Is Counting an Experiment?

[T'ai Chi]

If I define certain posts that I am interested in, and then count those posts, does that make this an experiment?

Or, does it still remain an observational study?

[biglee]

What would be your comparison group? How would you control for confunding variables? How will your test your null hypothesis?

[Yahweh]

My amateur understanding is that a mere count is an observation...

Of course something to take note is the fact if a subject knows he is being counted, that will no doubt affect the subjects behavior. Results could deviate from their otherwise more accurate results.

[drkitten]

Originally posted by T'ai Chi
If I define certain posts that I am interested in, and then count those posts, does that make this an experiment?

Or, does it still remain an observational study?

Does this have to be an either/or question?

[garys_2k]

Nothing's being tested, so where's the experiment? Simply asking "Are there more of type x or type y? I'll count them." is not an experiment.

[T'ai Chi]

Let me add that I know the answer and am just taking a poll for those on this board who call things I've counted 'experiments'.

[jj]

Originally posted by T'ai Chi
Let me add that I know the answer and am just taking a poll for those on this board who call things I've counted 'experiments'.

Let it be known that T'ai Chi is deceiving us all by the nature of the question he has asked here.

He implies that his "clancie" metric has meaning. He has repeatedly (in his signature and in postings) raised it as an "issue" and suggested or constructed various accusations from it. I asked, then, if this was an experiment to determine the validity of his metric. He replied with a rather disputatious definition of an experiment, to which I replied that without experimental validation, his metric means nothing. Note that he has not provided any evidence or validation, rather he ran off into "what is an experiment", relying on some very lame semantics instead of answering the substance.

From there, we attempted to discuss validation several times, with him dodging the question each time, and also dodging the very simple and obvious bias in his "metric".

At this time, he further tried to shift the burden, claiming that rather than he showing that his metric means anything, I must show it has an error. This is a competely incorret statement on his part, as he is the one implying via his signature and other postings and actions that his metric (N ...) has meaning, and therefore the full and complete burden is on him. Acting as a reviewer, all I have to do (which I have done) is point out a possible bias mechanism. (At this time, I have done more than that, I have pointed directly to an article demonstrating the existance of this bias, fully indicting his metric.)

So, he's run off here to try argumentum-ad-populum.

As is usual for T'ai Chi, he has then reconstructed the question, shifted the bar, and over a period of time, evolved to the question he's asking in this poll.

That is, of course, not the original question, nor is it in the original context.

[T'ai Chi]

Originally posted by jj

Let it be known that T'ai Chi is deceiving us all by the nature of the question he has asked here.


You can tell them what to think, but I'll stick with asking them to evaluate it for themselves.


, relying on some very lame semantics instead of answering the substance.


The difference between an observational study and an experiment is pretty big.


From there, we attempted to discuss validation several times, with him dodging the question each time, and also dodging the very simple and obvious bias in his "metric".


I have a criteria where I count all fruit items in my fridge. I search my fridge and count the items that are fruit or made out of fruit. I count 19 objects. Do I need to validate this? Haven't I validated it by the actual count? Perhaps I counted ketchup by mistake, so there are actually only 18 fruit objects. Viewing it as an estimate, isn't 18 fairly close to 19 to still be useful?

I've said several times now that I have no problem considering N(x,y) as an estimate of a true parameter N. Your example was one that was, conveniently I'd say, found in the very thread right after you brought it up. I challenged you to find any examples prior to today's date; so far you have not, which I think is quite revealing. If you have more example, please, don't keep them secret. If you don't have more examples, then you only have 1, which when studying hundreds (soon to be in the thousands) of posts isn't bad bias at all.

You have failed to realize that an estimate is naturally 'off' by some number from the true parameter (else it wouldn't be an estimate). Therefore, you are overreacting when you claim that N(x,y) being 1 off from N is "serious". I mentioned that there are many surveys of populations where an estimate of a true population might be off by much, much, MUCH more than 1. Does this invalidate them and make them "serious" errors, jj?


