to know math online to make yourself feel better about your pathetic existance. F-off loser.
Dan Quayle (AN): You knew Jack Kennedy?
Dan Quayle (AN): You were a friend of Jack Kennedy?
Dan Quayle (AN): I am no Jack Kennedy?
Dan Quayle (AN): Prove it! Prove I am no Jack Kennedy!
ATTN TUISCLers: A grave error has been comitted. It is obvious that davids is in way over his head here and not supprisingly got a little flustered. His last post should have read:
I didn't know when I went out to my car this morning that it was still going to be the color xxxxxxxx which is was yesterday. Or to get even more technical that I think it was yesterday. (ANGRY GLORY HOLE PATRONS HAVE BEEN VALDALIZING IT REGULARLY) However, I am as confident telling people that those UGLY OPEN SORES on their COCKS could not have possibly COME FROM MY MOUTH contacting their DICKS. However it is common knowledge that I AM A LIAR AND A BULLSHITTER.
If you've got a sore you know the score davids was patient zero.
P.S. "xxxxxxxx" is not a color it is RUST!
p.p.s. "xxxxxxxx" could also be the color of the SKID MARKS in Clifbar's underwear.
I didn't know when I went out to my car this morning that it was still going to be the color xxxxxxxx which is was yesterday. Or to get even more technical that I think it was yesterday. However, I am as confident telling people that those UGLY OPEN SORES on their COCKS could not have possibly COME FROM MY MOUTH contacting their DICKS. However it is common knowledge that I AM A LIAR AND A BULLSHITTER.
If you've got a sore you know the score davids was patient zero.
P.S. "xxxxxxxx" is not a color it is RUST!
No, you can't point out the problem that is why you won't respond. Tell me the problem if it is so obvious. This is the tactic you always fall back on when questioned, it's too obvious and beneath me to reply.
Interestingly I have the original post you cite as starting our arguments. You did the exact same thing;
http://www.tuscl.com/discuss-thread.asp?…
I raised a point, laid out my assumptions and why I reached them, and gave you a chance to expand on your original (very brief) posts or correct me and point out my faulty assumptions. The harshest thing I said was that your post had a serious logical flaw. Your response was that I was too much of an idiot to argue with and your arguments are as obvious as 2+2=4.
Now I read your post one way. You seemed to me to equate rising gas prices with a bad economy, but allow for the possibility that the economy wouldn't tank. You didn't explain why you weren't sure, you just left it as a hypothetical. You told me that was the wrong way to read it. I laid out my assumptions and I asked for a clarification if my assumptions were wrong. You could have ended the whole thing amicably by saying something like;
"AN, my "if" was that I am not sure that rising gas prices will make the economy tank, so we essentially agree."
But I never found out if that was what you meant. You only said it was hypothetical, which I didn't think changed the fact that you seemed to equate rising prices with a tanking economy, only that you were not predicting either. You immediately assumed my misreading was intentional and hostile, after all yow could your post have been anything less than 100% clear? It was pretty much all downhill from there, huh...
I'm not asking for number theory here. Point the "obvious flaw" in my synopsis of your syllogism. What is it that you mean that I am not reading?
I know theoretical mathemeticians and their intuitive powers.
Theoretical mathemeticians are research scientists.
You do not posess the knowledge and intuitive powers of theoretical mathemeticians.
You are not a research scientist.
If not outline (maybe without recourse to Godel's theorem) exactly how you "know" I am lying about being a researcher. Here I'll even do the syllogism for Godel.
You don't know Godel's theorem
You are not a research scientist.
The implied first step is that all research scientists know Godel's theorem. Is that your contention? All your arguments seem to come down to You can't be a researcher because all researchers should know x off the top of their heads.
I know theoretical mathemeticians and their intuitive powers.
Theoretical mathemeticians are research scientists.
You do not posess the knowledge and intuitive powers of theoretical mathemeticians.
You are not a research scientist.
P.S. I will gladly confess that I am not a theoretical mathmetician and never have been. My knowledge of math is almost entirely applied and practical in nature.
This is just bait to get into a debate on nihilism, which I am sure you pretty good at. Let's put it this:
I don't know when I go out to my car this morning that it is still going to be the color xxxxxxxx which is was yesterday, or to get even more technical that I think it was yesterday. However, I am as confident telling people and beleive that my car is really the color xxxxxxxx as I am that you are NOT a scientist. Is there some minor chance that I am wrong? Did I perhaps forget the color of my own car? Did some prankster come along and paint it overnight on me? Maybe it was stolen and repainted so they could sell it? When I say my color is xxxxxxxx, I actually do not known that. But am I telling a flat out lie? You tell me....
Anyway, I have no intention of debating epistemology with a nihilistic, sophmoric, self-confessed troll so that is all I have to say on that matter. (Did way too much of that when I was younger.)
As for your "syllogism", well the problem is so obvious I am sure even you know what it is yourself, so I'll just ignore your latest troll.
As for you just lying to me: Nope, uh, uh. You told other you were a scientist. Ain't buying that one.
Well, I was mostly describing you. When you say things like "I know you are not a scientist" that is just a plain lie, you don't and can't. I applied the same standard in this latest thread to see how you'd react (you obviously didn't use Bayes theorem, you don't know 6th grade math, etc, etc...) Yep, I've overstated and shaded the truth on a few things recently to make life more interesting for you, but remember I'm YOUR troll, nobody elses, and for the most part I've been truthful even with you. For everyone else, I've been totally truthful.
As for your constant refrain "you can't be a physicist if you don't know Godel's theorem" that's just plain silly, and always has been. It really points out a very limited awareness on your part of the world outside university walls.
However, you've been doing it for way more than one month.
To be honest, I knew (well was 85% certain) that you weren't a science prof even before the stuff about Godel's theorem. You made some claim denigrating the value of post-secondary education, so I figured you didn't have any or were a drop out. But Godel's theorem clenched it. So I had you pegged as a liar very early on.
As for being a troll, I remember once I made a trivial statement about economics which you misread. But rather than just admit you had misread it you pretended that you hadn't and gave some lengthy sophistry why I was wrong. So again, early on I knew you were a troll.
