Cold Reading Demos at TAM2

[CFLarsen]

Claus... er, I mean "CFLarsen", ya got this one saved for the next time poseur tries the "I got an M.S. in stat" ploy again?

What do you think?

 

[Clancie]

Bill,

I'm entertained at how quickly you want to get into this new diversion. ROFL.

Anything but simply addressing the six questions listed above, clearly and sequentially, eh? Leopards...radioactive decay....obsession. You want to post about anything but the questions from Thanz and T'ai Chi.

What convenient "distraction from the topic at hand" (the "J" count issues, Bill...remember?) will you turn to next?

 

[CFLarsen]

And sigs can vary

But that's the point! Clancie's sig has not changed in ages! With every post she makes, she shows how obsessed she is with me.

You're a little over twice as obsessive over Clancie as she is over you, just going on post count. Well done.

If you need to believe that, then go ahead. It still does not make you right.

 

[T'ai Chi]

(snip)

Looks like Bill is upset with the actual data. No need to get emotional Jr., it's just data.

1. Ill-defined measurement criterion. Do you refer to Claus as "CFLarsen"? No. Does Clancie? No. Search again to find 225 instances of Clancie referring to Claus as Claus. Strange, no?

Uh dippy, I how would I know what Clancie refers to Claus as in all her historical posts? I know that recently at least she refers to him as Claus. CFLarsen is also technically Claus' user name here on this board. Therefore I searched by CFLarsen and by Claus. By doing this I was actually being harder on Clancie, not easier, as I've assessed her more post counts. If you take away N(Clancie, CFLarsen), the evidence is even more in Clancie's favor.

Also the 225 is up from my 224 because Clancie made another post with Claus in the post, obviously.

2. Normalization. Your "filler" excuse is exactly that "filler." Bullsh**.

LOL! Right, there are no filler posts on this board?? hahah, recipes? Simple one liners? Heck, I've recently got over 200 more posts just by bickering with Claus and you over the last couple of weeks.

You didn't do it because you're a greenhorn at stat.

If that comforts you, you are most welcome to keep believing that.

Claus... er, I mean "CFLarsen", ya got this one saved for the next time poseur tries the "I got an M.S. in stat" ploy again?

Please, bring it up.

Often.

Then we can see historical 'obsession' counts and calculate new ones. I'd bet that Claus would still have more. What do you think, Tottle?

 

[T'ai Chi]

You also continue to mess up even the most basic statistical analyses.

Such as... ?

You know who is the one who claims the counts are independent. But according to you, that is not messing things up. Whateva!

 

[T'ai Chi]

What do you think?

The above is an example of what I call a 'filler' post.

 

[Darat]

See my comments below on why I didn't include the sig files or consider (N/total posts)*100%..

By your own arguments Claus and Darat, Clancie could change her sig, and then Claus would be more obsessive again. LOL.

I would like to point out I have made no comment about whether Claus or Clancie are obsessed with each other or not (in this thread) so far.

All I did was to check your figures against your definition of an 'obsessive' post i.e. "which I will define as one poster posting about another poster."

Now using your definition at the time you posted your data your figure was incorrect about the number of times Clancie's posts mentioned Claus or CFLarsen. The correct figure was 3116 not 304.

And yes if Clancie changes her signature file it could alter that figure - and to that I have to say "so what?”. Claus could also change his signature file to say something about Clancie and then that would change the conclusion. But that is exactly as it should be!

I am amazed that you don’t seem to like the fact that new evidence may mean your conclusions have to change!

At the time of your post your conclusion of “Based on the definition of obsessed and the evidence, it would seem that Claus is more obsessed with Clancie, than Clancie is obsessed with Claus.". was incorrect.

At the time of composing this post (and your original post) the correct conclusion based on your definitions is:

“Based on the definition of obsessed and the evidence, it would seem that Clancie is more obsessed with Claus, than Claus is obsessed with Clancie.".