At this time, he further tried to shift the burden, claiming that rather than he showing that his metric means anything, I must show it has an error.


If you make a claim, jj, you have a burden too. You claimed it was a serious error, and you have yet to show that. You have yet to show that the situation you specified is rampant, that is, that it occured in places other than, a thread you mentioned it in. And how many times did it occur? Enough to take seriously? Even if it occured 10 times, say. Say N(x,y) counts 1000, but the count, N, is actually 990. Are you telling me that being off by 10 when we have 990 counts is a serious thing?


So, he's run off here to try argumentum-ad-populum.


I merely asked others what they think. I am open for discussion. If anyone thinks as you, they can voice their opinion on it.

It is not argument ad populum because, jj, I am not saying I am correct because people agree with me. This is another claim of yours; another one which evidence is lacking for.


That is, of course, not the original question, nor is it in the original context.

Besides failing to give me a straight answer on Bill providing evidence for his claim about statistical terminology, you've also failed to answer if me counting trees outside of my window is an experiment.

You've also failed, due to lack of data, in demonstrating that N(x,y) is not near N. Being just 1 off from N, you've actually managed to show the opposite, jj, that N(x,y) is actually a good statistic.

[jj]

Originally posted by T'ai Chi
You've also failed, due to lack of data, in demonstrating that N(x,y) is not near N. Being just 1 off from N, you've actually managed to show the opposite, jj, that N(x,y) is actually a good statistic.

The burden of proof is yours, and you are engaging in scientific misconduct when you say "just one off".

If you wish to rescue yourself from this charge, you will show full and convincing evidence that:

1) There is one and only one case where "clancie" occurs in a quote included in a Claus article, where the word "clancie" was originally included by another author.
2) That your "N(x,y)" is meaningful in the fashion that you have used it.
3) That the burden of proof lies with the reviewer instead of you, the author.

You are miseducating people about one of your claimed areas of expertise, and trying to pass off a quacked metric.

[CFLarsen]

If I quote someone who mentions Clancie, but don't mention Clancie myself, why should that be counted as "Claus mentions Clancie"?

Can anyone explain the fairness of that?

[Zep]

T'ai Chi, you are having the greatest of trouble counting properly in the first place. All the rest, including this poll, is therefore irrelevant.

[MRC_Hans]

Counting events in vivo is an observational study. That does not necessarily imply that it gives any meaning. In fact, conclusions can rarely be made from observational studies, because they are most prone to bias.

Just making an abstract observation, here ;).

Hans

[garys_2k]

Counting may demonstrate some correlation, if done often enough and systematically, but not causation.

[Dancing David]

Originally posted by MRC_Hans
Counting events in vivo is an observational study. That does not necessarily imply that it gives any meaning. In fact, conclusions can rarely be made from observational studies, because they are most prone to bias.

Just making an abstract observation, here ;).

Hans


And a good observational study is considered by the same protocol types as an experiement.

With Bias being first and formost, bias in the sample, bias in the measurements, bias in the observation.

Not that I dispute Tai's insistance on hounding Bill about stuff either.

[T'ai Chi]

Originally posted by jj

, and you are engaging in scientific misconduct when you say "just one off".


You've only showed it, so far, to be just one off.

Moreover, we're not doing science here jj, unless you believe me counting is science.


You are miseducating people about one of your claimed areas of expertise, and trying to pass off a quacked metric.

Wrong and wrong. 2 for 2.

[T'ai Chi]

Originally posted by Zep
All the rest, including this poll, is therefore irrelevant.

Including your post, Zep?..

This poll was because people have claimed counting was an experiment. I wanted to ask what people thought about that.

[T'ai Chi]

Originally posted by Dancing David

Not that I dispute Tai's insistance on hounding Bill about stuff either.

I have statistics books written by prominent statisticians that use the word 'skewness' (and use the word 'skew') to refer to the same thing.

But Bill...somehow...knows more than these experts, enough to assert that it is only called 'skew'.

But still... no evidence from him. Why?

[jj]

Originally posted by T'ai Chi


Including your post, Zep?..