Anyway, this thread was just playing around for me: I wanted to see just how bad your math was, and, dude, it's really bad:
Maybe at grade 12 level, but you definitely won't be passing a freshman course these days. Maybe you took some first year classes 20 years ago, but math (beyond high school level) is clearly not used daily in your work.
AN: "Sucks, don't it."
I guess what really sucks here, is that there is no one really challenging to debate with, or who has much of mind. Other than RL, and, unfortunately, he is does not interact too in threads. It's a real shame to see a discussion board about strip clubs not focus on how to outplay the strippers. Everyone here is just resigned to doing what they want. Bunch of pussies.
This thread was way too easy: you completely fell apart even when we got into any math beyond high school.
Other boards have much more knowledgable and polite posters.
"Can't imagine why you are unpopular on this board? I just gave you about a one month minor taste of what you've been doing for a year. And I wasn't nearly as insulting."
I am unpopular because my beliefs are diametrically opposed to those of your typical strip club regular. Not surprising that I should be unpopular.
No matter, truth is what is more important to me than popularity. And look at the people I would be winning popularity from? PLs, liars, and trolls. Should it matter to me what they think?
Now a couple of final points about Godel's theorem:
a) His incompleteness does not apply to predicate calculus. Predicate calculus is complete. In fact he was the one who proved it was complete with his lesser known completeness theorem.
b) Godel's theorem applies to any consistent logical system general enough to include number theory. Tacking on axioms (like the real number axioms, say) does not make it go away.
So does it apply to physics? Well in physics we are dealing with real and complex numbers, mostly, which are a superset of the integers, so it's applicable there too. If using real and complex numbers made the incompleteness go away, then people would just use real and complex numbers, to derive the result you couldn't get in number theory and no one would care.
Oh, as for your description of what being a research scientist is like: Dude I work with REAL research scientists every day (some of the world's top mathematicians). Have been doing so for about 15 years. I know what powers of intuition, raw knowledge, and intelligence they have: What they are likely to know and what they aren't. That's exactly why I was able to pegged you as a fraud so quickly. Or to borrow a famous political line:
"Abbie, I have known research scientists. Research scientists are friends of mine. You, Abbie, are no research scientist."
You have just experienced a troll. Someone so determined to be right that no statement would go unanswered or criticized, no post left stand without pulling up every contradictory statement or slip of the tounge you've ever made and demanding you account for every one of them. Someone who makes false statements about you and calls you an idiot and a liar and screeches to other board members that you are not to be believed or trusted. And they won't leave you alone. You can't post anything or have any discussion without the troll butting in and making it about himself and his superiority. Wether he is or isn't isn't even the point, you just want him to go away. You hope that ignoring him will do it, but then you fear he'll just keep on without any provocation declaring to anyone who will listen how he won.
Sucks, don't it.
Can't imagine why you are unpopular on this board? I just gave you about a one month minor taste of what you've been doing for a year. And I wasn't nearly as insulting.
ClifBar and davids are already blown. Stick with another handle and re-join the conversation.
http://www.tuscl.com/discuss-thread.asp?…
----------------------------------------------------------------------------------
----------------------------------------------------------------------------------
AN is Scared to Fight Me! Fri, Feb 17, 2006 @ 1:48 am
Posted by: davids
----------------------------------------------------------------------------------
Gentlemen, following parody's advice I've decided to change my name, my attitude my outlook on life, strippers, the universe and everything, so I'm taking a new name to go with my new attidue/identity. wish me well! (On and sorry about that last post, just had to do it once for old times sake.)
ClifBar (formerly, davids)
----------------------------------------------------------------------------------
----------------------------------------------------------------------------------
And your ratio produced exactly no new information. You already claimed 20% of scientists were theists. There's a shocker. You can take your numbers and assumptions and plug them into a formula and get the answer you want. WOW. I'm impressed. I may have to take it back, you may be a scientist after all since a lot of them seem to think that proves something too.
My point was that your "proof" was nothing short of BS. If you were applying the strict mathematical standards of a proof, which you explicitly claimed, it fell like a house of cards based on your first assumption. To quote you;
"AN: Are you familar with Bayes's formula? If not you'll need to review it before you understand this PROOF."
"The EXACT result is given by:"
[empasis mine] Your post neither proves anything nor is it anywhere close to exact. End of story on that. I was never talking about limit theory, you were. As for your assumptions you could easily have solved for P(S|T) (as you eventually did) and then produced the ratio that way, but then I guess that didn't look as scary or impressive in your opinion.
You assumed you could frighten me or bully me into silence, I decided to make you defend every single statement at the standard of a proof, at which point you can't claim it is anything like a proof. It was just you plugging in numbers and making assumptions.
Hey, you know some math. Wow. You can also look things up on Wikipedia. I guess I should tremble in fear and admit that I am an unworthy liar. Except that I'm not.
I said I was a part time college professor and a full time researcher about 2 or 3 times in response to questions. You are the one who keeps obsessing about it. I don't really care for the purposes of this board and don't bring it up, except when you make laughably wrong scientific statements at which point I feel I have the right to correct you.
You want to freeze me out, here's a hint. Don't start posts about me, don't call me a liar, don't make silly claims like "anyone who doesn't know Godel's theorem or have dozens of names of theorems memorized MUST be lying about being a physicist." That implies you know something about both physics and Godel's theorem and how they relate. If you can't prove it, just don't keep posting on it as if it is a fact. It is BS and I know it, as do you.
Anyone as knowledgable about Godel as you would know that Godel's theorems are written in first order logic and predicate calculus, and strictly speaking only apply to that specific branch of mathematical logic, so a person who isn't well versed in those fields knows Godels theorems at only a general descriptive level, as I do. Also to apply Godel to any science other than theoretical math is not really possible unless that science can be expressed in first order logic.