**************************


I will now make a comment about this silly, simplistic view of “obsession” and that it is just silly and simplistic.

Clancie and Claus have a “history” and in my opinion neither of them will let bygones be bygones and posts like mine here can’t possible help so I’m not going to make anymore comments about “Who is the most obsessed with whom?”

I apologise to both Clancie and Claus for continuing this playground nonsense.

 

[T'ai Chi]

If you need to believe that, then go ahead. It still does not make you right.

You're a little over twice as obsessive over Clancie as she is over you, just going on post count.

I am not believing anything, other than the fact that the numbers came out the way they did.

 

[CFLarsen]

“Based on the definition of obsessed and the evidence, it would seem that Clancie is more obsessed with Claus, than Claus is obsessed with Clancie.".

But evidence is not a factor in T'ai Chi's calculations. And, as we have seen, his definitions will change, depending on the mess he has gotten himself into.

I apologise to both Clancie and Claus for continuing this playground nonsense.

Not a problem with me. Facts, always.

 

[T'ai Chi]

But evidence is not a factor in T'ai Chi's calculations.

Nice try, believer.

Not a problem with me. Facts, always.

Great! Then look at the post count, and one draws the conclusion that even the sum N(Clancie, CFLarsen) + N(Clancie, Claus) = 80 + 224 = 304 is still smaller than N(CFLarsen, Clancie) = 612.

Based on the definition of obsessed and the evidence, it would seem that Claus is more obsessed with Clancie, than Clancie is with Claus, in fact about twice as much.

 

[T'ai Chi]

Now using your definition at the time you posted your data your figure was incorrect about the number of times Clancie's posts mentioned Claus or CFLarsen. The correct figure was 3116 not 304.

A big fat 'Wrong' should suffice. I said:

"There is a simple test we can do to rate a poster's 'obsession' with another poster, which I will define as one poster posting about another poster."

If anything, since a sig file is only typed in once, it should only be counted as 1 additional post, instead of as the total post count number of additional posts.

There is also something quite stupid about your, Darat, belief that Clancie's historical posts done way before she typed in her sig file should suddenly count as posts about Claus when this sig file is submitted.

All of this stuff, and for other reasons I've mentioned, is why sig files were not included, only actual post content.

I am amazed that you don’t seem to like the fact that new evidence may mean your conclusions have to change!

I'm amazed you protest a simple count! Wait a sec, I'm not amazed really, that is standard occurance when belief is involved.

At the time of your post your conclusion of “Based on the definition of obsessed and the evidence, it would seem that Claus is more obsessed with Clancie, than Clancie is obsessed with Claus.". was incorrect.

Wrong. You don't seem to think too highly of reality.

Claus got schooled, in a major way. He should get a Tottle award, and a SkepticReport article should be written about it.

 

[Clancie]

It's a pain to go through old threads and posts. But, yes, I think Claus is totally obsessed with posting to and about me. So, in the spirit of science, I will keep a tally, from this post on. (I'm curious myself, although I think the result will be...disturbing).

Tally as of 2:23 PM....

Clancie: 1
Claus: 0

Hey, I'm already ahead with this post!

I just hope we keep it that way.



*end of my posts on this topic...unless, of course, a new pattern emerges over time....

 

[BillHoyt]

Looks like Bill is upset with the actual data. No need to get emotional Jr., it's just data.



Uh dippy, I how would I know what Clancie refers to Claus as in all her historical posts? I know that recently at least she refers to him as Claus. CFLarsen is also technically Claus' user name here on this board. Therefore I searched by CFLarsen and by Claus. By doing this I was actually being harder on Clancie, not easier, as I've assessed her more post counts. If you take away N(Clancie, CFLarsen), the evidence is even more in Clancie's favor.

Also the 225 is up from my 224 because Clancie made another post with Claus in the post, obviously.