This poll was because people have claimed counting was an experiment. I wanted to ask what people thought about that.

Argumentum ad populum, with a skew-phrased question that does not encapsulate the sixtifour of the issue.

In other words, an attempt to gather more quack statistics.

[jj]

Originally posted by T'ai Chi


You've only showed it, so far, to be just one off.


Again, you are the proponent of the metric, it is your obligation to fully verify the metric, including addressing the issues raised by critics. I am not obligated in any fashion to gather any data whatsoever.

Please show that your metric has meaning, or retract it.


Moreover, we're not doing science here jj, unless you believe me counting is science.


This is a skeptic's board, son, if you're not doing science, then you don't get the privilege of insisting that your quack metric means anything.

As Claus has posted in one of the many threads where you're stalking him, you've implied repeatedly that your metric means something.

Since you have, you have taken on the responsibility to fully validate it, to address (sensibly) the comments of critics, and to show your work.

You haven't done any of that.


Wrong and wrong. 2 for 2.

And you continue to miseducate people about statistics, which you imply some expertise in in the "what do we do" thread.

[Gregory]

Originally posted by T'ai Chi
I have statistics books written by prominent statisticians that use the word 'skewness' (and use the word 'skew') to refer to the same thing.

But Bill...somehow...knows more than these experts, enough to assert that it is only called 'skew'.

But still... no evidence from him. Why?

What I don't understand is why you care about this to the extent that you seem to. Bill was wrong? Okay. He's refused to admit it? Bad form. But you're still complaining about a semantic error somebody made several months ago; don't you think it's time to let it go?

[Dancing David]

Originally posted by T'ai Chi


I have statistics books written by prominent statisticians that use the word 'skewness' (and use the word 'skew') to refer to the same thing.

But Bill...somehow...knows more than these experts, enough to assert that it is only called 'skew'.

But still... no evidence from him. Why?

Why ask why?


As they taught me in mental health, don't ask why, ask what.

[T'ai Chi]

Originally posted by jj

In other words, an attempt to gather more quack statistics.

Interesting tactics jj. I reply to you in a thread and it is stalking. You reply to me in a thread and it isn't. Well interesting really isn't the right word.. ;)

Anyway, more emotional speculation about what I supposedly attempt.

So far, the poll shows that the vast majority of people who voted think counting to be observational.

[T'ai Chi]

Originally posted by jj

Argumentum ad populum,

By the way jj, you need to study up on your fallacies. An argument ad populum is saying something is right because a lot of people believe it.

I just asked people for what they thought. I didn't say I was correct because of what the people thought. Therefore, you are incorrect, again.

[T'ai Chi]

Originally posted by Gregory

What I don't understand is why you care about this to the extent that you seem to. Bill was wrong? Okay. He's refused to admit it? Bad form.

It is because Bill constantly tells others to back up their claims by providing evidence for their assertions. Then he doesn't. In fact he calls names, degraded me, then ran when I called him on it (and is still running). He needs to play by the same standards that he holds others to before I take him seriously again.

Moreover, he has called people a's's h'ole, kaffe klatch, dodger, woo woo, tse tse fly, balloon, and other names.

His skepticism appears to be 'do what I say, not what I do, and wihle you're at it, think I am a great skeptic', and I'm pretty surprised that not more skeptics have called him out for such behavior.

Your'e right, he refused to admit it. The bigger question is "Why?".

[jj]

Originally posted by T'ai Chi


By the way jj, you need to study up on your fallacies. An argument ad populum is saying something is right because a lot of people believe it.


Duh.


I just asked people for what they thought. I didn't say I was correct because of what the people thought. Therefore, you are incorrect, again.

Of course, citing the opinion of random people is also appeal to false authority, but you knew that, didn't you? Well? Didn't you? Asking it isn't QUITE citing it. Oh, wait, we can all see the results. OOOPS, they're self-citing. Guess it IS appeal to false authority.

Tsk, tsk!

It is, of course, also ad-populum by its very nature. If anyone, anywhere, can see the results, like we all can, it's ad-populum to the heart.