Here's a little fact about scientists, what they're really like as opposed to your vision or projection. Well, it's mostly about me but most of it can be applied generally. I'm interested in my research. It is what I do. I have a few things memorized. These are the things I use all the time. I am familiar with a larger body of knowledge I use infrequently. Anything else, I look up. I no longer have the forms of derivatives and integrals memorized like I did in college. I don't have the names of dozens of theorems on the tip of my tongue. I no longer know off the top of my head the rest mass of a proton or electron, or the value of Planck's constant in both joules and Ev, or any of dozens of other constants that I used to know in both the cgs and mks systems. It just doesn't happen like that. If you use it or teach it every day or frequently you remember, otherwise, you look it up when you need it and remember the class you took 20 years ago. At the end of the day, I go home and enjoy my other hobbies. I may take an extra interest in something scientific, but I don't spend nights reading scolarly journals or learning first order logic and predicate calculus for fun. There are some of those people around, but the vast majority of scientists are much more like me. Deal with it.
Also, one last time, when I go to a strip club, I want to have fun, not date strippers. To get lapdances I need to spend money on strippers. I really don't care what these strippers think of me when I'm done and go home. In the majority of cases I'll never see them again.
Now, since nobody but you or I are reading this thread anymore (I don't count your alter-ego Clifbar, see above) I'll say again, as I've said before, please, freeze me out. I've tried ignoring you many times before, but you refuse to return the courtesy. Hell, I organized the first boycott, I'd be glad to ignore you again. Keep to your flames to your own threads and confine your comments on other peoples threads to the topic, and not to telling everyone how idiotic everyone on the board except you are, and I have absolutely no problem with you.
y=kx/(ax+b(1-x))
not
y=kx(ax+b(1-x))
above
lim(x->0) y/x = k/b when y=kx(ax+b(1-x)) and b != 0, in which case he will have proven that he own claims regarding the LHS were going to infinity were wrong,
OR he will deny this somehow, in which case he doesn't understand limits AT ALL. In either case we have shown that he doesn't have enough math intuitive/ability to be a science professor, and not only that his claim of having a master in astrophysics are pure bunk. I also beleive he claimed to have a bachelor's degree in math, though it was perhaps a dual major. If that is case, another lie by the fat, old, alcoholic loser.
Anyway Clif, this thread succeeded beyond my wildest dreams if he had just shut his face I would still have the odds of him being an anonymous physical science professor at 1 in 10,347, but right now it's around like 1 in 100,000,000. The guy needs to learn to control his strawman and academic pretentions.
P.S.: Thanks for all your help in this thread, Clif. Look forward to reading more of your posts here on TUSCL. You seem to be one of very few posters here who actually has a clue about anything.
Didn't say there. Was a counter example to that your contention that
y/x -> infinity or is indetermenant as x goes to zero is flat out wrong (except when y=k*x).
AN, what if y= kx/(ax+b*(1-x)), where k, a, b are all constants? then what is lim (x->0) of y/x? could it be k/b? What do you think?
y=k*x
y=x^2
y=x^2+x"
I dealt with the first in my next sentance, but I should have mentioned that it can only be arrived at via your questionable algebra. The other two are impossible by Bayes theorem if you set P(S) = x
y = x * P(T|S) / ( x * P(T|S) + P(N)*P(T|N) )
There is no function you can substitute in for P(T|S) to make y = x^2 or x^2 + 1
In other words if you knew Bayes theorem and wanted to approximate P(S|T) for P(S) -> 0 you would simply say 0 rather than go through the division, form a ratio, approximate P(S) = 0 on the right but not the left, then substitute the result back in.
I've followed your thinking and found it lacking.
Now about that algebra.
should read
"your claim that the limit on the LEFT was either INFINITE OR non-determinant would show that bayes's formula was wrong."
How about these counter examples:
y=k*x
y=x^2
y=x^2+x
What you are missing is that P(S|T) is a function of P(S). How do you work out what function it is? Using bayes's formular. Sorry but it does not stay constant like you want it to.
And in any case if your claim WAS true you would have proved that bayes's formula is wrong: Since the limit on the right is finite and determinant, your claim that the limit on the right was either finite and non-determinant would show that bayes's formula was wrong. I think that was the point of Clif's proof.
But go ahead, try and prove bayes's formula is wrong. Let me know how that works out for you. Good luck, man.
Above you made the claim that when P(S)=0
P(S|T) = 0 * P(T|S) / 0 * P(T|S) + P(N) * P(T|N)
That's not true since the rhs is undefined since P(T|S) (do you see it? It's in the numerator on the right) is undefined when P(S)=0. This "equality" of yours is meaningless (along with most of what you post here, since you argue against claims I never even make).
Now P(S|T) does indeed equal 0 when P(S)=0 and P(T)!=0. But the way to prove that is from the definition P(S|T)=P(S&T)/P(T)<=P(S)/P(T)<=0/P(T)=0.
(As a minor nit you left out some brackets on the right should be "(0 * P(T|S) + P(N) * P(T|N))" (if it were you I am sure you would be arguing that I really meant it w/o the brackets, and then if I corrected my typo later you would accusing me of backtracking or something. However , unlike you I don't need to attack strawmen or typos, I am quite capable of defeating what you really intended w/o resorting to dirty tricks.)
RIP, AN. RIP.
That is flat wrong. Any limit is either indeterminant or infinity for y/x as x -> 0. If y is any nonzero number, y/x approaches infinity. If y is 0, then y/x is indeterminant. The ONLY way you can make it work right to the limit is if y is a factor of x. In that one case the function kx/x is k with a discontinuity at x=0.
You can argue all you want but at the limit (the one you used explicitly on the right) the formula, your formula derived from Bayes breaks down because you are not solving for a probability, but a ratio. My point, and you missed it again, was that you pooched it by messing with Bayes formula. If you want to approximate P(S|T) as P(S) -> 0 you can do it in your head. ZERO!! No substitution, no approximation, it is a simple concept. The probability of being a scientist given ANY other conditional or independant probability will approach 0 as the probability of being a scientist approaches 0.
Now as for your approximation, as best I can figure, here's how you did it.
y/x = .2/(1+x)
let x=0, but only on the right
y/x = .2/(1+0) ( x = 0 is still only true on the right)
y/x = .2 ( x = 0 is still only true on the right)
y = .2x (but we won't call it 0 this time just because it moved to the right, now x !=0)
Now, even though y=0 is a better approximation substitute back in
.2x/x = .2/(1+x) (what the heck, just add that term back in now)
Take that over to the math boards for a laugh.