Upset? Not at all. You've laid down more evidence of being a poseur. As Claus has pointed out, your premise is utterly flawed. He was referring to her sig line. And, not to confuse you with more facts, but the fact of the matter is you ignored the need for data normalization, something a first year stat student would rarely do. Claus posts more. Whether you wish to paint the posts as filler or not, they need to be factored into any proper measure of this stupid thing whose measurement you are so badly mangling.

You totally ignored the need to normalize, and then tried to cover that basic error with this "filler" blunderbuss and a link to a whole page of papers. How McCarthyesque!

 

[Clancie]

Bill,
You're really into the "Clancie count"! rofl.

I guess anything beats responding to those pesky "J" questions T'ai and Thanz have asked you, doesn't it?

 

[CFLarsen]

Bill,
You're really into the "Clancie count"! rofl.

I guess anything beats responding to those pesky "J" questions T'ai and Thanz have asked you, doesn't it?

I think this post is very symptomatic: T'ai Chi is the one who brings up his (flawed) analysis, but when Bill Hoyt replies to it, you are all over him, mocking him for it.

Hey, you know what? I am beginning to think it's not just me, you are obsessed with. I mean...just look at how you hound Hoyt from thread to thread....

 

[T'ai Chi]

Upset? Not at all.

Yes, you are, and now you are in denial apparently. Just to recap:

N(Clancie, CFLarsen) + N(Clancie, Claus) = 80 + 224 = 304,

304 < N(CFLarsen, Clancie) = 612.

To summarize in English for you Jr.: Claus writes more about Clancie than Clancie about Claus, about twice as much based on a count of posts, and that is even after considering (in addition) Clancie referring to Claus by his username of CFLarsen. The above evidence (note Bill: NOT belief) makes Claus's claim of Clancie being obsessive, rather empty and stupid. I guess he didn't want to do work and examine the data for some reason.

As Claus has pointed out, your premise is utterly flawed. He was referring to her sig line.

For reasons already pointed out, your and his thinking is utterly flawed, as the sig line is not counted as post content, because:

a) Clancie doesn't type her sig line anew every post, only once in her sig file, then the computer program simply tacks it on. A user doesn't create a sig line in a post, they create it in their user panel.
b) Using your and Claus's own logic say Clancie now has a higher count than Claus. Well, just by changing her sig file content, you are saying, by your own logic (or lack of) that suddenly it could flip and have Claus have a higher count than Clancie? That method of counting doesn't seem too sensible. In fact, it doesn't even seem remotely helpful in answering any question.
c) By Clancie adding material to her sig file (about Claus) suddenly that adds all the posts to her total count of posts about Claus, including posts that she made before adding material about Claus to her sig file???
d) I wanted a pure count, since considering something like (N/total number of posts)*100% is probably not helpful since a lot of posts on bulletin boards are what I classify as 'filler' or 'noise', therefore deflating what might be an accurate measure. If you deny there is such filler, you must have a pretty good belief system in place.
e) The bulletin board Search feature doesn't search sig files.
f) I wanted a simple count, and morover, I'm not doing any inference with this, just descriptive statistics.
g) There is a quote from Claus to Clancie in Clancie's sig file, gee, so using your logic, maybe all of Clancie's posts should also count as Claus talking about Clancie? And moreover, it was something which Claus suggested Clancie include in her sig line.
h) It is an option for users to turn sig files off. Therefore, in addition to all the above reasoning, sig files are distinct from post content.

What I really want to know, is if you and Claus agree with b) or not. That is, if you think that Clancie is currently more obsessive than Claus, do you AGREE or DISAGREE, that if Clancie removes content about Claus from her sig file, that suddenly Claus will be more obsessive than Clancie?

AGREE or DISAGREE?

And, not to confuse you with more facts, but the fact of the matter is you ignored the need for data normalization, something a first year stat student would rarely do.

See above for reasons.

Second, please Jr., show us your actual evidence (something which you seem very pressed to do as of late) that that is something a 'first year stat student would rarely do'. I seek numbers out of your fingers, not words, not your attempt at comparing these new counts to Markov chains, Poisson processes, Brownian motion, queue system, radioactivity, or any other stochastic process that don't really have to do with what we are specifically talking about here.