[jj]

Originally posted by T'ai Chi


It is because Bill constantly tells others to back up their claims by providing evidence for their assertions. Then he doesn't. In fact he calls names, degraded me, then ran when I called him on it (and is still running). He needs to play by the same standards that he holds others to before I take him seriously again.

Would that be an argument for ethical relativism, T'ai Chi? Just maybe?

[jj]

Originally posted by T'ai Chi


Interesting tactics jj. I reply to you in a thread and it is stalking.

#define SARCASM 1

Yeah, wander over to CT, post a snippy, false reply in my discussion with someone else, right, that's "reply to you in a thread" and nothing more. Got a bridge for sale, too?

#undef SARCASM

Sorry, bub, you're transparent at this point.

[T'ai Chi]

Originally posted by jj

Of course, citing the opinion of random people is also appeal to false authority, but you knew that, didn't you? Well? Didn't you?


I asked people to vote. Apparently you have a problem with seeing peoples' opinions.


Guess it IS appeal to false authority.


That is incorrect, as mentioned to you already, because I am not saying I am correct because the majority of them agree with me. I am merely seeing what people think.


It is, of course, also ad-populum by its very nature. If anyone, anywhere, can see the results, like we all can, it's ad-populum to the heart.

Just asking for people's opinions is ad populum? That's a new one! You're trying to define the terms the way you want them to be defined. I'm not buying it.

[T'ai Chi]

Originally posted by jj

Would that be an argument for ethical relativism, T'ai Chi? Just maybe?

jj, you keep straying from the issue. Just address it for once.


The issue:
--
It is because Bill constantly tells others to back up their claims by providing evidence for their assertions. Then he doesn't. In fact he calls names, degraded me, then ran when I called him on it (and is still running). He needs to play by the same standards that he holds others to before I take him seriously again.
--

[T'ai Chi]

Originally posted by jj

Yeah, wander over to CT, post a snippy, false reply in my discussion with someone else, right, that's "reply to you in a thread" and nothing more. Got a bridge for sale, too?


Is that anything like when I post a poll, and you come in and snipe? Or when I post to another poster, and you reply to that post?

A "false reply"; you're inventing new terminology right and left. :)

[biglee]

Originally posted by jj


Of course, citing the opinion of random people is also appeal to false authority, but you knew that, didn't you? Well? Didn't you? Asking it isn't QUITE citing it. Oh, wait, we can all see the results. OOOPS, they're self-citing. Guess it IS appeal to false authority.

Tsk, tsk!

It is, of course, also ad-populum by its very nature. If anyone, anywhere, can see the results, like we all can, it's ad-populum to the heart.

Excuse me, but JJ are you saying counting constitutes an experiment?

[CFLarsen]

If I quote someone who mentions Clancie, but don't mention Clancie myself, why should that be counted as "Claus mentions Clancie"?

Can anyone explain the fairness of that?

T'ai Chi, can you explain the fairness of that?

[The Don]

Originally posted by CFLarsen
Can anyone explain the fairness of that?
Of course it's fair. To do a half a$$ed job where you just count the number of posts with a text string in is easy. To do a proper job where the text string context is considered as well is hard and time consuming.

Therefore, in the interests of being able to spout rhetoric with some plausable back-up for a minimum of effort, half a$$ed it is. Of course you could confound TaiChi by referring to the person (either directly or indirectly) as C*ancie (varying the position of the '*') and making the count more difficult.

[biglee]

Originally posted by CFLarsen
If I quote someone who mentions Clancie, but don't mention Clancie myself, why should that be counted as "Claus mentions Clancie"?

Can anyone explain the fairness of that?

T'ai Chi, can you explain the fairness of that?

What does fairness have to do with whether or not counting is an experiment?

[The Don]

I think this is an extension of the "Is Claus stalking/hounding Clancie" threads which are spread all over the forum.