Here's a better approximation
y/x = .2 / (1+x)
y = .2x / (1+x)
y = 0
That is how you approximate y as x -> 0
Now the funny part is that by definition .2x/x = .2 for all x!=0 and .2/(1+x) -> .2 as x -> 0 they will never be equal, and the only time they are anywhere near close is the special case x -> 0
Wrong.
Again you need to try things with real numbers.
I won't have had a very small number,
I would have had k * P(S) where k was about .2.
But I don't know what P(S) is. But I don't care either, as long as it's small but non-zero.
I wasn't interested in the exact probability but how much the probability decreased. That's why I wanted the ratio. That's the 2nd time I am having to say this... Please clue in at some point.
Yes if I had said that P(S)=0, P(T|N)=1, and P(N)=1 and that P(S|T) is exactly equal to P(T|S) * P(S) then I am sure you can find a ton of contradictions. But I never claimed these things so go ahead and have a field day....
Btw, the fact that I said the formula was an approximation, couple with your math (hell I'll even assume you got it right) only implies that P(T) is about equal to 1 in the US. This is not far off: P(T)=0.94 in the US.
Dude, I did use it straight. You are allowed to divide both sides of an equation by non zero numbers. That is like grade 7 algebra.
I want to see how much the probability of someone claiming to be a scientist decreased once they also claimed to be a theist. By what ratio? P(S|T)/P(S) is exactly the ratio I was after, and feel right out of bayes's theorem with a simple legal division. I can't see why this division is so troubling to you: it's a non-zero number involved.
P(A|B) = P(A n B) / P(B)
And yes it is undefined for P(B) = 0 because if A depends on B and B is impossible it doesn't make much sense to do the calculation, does it.
Now as for your previous post, you either stupidly or in an attempt to confuse the issue switched the order of the probabilities.
P(A|B)=P(A&B)/P(B) when P(B)>0.
Back to what we were doing
P(S|T) = P(S n T) / P(T) when P(T) !=0
So in your example P(B) is P(T), not P(S), and the conditional probability is defined in both dependant and independant cases,
For independant events (by definition)
P(A n B) = P(A) * P(B)
So substituting back in
P(A|B) = P(A) * P(B) / P(B)
Or, if A and B are independant P(A|B) = P(A), which is the expected and trivial solution.
If the event A is dependant on the event B, then we must use Bayes formula, which is;
P(A|B) = P(A) * P(B|A) / P(A)*P(B|A) + P(!A)*P(B|!A)
In the case of independant events or dependant events the conditional probability of P(A|B) = 0 if P(A) = 0 and undefined if P(B) = 0, but you were not talking about taking P(B) to a limit of 0, you were talking about taking P(A) to a limit of 0.
"Now, why did I divide? Why not? It's the same formula."
Why not? Because then you are not solving Bayes Formula, you are solving for a ratio which has no meaning whatsoever. If you wanted to solve for a conditional probability you would have used Bayes formula straight and had an answer of a very small number rather than a ratio which was meaningless since you had no idea what P(S|T) was, since that was what you were solving for.
Just for fun, lets go back to your original conclusion;
"Since P(S) is close to zero and P(T|S) is about .2 the demoninator of the rhs is about 1. Thus
P(S|T)/P(S) is about equal P(T|S)"
Now first of all your proof in plain language says "the ratio of the probability of being a Scientist given one is a theist to the probability of being a scientist is approximatly equal to the probability of being a theist given one is a scientist". QED indeed. It is virtually meaningless so I let's put it another way.
The probability that one is a scientist given one is a theist is approximately the same as the probability one is a scientist times the probability that one is a theist given one is a scientist, or P(S|T) = P(S) P(T|S)
Hmm, well you already defined P(S) as nearly 0 so what was the point?
OK, just for fun let's follow the logic...
Use the definition of conditional probability you provided.
P(S|T) = P(S n T) / P(T)
P(T|S) = P(T n S) / P(S)
Take your approximation
P(S|T) = P(T|S) * P(S)
Do some algebra, substituting the definitions
P(S n T) / P(T) = P(T n S) * P(S) / P(S)
P(S n T) = P(T n S) P(T)
But by definition (A n B) = (B n A) so your proof consists of saying that
a) P(T) = 1, i.e. all persons believe in god, or
b) You had no idea what you were doing
http://en.wikipedia.org/wiki/Conditional…
In particular pay very close attention to this:
If P(B) = 0, then P(A | B) is left undefined.
I would also recommend that you read the articles on limits, since you seem to be clueless about them.
"The conditional probability of S given T where they are not independant is in fact 0 if the probability of S is defined as 0."
Wrong conditional probabilites are only here when P(S) != 0. Why b/c P(S|T)=P(S&T)/P(S). "
Ooops, misread what you original wrote. You are almost right:
P(S|T)=P(S&T)/P(T)
so your statement is no quite true since you need to qualify it by saying that that P(T) != 0 as well...
which is of course zero.
Wrong the inescapable conclusion is that it is you who doesn't understand bayes's theorem even now.
"Apply the actual form of Bayes formula, not your modified "proof" and it is very consistent;
P(S|T) = 0 * P(T|S) / 0 * P(T|S) + P(N) * P(T|N)"
What a joke. P(T|S) is not defined when P(S) is equal to zero (although it is defined for any other value of P(S)). So your "equality" is wrong. Inescapable conclusion: You do not understand what conditional probability is and hence do not understand Bayes theorem.
"The conditional probability of S given T where they are not independant is in fact 0 if the probability of S is defined as 0."
Wrong conditional probabilites are only here when P(S) != 0. Why b/c P(S|T)=P(S&T)/P(S).
"This is what we call a trivial result. In other words if you define S as impossible, you don't need Bayes Formula. "
I agree with this. Bayes formula doesn't apply when P(S)=0. But I never said P(S)=0, I just said some term were negilible compared to others: e.g .2*.01+0.95*.99 is about equal to .95*.99.
"Your approximation blew up because; "
The appoximation does not blow up. Try it with real number to see that it doesn't. Or post real number showing that it does.
"1) You didn't use Bayes Theorem."
I most certainly did.
"2) You don't understand Bayes Theorem."
No you are the one who doesn't understand it.
"3) You either didn't google the proper page or decided to try and seem smart with your modification."