I'll wait for you.

...blunderbuss and a link to a whole page of papers.

I linked to a whole page of papers in our discussion on these current counts? Evidence please?

You seem to have a problem with that lately... yet I'm accused of being against science?

 

[T'ai Chi]

Bill,
You're really into the "Clancie count"! rofl.

I guess anything beats responding to those pesky "J" questions T'ai and Thanz have asked you, doesn't it?

Oh, I'm sure he'll compare the counts to something, like logistic regression, correlation coefficients, variance inflation factors, multiple linear regression, matrix scatterplots, Pearson residuals, whatever; just as long as he avoids addressing the actual issue and the actual data.

I still can't believe he expects us to accept his belief of independence just because he says it is. Shah! Right! As if!

 

[CFLarsen]

a) Clancie doesn't type her sig line anew every post, only once in her sig file, then the computer program simply tacks it on. A user doesn't create a sig line in a post, they create it in their user panel.

Lame, lame, lame.... Having a computer "write for you", so to speak, does not change the intention. Why does Clancie have it in her sig, if not to state clearly that she ignores me, with very post she makes?

b) Using your and Claus's own logic say Clancie now has a higher count than Claus. Well, just by changing her sig file content, you are saying, by your own logic (or lack of) that suddenly it could flip and have Claus have a higher count than Clancie? That method of counting doesn't seem too sensible. In fact, it doesn't even seem remotely helpful in answering any question.

Why not? We have to look at the data as they are, the moment we look at them. As of now, Clancie mentions me far more than I mention her.

c) By Clancie adding material to her sig file (about Claus) suddenly that adds all the posts to her total count of posts about Claus, including posts that she made before adding material about Claus to her sig file???

Yes. Precisely.

d) I wanted a pure count, since considering something like (N/total number of posts)*100% is probably not helpful since a lot of posts on bulletin boards are what I classify as 'filler' or 'noise', therefore deflating what might be an accurate measure. If you deny there is such filler, you must have a pretty good belief system in place.

You use this a lot, T'ai Chi: If you cannot refute an argument, you simply call people believers, and reject their point altogether.

e) The bulletin board Search feature doesn't search sig files.

It doesn't have to: We know that everytime a poster posts, the sig is attached.

f) I wanted a simple count, and morover, I'm not doing any inference with this, just descriptive statistics.

But why not a simple count? Why make it more complex than necessary? Is that to obscure the result?

g) There is a quote from Claus to Clancie in Clancie's sig file, gee, so using your logic, maybe all of Clancie's posts should also count as Claus talking about Clancie? And moreover, it was something which Claus suggested Clancie include in her sig line.

I suggested that Clancie used my name, instead of making a rather rude innuendo about someone she had on ignore. I found it dishonest to hint, instead of coming clean, and Clancie (finally) agreed.

h) It is an option for users to turn sig files off. Therefore, in addition to all the above reasoning, sig files are distinct from post content.

Ah, but here you dig your own grave, my friend: Until you know the exact number of posters who have turned their sigs off, your numbers are worthless.

Second, please Jr.,

Who are you referring to? Please address people by their names.

You asked a lot of questions here. I only have one for you:

Why does Clancie have it in her sig, if not to state clearly that she ignores me, with very post she makes?

 

[BillHoyt]

T'ai Chi,

You have been asked to account for not doing something any first-year stat student wouldn't have known enough to do: normalize the data. You changed your sig line in response and twice posted the same statement by way of lame explanation:

"See my comments below on why I didn't include content of sig files or consider (N/total posts)*100%.
__________________
http://www.library.adelaide.edu.au/.../stat_math.html"

Following that sig line link, we find a collection of papers, none of which appear to have anything to do with sig lines or reasons why one should not do data normalization. Please explain this lamest of dodges.

 

[Interesting Ian]

I have no formal education in statistics.

I've had none at all, formal or otherwise.