Tai Chi says "Yes" because a high proportion of Claus' posts contain the chanracter string 'c-l-a-n-c-i-e'. Other say "you can't tell" on the basis that a bare count doesn't take into account context (for example quoting a post where someone refers to Clancie). Claus says "No" on the grounds that he's not that kind of chap, he's looking for answers to one of the "Larsen Lists".

[CFLarsen]

Originally posted by The Don
I think this is an extension of the "Is Claus stalking/hounding Clancie" threads which are spread all over the forum.

Tai Chi says "Yes" because a high proportion of Claus' posts contain the chanracter string 'c-l-a-n-c-i-e'. Other say "you can't tell" on the basis that a bare count doesn't take into account context (for example quoting a post where someone refers to Clancie). Claus says "No" on the grounds that he's not that kind of chap, he's looking for answers to one of the "Larsen Lists".

Yup.

[T'ai Chi]

Originally posted by The Don

Of course you could confound TaiChi by referring to the person (either directly or indirectly) as C*ancie (varying the position of the '*') and making the count more difficult.

Well sure. That is certainly a possibility. However, any such deceit would be easily spotted and exposed, and would made the doer look even worse.

[CFLarsen]

Perhaps I should have used a more fitting term. "Fair" also has a meaning in Danish that is close what I tried to say.

How's this:

Can we say anything about whether or not I stalk someone?

[CFLarsen]

T'ai Chi,

I would like you to comment on the two previous posts of mine. In the first, I respond to The Don, but doesn't talk about You Know Who. That would, with your method, increase the count.

However, later, I do, without mentioning You Know Who by name. That would not, with your method, increase the count.

I used no deceptive methods whatsoever.

Don't you see the inherent problem with your method??

[The Don]

I lack all of the skills to answer the following question, maybe someone else doesn't.

How does T'ai Chi's (apologies for earlier mis-spellings) "Clancie count" comapre to Claus' over the past few weeks ? For heaven's sake, about 25% of my postings today contain the name and as I recall I've never engaged Clancie in debate (bu99er, ther I go again, I must be obsessed).

[T'ai Chi]

Originally posted by The Don
I think this is an extension of the "Is Claus stalking/hounding Clancie" threads which are spread all over the forum.


Actually, this particular thread is in response to a few people people claiming that counting is an experiment. So far, it looks like people think that counting is indeed observational.


Tai Chi says "Yes" because a high proportion of Claus' posts contain the chanracter string 'c-l-a-n-c-i-e'. Other say "you can't tell" on the basis that a bare count doesn't take into account context (for example quoting a post where someone refers to Clancie). Claus says "No" on the grounds that he's not that kind of chap, he's looking for answers to one of the "Larsen Lists".

I wanted to see if the claim of Clancie being "obsessive" or "stalking" or "hounding" has any truth to it, or if it is, in fact, the other way around. These words are so vague as to be useless, so actual numbers are needed. People can say whatever they want. I'll personally let the data tell the story.

[The Don]

People can say whatever they want. I'll personally let the data tell the story. [/B]
In which case all you will be able to tell is the proportion of Claus' posts which contain the character string "clancie". Aside from that you can't tell anything.

After all, I'm hvaing a go at you here and yet you're not named

[T'ai Chi]

Originally posted by The Don

In which case all you will be able to tell is the proportion of Claus' posts which contain the character string "clancie". Aside from that you can't tell anything.

After all, I'm hvaing a go at you here and yet you're not named

True, but you've dishonestly modified your behavior to try and "beat" the count, now haven't you?

To show how dishonest that is, all the rest of your posts to or about me in this thread do in fact contain my name. So, you've basically, like jj, constructed a 'special case' which doesn't really occur in reality.

Are you seriously trying to convince me that people respond to other people by quoting, but then removing their name from the post entirely?

[The Don]

Originally posted by T'ai Chi


True, but you've dishonestly modified your behavior to try and "beat" the count, now haven't you?

To show how dishonest that is, all the rest of your posts to or about me in this thread do in fact contain my name. So, you've basically, like jj, constructed a 'special case' which doesn't really occur in reality.