Huh? No modification was involved. Just a division by of both sides by a non-zero number, which is fully legal in algebra.
"4) Your "proof" was complete BS, just a straight application of a formula, and you couldn't even do that "right."
You strawman representation of my proof was was the complete BS. And you couldn't even get that right.
"5) You just keep digging..."
You're the one who digging baby: Add
"doesn't understand the definition of conditional probabilites to the list"
"doesn't think it's legal to divide both sides of an equation by a non-zero number"
".2x/x is about equal to .2/(1+x) for small positive values of x"
from
P(S|T) / P(S) = P(T|S) / P(S)*P(T|S) + P(N)*P(T|N)
I get the feeling you won't answer."
I think I left out the factor of P(T|S) in .2/(1+x). It should be .2/(1+0.2*x). I intuitively knew it made very little difference to the approximation, since P(S) is small: The limit on the right is still .2: the first term is negilible compared to the first, which was the whole point of the approximation, which you completely missed. Never ever did I claim that P(S) was equal to zero or that I had exact equality. I was never thinking of taking limits. Just trying to make a good approximation. And if you had ever tried things with real numbers you would see how good the approximation is.
Now I guess you are asking about the left hand side, How did I get it as .2x/x? This is also an approximation from bayes's theorem (and I used an approximation, since I only claimed "about equal" in the first place and for the special values of P(S) and P(T|N) and P(N) in our problem space). By bayes theorem, and our special values of the number mentioned
P(S|T) is approximately equal to P(T|S) P(S) is or domain or .2*P(S).
The ratio actually varies a little bit as x drops from 0.1 to 0 but is always very close to P(T|S)/ P(T|N).
If you try it with real numbers you will see that it doesn't vary more than a % or two: as Clif pointed out.
The whole point of the approximation was to get you to concede that a limit of P(S|T)/P(S) does not have to go to infinity or be indeterminate as P(S) goes to zero. It can easily be, and indeed, is a finite number.
y/x = .2 / ( .2x + 1)
and came up with
.2x/x = .2 / (1+x)
I've been trying to get you to explain it for a while now. I will even throw you a bone and say I dont care about the factor of .2 and will allow that for the purposes of a limit
.2x + 1 ~ 1 + x
P(A|B)=P(A&B)/P(B) when P(B)>0.
A conditional probability it's not even defined when P(B)=0. Since P(B) != 0 in our case (there is at least one scientist in the US) it's fine to divide by P(B). Now, why did I divide? Why not? It's the same formula. If I have a fomula which says E=1/2 m v^2 when m!=0. You are telling me that I cannot write it as E/m=v^2/2. It's the same thing. Since I was looking for the ratio of the probabilities, I wrote it my way. Why should the division have confused you?
".2x/x is about equal to .2/(1+x) for small positive values of x"
from
P(S|T) / P(S) = P(T|S) / P(S)*P(T|S) + P(N)*P(T|N)
I get the feeling you won't answer.
|
| 0 * P(T|S)
| P(S|T) = --------------------------
| 0*P(T|S) + P(N)*P(T|N)
|
Take your formula
P(S|T) / P(S) = P(T|S) / P(S)*P(T|S) + P(N)*P(T|N)
P(S|T) / 0 = P(T|S) / 0 + P(N)*P(T|N)
Therefore for the special case you cite where P(N) and P(T|N) are both 1 (which P(N) must be if P(S) = 0 and if as you said all non scientists are theists)
P(S|T) / 0 = P(T|S) / P(T|N)
Which means that for the case you defined, a 0 probability, Bayes Theorem is inconsistent. Except for one thing, you didn't use Bayes Theorem.
This is Bayes Theorem:
P(A|B) = (P(A) * P(B|A)) / (P(A) * P(B|A) + P(!A) * P(B|!A))
Where !A represents the probability of "NOT" A, or even more
simply if A = .95 then !A = .05
Or look it up http://plato.stanford.edu/entries/bayes-…
It is under (1.3) Bayes's Theorem (2nd form).They use a slightly different notation, but it is clear that it solves for only one conditional probability, not a ratio.
or to make my next point next clear, this is Bayes Theorem;
P(A|B) = P(A) * P(B|A) / P(A)*P(B|A) + P(!A)*P(B|!A)
Which you change the variables to represent as you said
"Where S=is a scientist, T=is a theist, N=is a non-scientist."
So the substitution is
S = A
T = B
N = !A
P(S|T) = P(S) * P(T|S) / P(S) * P(T|S) + P(N) * P(T|N)
And then you apparently apply some algebra to give the form
P(S|T) / P(S) = P(T|S) / P(S) * P(T|S) + P(N) * P(T|N)
Why?
It makes no sense since you've defined everything except P(S|T) as a known quantity and the traditional form gives you nothing except your known quantities on the right side. Doesn't make sense from the point of an application or a proof since you are explicitly trying to calculate P(S|T) by your own words and Bayes' Theorem gives you that form unmodified.
The inescapable conclusion is that you don't understand Bayes Theorem.
Apply the actual form of Bayes formula, not your modified "proof" and it is very consistent;
P(S|T) = 0 * P(T|S) / 0 * P(T|S) + P(N) * P(T|N)
The conditional probability of S given T where they are not independant is in fact 0 if the probability of S is defined as 0. This is what we call a trivial result. In other words if you define S as impossible, you don't need Bayes Formula. Your approximation blew up because;
1) You didn't use Bayes Theorem.
2) You don't understand Bayes Theorem.
3) You either didn't google the proper page or decided to try and seem smart with your modification.
4) Your "proof" was complete BS, just a straight application of a formula, and you couldn't even do that right.
5) You just keep digging...
P(S) * P(T|S)
P(S|T) = ------------------------------
P(S)*P(T|S) + P(N)*P(T|N)
I'm not sure where you keep coming up with this representation of 0.2x/x for the left side, but it is beside the point for what I was saying.
OK, so let's say I am dealing with small number positive numbers. On the right I have
0.2 /(1+x) , since x is small I approximate the right as 0.2/(1+0)=0.2/1=0.2.
Now you are saying that I cannot claim that this is equal to a LHS of 0.2x/x when it is clear from context that we are talking about cases when x is not zero? Yeah I treated it as zero on the right, but you are saying the left "goes to infinity"? When? Not for any x in the context of what we are talking about.
Gotta be more careful with those strawmen in the future, AN: you fucking idiot.
AN: "You treated P(S) as zero on the right side, so it has to be treated as 0 on the left side."
Fucking strawman. I treated it as a small number on the lhs, small enough that it does not effect the denominator much especially when multiplied by 0.2, so small that the first term in the denominator on the rhs can be ignored: W/ if the denominator is .9425 w/o it is .9405.
P(S|T)/0= .2
So you are saying that a number approaching infinity is equal to .2"
First of all I never defined P(S) as "approaching zero", that's your strawman. Usually when you create strawmen here you are at least able to create ones that you can defeat (and then you hope no one noticed your strawman in the first place). But this time the strawman blows up in your mongloid face:
Let's say I had defined P(S) as approaching zero:
EVEN AS P(S) approaches zero the lhs does not increase, it stays a constant 0.212. In fact, the only time the lhs is anything other than 0.212 is when when P(S)=0: then lhs is undefined. P(S) would only be zero if there were zero scientists in the USA. But since there is clearly at least one scientist in the USA (it's not you, however), I can divide by P(S) if I want.
Anyway, as you say only two people are following this argument: me and you and we both know what a fucking fraud you are. Fuck you, you god damn loser. Back to the bar for yet another drink in your life of alcoholic misery.
Yawn.
"(Now it's completely beside the point, but your "counter example" is wrong even if I was talking about limits which I wasn't: What is the limit as x goes to zero of
y/5x? Is it infinity? Perhaps? But what if y=x?"
I said, by your own reasoning that if P(S) was nearly 0 (which has a very specific meaning to mathmeticians) then the statement you made
"P(S|T)/P(S) is about equal P(T|S)"
Is plain wrong.
You treated P(S) as zero on the right side, so it has to be treated as 0 on the left side.
Each time you get caught you change the argument.
I'm tired of it. I don't know who you are trying to convince since you and I are the only ones reading this thread anymore.
Be sure to use lots of caps, then everyone knows you are better than me.
LESSON #4 FOR AN:
JUST ACCEPT YOUR LOT IN LIFE:
YOU ARE A COLLEGE DROPOUT BUM WHO IS A SOPHMORIC ALCOHOLIC AND MERELY A BARTENDER.
YOU CAN'T EVEN LIVE YOUR FANTASY OF HAVING BEEN A SCIENCE PROFESSOR ON THE INTERNET SINCE YOU WILL GET CALLED DOWN AND HUMILIATED EVEN THERE.
YOU CAN'T GET ANY PUSSY FOR FREE BECAUSE YOU ARE TOO OLD, FAT, STUBBORN, ILL-TEMPERED AND HAVE DRINKING PROBLEMS. THEREFORE YOU ARE CONDEMNED TO SPEND THE REST OF YOUR REMAINING DAYS AS
A MERE BARTENDER
A PATHETIC LAP-DANCE BUYING, STRIP CLUB REGULAR....
HEY BUT AT LEAST YOU HAVE THE REST OF THE PATHETIC LOSERS ON TUSCL TO KEEP YOU COMPANY AND CONSOLE YOU AND THEY MIGHT EVEN BUY INTO YOUR FANTASIES...
Here Lies
AbbieNormal
??? - Apr 01/2006
RIP
"A 2005 poll by AP/Ipsos surveyed ten countries. Of the developed nations, people in the United States had most certainty about the existence of god or a higher power (2% atheist, 4% agnostic)": Hence 96% theists, makes no difference to the approximation.
As for 1% of the population being scientists, I was probably being generous there. As the number gets lower the formula is closer, even if it was a high as 5% though, my approximation would still be good (plug in real nubmer to see why).
Well, I hope you've learned your lessons AN:
1) Don't debate me on science/math AN. You'll get totally crushed and lose all credibility
2) Don't claim to be a professor of science on the internet when your not: someone is going to call you down
and your most important lesson in this thread (I hope):
3) the limit as x goes to zero of k * x / x where k is a constant is K not infinity. (Fuckin retard... going to laugh my ass about that one for the next year.)
You know if you try and engage anymore you will be SMASHED TO EVEN GREATER SMITHERINS. At least you realize that now so you are disengaging. But too late: The damage is already done and your claims of being a professor are forever blown out of the water. The thread ID has been saved:
You started off the thread in bad shape but by the end we learned the following:
You don't know what bayes's theorem is;
you don't understand basic notations used in probability (what is P(S)?);
you don't understand limits (To show my proof is wrong you claim that the limit as x goes to zero of 0.212 x / x is infinity);
you don't know how convert between number percentage and real numbers
The point is that someone claiming to be a professor of a physical science who thinks that the limit as x goes to zero of 0.212 x / x is INFINITY IS A blatant fraud.
Never mind. Keep your fued in here to save the rest of this.
Classic davids. Always good for a laugh.
"Not understand the notion of limits"
to our ever increasing list of evidence that AN is a FRAUD.
Try it with real numbers to see why you "counter example" is wrong:
I estimate P(S)=0.01: therefore P(N)=P(~S)=0.99.
P(T|N) is about 0.95. So the denominator of the RHS is 0.945, so the RHS itself is 0.2/0.945 or 0.212. Therefore the LHS is not even close to infinity as you claim. If P(S) is 0.01 then P(T|S) is 0.00212 or 1 in 472.
(Now it's completely beside the point, but your "counter example" is wrong even if I was talking about limits which I wasn't: What is the limit as x goes to zero of
y/5x? Is it infinity? Perhaps? But what if y=x?
That's where your reasoning breaks down. I guess you are good enough to fool parodyman, but since his IQ is only 60 that ain't so hard to do.
The more you post he more convinced I become that you not only not a professor of an anonymous physical science but that you are merely a bum who dropped up in his (or her) sophmore year to become an alcoholic and a bartender.
But it is amusing to see you thrash around like a fish out of water.
Back to the bar for a drink, old man!
Laughing
my
fucking
ass
off
at
what
a
clueless
retarded
loser
abbie
normal
is
)
Seriously, break character and let everyone know that you know davids is a joke.
Where to start...