Are you seriously trying to convince me that people respond to other people by quoting, but then removing their name from the post entirely?
No, what I'm attempting to do is to show that the measure you have selected (number of records with that character string) does not relate to the thing you are seeking to investigate (whether or not Claus is harassing Clancie).

A better method (which has already been proposed elsewhere) would be to sample a random selection of Claus posts and to classify them as "Clancie harassing" or not. This would give a metric which could be a basis for a discussion.

There are some flaws in this approach. Firstly, your opponents will insist that the sample size you have taken is too small. A rough calculation shows that for a population of 6000 (Claus' posts) and for a confidence level of 95% and a confidence interval of 3% you'll need to sample around 900 posts. The second problem is that assessment will be subjective to an extent. A comment that one person may think is perfrectly innocent could be classified by someone else quite differently. I don't think there is a way around this one. Thirdly, there's the question of what counts as harassing, 5%, 10%, 20%, 50% of posts. What is a resonable value ?

My failing to include your name was in fact a coincidence. I got to the end of my post and noted it. If someone were to check, I guess that around a third of my posts refer to someone by name, the others are just tossed out there. In terms of quoting, it depends whether the post was initially triggered by pressing the quote button or whether, at a later time, it occurs that some text from another post may be required.

[T'ai Chi]

Originally posted by The Don

No, what I'm attempting to do is to show that the measure you have selected (number of records with that character string) does not relate to the thing you are seeking to investigate (whether or not Claus is harassing Clancie).


First, he claimed that Clancie was being "obsessive".

Second, the pount counts are one small portion of gauging that. Post content is an entirely different matter.

[The Don]

Okey-dokey

No, what I'm attempting to do is to show that the measure you have selected (number of records with that character string) does not relate to the thing you are seeking to investigate (whether or not Claus is obsessive about Clancie).

A better method (which has already been proposed elsewhere) would be to sample a random selection

blah

blah

blah

What is a resonable value ?

[Dancing David]

Gosh Don, stop trying to bring reason into this.

[T'ai Chi]

Originally posted by The Don
Okey-dokey

No, what I'm attempting to do is to show that the measure you have selected (number of records with that character string) does not relate to the thing you are seeking to investigate (whether or not Claus is obsessive about Clancie).

A better method (which has already been proposed elsewhere) would be to sample a random selection

blah

blah

blah

What is a resonable value ?

If you can think of a good way to sample posts randomly, I am all for trying it out.

[Zep]

Originally posted by T'ai Chi

If you can think of a good way to sample posts randomly, I am all for trying it out. Or perhaps YOU can? Hmmm?

[T'ai Chi]

Originally posted by Zep
Or perhaps YOU can? Hmmm?

He is the one who suggested it, not I, Zep.

[The Don]

When you say mechanism, presumably you mean "how can I get access to the text of all 10000(ish) of Claus' posts and from these 10000 select the 900 or thereabouts that will be the right sample size to give a 95% confidence level and a 3% confidence interval".

I think anyone who has seen my posts, my pathetic attempts to attach files and urls and the fact that my avatar is still a Christmas avatar cuold draw the conclusion that I'm not hte number one candidate when it comes to asking questions about making stuff happen on the Internet.

Perhaps someone who is more knowledgable could help me out.

One thing on which I am an expert (insofar as I make a living out of it) is data analysis, data mining and key performance indicator design. I fear that you have fallen into a trap that many of my clients fall into (naturally before they engage me). You have decided to base your metric (which you will use to determine whether or not Claus has been guilty of being obsessive about Clancie) on the information that you can get (proportion of posts containing a certain character string) because the information you really need (the number or proportion of posts in which Claus attacks, hassles whatever, Clancie) is not available to you.

This often happens in business. A business is ACTUALLY interested in their margin information but becuase they haven't got detailed cost inforamtion at a transactional level, they calculate margin against a nominal standard cost. This results in them obsessing about whether or not they've calculated the standard cost correctly (something which doesn't help their bottom line one jot as it's based on historical costs) and not worrying about making sure the unit price is the best they can achieve.