I didn't actually look up Baye's Theorem, I just looked at your "proof". I probably should have just read it last night, but I was tired and assumed you would write something plausible that I'd need to actually look at seriously. How silly of me.
Your "proof," from your own words is this.
P(S|T)/P(S)=P(T|S)/(P(T|S)*P(S)+P(T|N)*P(N))
P(S) ~ 0 If I'd actually read your post I would have stopped there, but let's go on for fun.
P(T|S) = .2
P(T|N) = 1
P(N) = 1
The denominator of the right side is indeed 1.
(.2)*0)+(1*1) = 1
The numerator is .2, so the right side of the equation is .2
P(S|T)/P(S) = .2, but you've defined P(S) as approaching 0, therefore:
P(S|T)/0= .2
So you are saying that a number approaching infinity is equal to .2
You are a laugh riot.
http://en.wikipedia.org/wiki/Percentage
Note the introduction:
"A percentage is a way of expressing a proportion, a ratio or a fraction as a whole number, by using 100 as the denominator. A number such as "45%" ("45 percent" or "45 per cent") is shorthand for the fraction 45/100 or 0.45.
As an illustration,
"45 percent of human beings..."
is equivalent to both of the following:
"45 out of every 100 people..."
"0.45 of the human population..."
One way to think about percentages is to realize that "one percent", represented by the symbol %, is simply the number 1/100, or 0.01.
A percentage may be a number larger than 100; for example, 200% of a number refers to twice the number. In fact, this would be a 100% increase, while a 200% increase would give a number three times the original value. Thus one can see the relationship between percent increase and times increase."
Note that I was a bit informal in trying to explain the precise change in the probabilities in my original post: I hope I made it clear exactly what I meant in my post using bayes's theorem.
Not sure what you can do now, AN. It is completely clear to me what a quack you are now. I guess you just need to hope that no one who knows anything about math reads this thread, b/c if they do you are a exposed as a fraud to them. I am saving the thread ID for future reference. :-)
Since parody has an IQ of 60 you are probably off the hook with him: If you told him that my claim that 7 was prime was wrong and called me an idiot he would just agree and laugh since that's what he wants to believe and reasoning is something he everything can't grok and/or doesn't care about.
Guess all you can hope for now is to coninue to convince idiots like him that you know something.
You and I both know that anyone who knows anything knows: you are a complete fraud. But I guess your average TUSCL'er is not going to fall under the category of knowing anything, so you are probably safe here w/ the rest of them. But still it was funny to see you totally commit suicide in front of me today. I think this thread is as classic the godel's theorem one. Forever and ever now I you will be the "professor of an anonymous physical science who doesn't know godel's theorem, what a turing award is, what bayes's theorem is, what the P(x) means, how to convert a percentage to a fraction..."
Holy fuck...
well you aren't the worst I've killed anyone in a debate online... casualguy went down pretty hard, when I got him to start calling himself a moron last time, then on another board, I had a guy tangle himself up when he admitted that his logic implied that hitler and stalin were people who had good ethics...
COME ON TUSCL, IS AN THE BEST YOU CAN PUT UP AGAINST ME. GIVE ME SOMEONE WHO CAN HOLD THEIR OWN EVEN A BIT IN A FUCKING FLAME WAR!
"Well I obvious meant by a "factor" of 400%"
A factor is a number by which something is multiplied, a percent is a fraction of 100, parts per hundered. At the very least it means you are utterly ignorant when dealing with mathematical language.
You obviously have taken NO MATH. Probability.
What is rhs? Right hand side?
What is P(S) probability of being a scientist.
What is P(S|T) it's a conditional probability: the probability of being a scientist given that one is a theist.
Dude: seriously: stop pretending to be a scientist. I had the odds at 1 in 10347 before today. now you are like 1 in 10,000,000.
Google for bayes's theorem. THis is basic stuff. Come on man.
Fuckin' fraud.
Sorry if omitting the word "factor" threw you, thought it would be pretty obvious to a professor of an anonymous physical science, you did seem to kind of figure it out in your 6AM post. :-)
Ah-ha
now this is way to fucking funny now:
an alleged professor of an anonymous physical science who doesn't know
what a turning award is
what godel's theorem is
what bayes's theorem is
too bad you only made it to stats 101. think you need to go a bit beyond that to learn bayes's theorem. for me it was a senior level probability course.
anyway, that once again seals it on you not be a professor of an anonymous physical since,
i mean the probability such a person could not know godel's theorem is like
1 in 100
could not know what a turing award is
1 in 20
coudl not know bayes's theorem?
what the fuck?? that's near impossible that is basic shit.
maybe 1 in 1000.
fuckin' fraud.
bye-bye, AN.
PS, what does P represent?
P(S|T)/P(S)=P(T|S)/(P(T|S)*P(S)+P(T|N)*P(N))
Is P a function or a sample or a space?
Also this seems wrong as a premise "P(T|N) is about equal 1: nearly all non-scientist Americans are theists."
However you want to define theist the best you can do is 75%. If you have a better source please post it.
Also when you say P(S|T) are you representing a liniar equation, a sample space, a moment? This would all help me along in my attempt to understand.
"Since P(S) is close to zero and P(T|S) is about .2 the demoninator of the rhs is about 1"
You need to define P, please, those of us not on your level need some help.
Oh, and what is rhs?
Explaining these concepts should be fairly simple for such an intellect as yours. I'll wait for a clarification before I post.
In the meantime lets go back to statistics 101.
Probability is between 0 and 100%. Zero is impossible, 100% is certainty. That is statistics 101. Something every poster can understand and look up.
"In addition, doesn't the fact that the said POSER simultaneously claims to be a church goer lower the probability a further 400%? What is up with the said POSER?"
Amazing, you can lower probability by 400%. At best that means a probability of -300%. Please enlighten me. You seem to be into a new branch of statistics here.
No, I don't know Bayes's formula off the top of my head, but there is a difference between you and I. I have completed statistics 101. I understand the concept of probability.
The exact result is given by:
P(S|T)/P(S)=P(T|S)/(P(T|S)*P(S)+P(T|N)*P(N))
Where S=is a scientist, T=is a theist, N=is a non-scientist.
Now I am going to grant you that you are an American, so the following argument makes sense in the context of Americans:
P(T|N) is about equal 1: nearly all non-scientist Americans are theists. P(N) is also about equal to 1 given that so few Americans are scientists (shame on them).
Since P(S) is close to zero and P(T|S) is about .2 the demoninator of the rhs is about 1. Thus
P(S|T)/P(S) is about equal P(T|S)
QED.
If you apply bayes's theorem to your other analogies you will see why your analogies were flawed.
Good luck.
-davids
Explain it again, please. You start with the assumption that someone who claims to be part of a group that represents a subsample of a larger group, and then produce somehow a 400% greater likleyhood they are lying not a member of the group at all? Please, enlighten me.
OK, you have a working knowledge of basic math at the 6th grade level davids, I apologize. You just don't know how to apply it. Keep replying though, every statement discredits you more.
That davids is how one makes a scientific argument.
P.S. I learned to do my percentages in 6th grade, and I didn't need Godel's theorems to do it. It is clear you have absolutely no working knowledge of general mathematics, let alone statistics from your 400% figure. I figured it was a joke till I saw your "explanation".
Here is where the 400% number comes from:
http://www.stephenjaygould.org/ctrl/news…
Look at Table 1 "Comparison of survey answers among "greater" scientists" (given the quality of AN's thinking here you do not doubt that he qualifies as a 'greater' scientist do you?) Anyway, 72% of these scientist do not beleive in god, 7% do beleive and, 21% are unsure. I wasn't sure if agnostics would go to church or not (probably not) so I was generous and said half would go (just to help AN out a bit). So that gives at least 80% who aren't going to church, or at most 1 in 5 who do go. That's how you get 400%, you fucking moron.
Probability? 400% Lower? Did this number fall out of your ass? Where is the scientific accuracy?
Also isn't the poster's belief in palm reading, fortune telling and NLP proof that he knows nothing of science?
What is up with the poster? Will he continue to jerk off the entire TUSCL list with his made up facts & figures? Check back here to find out.
http://www.tuscl.com/discuss-thread.asp?…
----------------------------------------------------------------------------------
----------------------------------------------------------------------------------
AN is Scared to Fight Me! Fri, Feb 17, 2006 @ 1:48 am
Posted by: davids
----------------------------------------------------------------------------------
Gentlemen, following parody's advice I've decided to change my name, my attitude my outlook on life, strippers, the universe and everything, so I'm taking a new name to go with my new attidue/identity. wish me well! (On and sorry about that last post, just had to do it once for old times sake.)
ClifBar (formerly, davids)
----------------------------------------------------------------------------------
----------------------------------------------------------------------------------
So I hope we can get that nonsense of two posters out of the way, it makes you seem even more deranged.
Now as for your argument;
"Let's accept AN's claims and arrive at a contradiction:
The lhs is equal to the rhs except when P(S)=0, by bayes's theorem.
Therefore the limit of the lhs as P(S) goes to zero must be equal to the limit of the rhs as P(S) goes to zero.
But the limit of the rhs as P(S) goes to zero is
P(T|S)/P(T|N)
which is obviously a finite, determinate number. (ie P(T|N) != 0 since some non-scientists believe in god.)
Or, perhaps, AN is claiming that bayes's theorem is wrong? Tehehe..."
Nope, I'm about to prove you are clueless.
Take your formula
P(S|T) P(T|S)
------ = -------------------------
P(S) P(S)*P(T|S) + P(N)*P(T|N)
P(S|T) P(T|S)
------ = -------------------------
0 0 + P(N)*P(T|N)
Therefore for the special case you cite where P(N) and P(T|N) are both 1 (which P(N) must be if P(S) = 0 and if as you said all non scientists are theists)
P(S|T) P(T|S)
------ = --------
0 P(T|N)
Which means that for the case you defined, a 0 probability, Bayes Theorem is inconsistent. Except for one thing, you didn't use Bayes Theorem.
This is Bayes Theorem:
P(A|B) = (P(A) * P(B|A)) / (P(A) * P(B|A) + P(!A) * P(B|!A))
Where !A represents the probability of "NOT" A, or even more
simply if A = .95 then !A = .05
Or look it up http://plato.stanford.edu/entries/bayes-…
It is under (1.3) Bayes's Theorem (2nd form).They use a slightly different notation, but it is clear that it solves for only one conditional probability, not a ratio.
or to make my next point next clear, this is Bayes Theorem;
P(A) * P(B|A)
P(A|B) = -------------------------------
P(A)*P(B|A) + P(!A)*P(B|!A)
Which you change the variables to represent as you said
"Where S=is a scientist, T=is a theist, N=is a non-scientist."
So the substitution is
S = A
T = B
N = !A
P(S) * P(T|S)
P(S|T) = ------------------------------
P(S)*P(T|S) + P(N)*P(T|N)
And then you apparently apply some algebra to give the form
P(S|T) P(T|S)
------ = ----------------------------
P(S) P(S)*P(T|S) + P(N)*P(T|N)
Why?
It makes no sense since you've defined everything except P(S|T) as a known quantity and the traditional form gives you nothing except your known quantities on the right side. Doesn't make sense from the point of an application or a proof since you are explicitly trying to calculate P(S|T) by your own words and Bayes' Theorem gives you that form unmodified.
The inescapable conclusion is that you don't understand Bayes Theorem.
Apply the actual form of Bayes formula, not your modified "proof" and it is very consistent;
0 * P(T|S)
P(S|T) = --------------------------
0*P(T|S) + P(N)*P(T|N)
The conditional probability of S given T where they are not independant is in fact 0 if the probability of S is defined as 0. This is what we call a trivial result. In other words if you define S as impossible, you don't need Bayes Formula. Any sentient mathemetician knows that. Your approximation blew up because;
1) You didn't use Bayes Theorem.
2) You don't understand Bayes Theorem.
3) You either didn't google the proper page or decided to try and seem smart with your modification.
4) Your "proof" was complete BS, just a straight application of a formula, and you couldn't even do that right.
5) You just keep digging...
When I ask a question it isn't always because I don't know. Sometimes I just want you to answer because your answers and explanations truly expose what you know and don't know. I learned that as a professor.