So you have suggested that the post count where Claus refers to Clancie is an acceptable proxy for the information you really need. If this is the case then what is the level which indicates obsession ? 1%, 2%, 5%, 10%, 20% 50% ?

[T'ai Chi]

It is nice to see that most people do not feel counting is an experiment.

[CFLarsen]

Originally posted by The Don
So you have suggested that the post count where Claus refers to Clancie is an acceptable proxy for the information you really need. If this is the case then what is the level which indicates obsession ? 1%, 2%, 5%, 10%, 20% 50% ?

T'ai Chi, do you have an answer to this?

[jj]

Originally posted by T'ai Chi
It is nice to see that most people do not feel counting is an experiment.

Appeal ad populum, after you claimed it wasn't.

Now, let's discuss how your question was misleading.

bzzzt.

[T'ai Chi]

j-square, I said it was nice to see, not that it was proof of anything.

Yeah, let's discuss some questions:

1) Should Jr. provide evidence re: his statistical terminology claim?

2) Should Clauzzz provide evidence re: his claim about Lucianarcy's IP address?

[T'ai Chi]

Originally posted by CFLarsen


T'ai Chi, do you have an answer to this?

Do you have an answer to your claims about Lucianarchy's IP address?? You raging hypocrite...

---
So you have suggested that the post count where Claus refers to Clancie is an acceptable proxy for the information you really need.
---

Claus is the one who claimed Clancie is "obsessive". I merely investigated his claim and asked, How can someone who has less post counts to or about that person be more obsessive that someone who has more post counts to or about that person?, and suggested that estimating N via N(x,y) is just one part of guaging "obsessive"-ness. The actual content of the posts is useful too, as well as some other things, but I had to go with what tools I had and use the forums' search feature.

---
If this is the case then what is the level which indicates obsession ? 1%, 2%, 5%, 10%, 20% 50% ?
---

Ask Clauzzzzz. He's the one who called Clancie "obsessive" in the first place. Perhaps he can clarify...

[jj]

Originally posted by T'ai Chi
j-square, I said it was nice to see, not that it was proof of anything.


Anibus fidat.

[T'ai Chi]

Originally posted by jj


Anibus fidat.


As long as we're doing other languages, I think "Yu so dum" is Chinese.

j^2, care to address these:

1) Should Jr. provide evidence re: his statistical terminology claim?

2) Should Clauzzz provide evidence re: his claim about Lucianarcy's IP address?

[jj]

Originally posted by T'ai Chi



As long as we're doing other languages, I think "Yu so dum" is Chinese.

j^2, care to address these:

1) Should Jr. provide evidence re: his statistical terminology claim?

2) Should Clauzzz provide evidence re: his claim about Lucianarcy's IP address?

Q. Why is T'ai Chi asking me stupid questions like this?

A. So he can claim I won't answer his questions.

[T'ai Chi]

Originally posted by jj

Q. Why is T'ai Chi asking me stupid questions like this?

A. So he can claim I won't answer his questions.

More evasion.

I don't need to claim you won't answer them, because you really won't answer them.

Here's another: Do you still claim that me counting via N(x,y) is "performing an experiment"? jj, do you still believe that?

[T'ai Chi]

Another vote in for 'Observational' I see.

[CFLarsen]

Originally posted by T'ai Chi
Claus is the one who claimed Clancie is "obsessive". I merely investigated his claim and asked, How can someone who has less post counts to or about that person be more obsessive that someone who has more post counts to or about that person?, and suggested that estimating N via N(x,y) is just one part of guaging "obsessive"-ness. The actual content of the posts is useful too, as well as some other things, but I had to go with what tools I had and use the forums' search feature.

...

Ask Clauzzzzz. He's the one who called Clancie "obsessive" in the first place. Perhaps he can clarify...

Not really. It is my opinion that Clancie was obsessed, both with me and BillHoyt. I did not try to quantify it, as you are trying to do.

BillHoyt did, however, and I think it was extremely clear from his numbers that Clancie was obsessed with him: 100:1 ratio. Good for him (bad for Clancie, though).

You know very well that your method is fundamentally flawed and says nothing about how often people address each other, or even talk about each other. But by quantifying it, you also try to qualify it: By pointing to numbers, you also want to slap a number on "obsession".

So, what is that number, T'ai Chi? What number indicates obsession? BillHoyt has 100:1, which clearly is obsessed. You, OTOH, points to a much lower ratio - so what is that number?

If you cannot answer, then your graph means nothing. As it is, it is a classic example of misleading statistics. But without that number?

Why, one could be led to believe that it is an exercise in propaganda...

[The Don]

T'ai Chi,

I'm very disappointed that you haven't chosen to either acknoledge the content of my last post or answer the question I posed. You have stated your expertise with respect to statistics and statistical analysis and I wanted to get your opinion regarding:

- Whether it's appropriate to base analysis on whatever can actually be measured rather than the correct thing to measure
- Whether my calculation of sample size was about right (after all it was a back-of-the-cigarette-pack caluclation performed in a non-smoking office
- The % of posts required before Claus Larsen is considered obsessive

[T'ai Chi]

Originally posted by CFLarsen

Not really. It is my opinion that Clancie was obsessed, both with me and BillHoyt. I did not try to quantify it, as you are trying to do.


As I've said, there is more to "obsessive"-ness than just post counts, like the content of posts, language used, etc. You're the one who claimed someone was obsessive.

The whole point is is that I asked you to quantify it because you were and still are being incredible vague. You failed to to so, so I have to do your homework for you.


BillHoyt did, however, and I think it was extremely clear from his numbers that Clancie was obsessed with him: 100:1 ratio. Good for him (bad for Clancie, though).


Don't change the topic. We're talking about examining YOUR claim that Clancie is obsessive over you, dear.


So, what is that number, T'ai Chi? What number indicates obsession?


You're the one who claimed obsessive. Why don't you tell us? I also never claimed that one number or any number says something is obsessive, dear. I'm simply counting posts and comparing them to Clancie's count, and noticing that:

1)over 4 months you have many more posts to or about Clancie than she does to or about you.

2) others are noticing that you post to or about her a lot, and they started a long tread about your behavior

3) you cataloging entire fourms, and

4) you posting to or about Clancie on other boards

5) you doing similar things to some other posters, ie telling Lucianarchy that wherever he goes, you'll be there, etc.

That is convergence of evidence of some interesting behavior.


Why, one could be led to believe that it is an exercise in propaganda...

Why don't you show us your evidence for IP address similarity... One could be led to believe that it is an exercise in...

[T'ai Chi]

Originally posted by The Don
whether or not Claus has been guilty of being obsessive about Clancie) on the information that you can get (proportion of posts containing a certain character string) because the information you really need (the number or proportion of posts in which Claus attacks, hassles whatever, Clancie) is not available to you.


From the thread discussions, The Don, do you feel that N(x,y) is a good estimate of N?

And again, post count is just one aspect.


So you have suggested that the post count where Claus refers to Clancie is an acceptable proxy for the information you really need. If this is the case then what is the level which indicates obsession ? 1%, 2%, 5%, 10%, 20% 50% ?

The level of what?

[CFLarsen]

Originally posted by T'ai Chi
The whole point is is that I asked you to quantify it because you were and still are being incredible vague. You failed to to so, so I have to do your homework for you.

But then, you must have decided what that number is. What is it?

Originally posted by T'ai Chi
Don't change the topic. We're talking about examining YOUR claim that Clancie is obsessive over you, dear.

Don't change the topic?? You have been for a long time derailing threads. Hypocrite!

Originally posted by T'ai Chi
You're the one who claimed obsessive. Why don't you tell us? I also never claimed that one number or any number says something is obsessive, dear. I'm simply counting posts and comparing them to Clancie's count, and noticing that:

Excuse me? You count numbers, claim that I have more posts to Clancie that she has to me, and you don't claim that it means anything?

Originally posted by T'ai Chi
1)over 4 months you have many more posts to or about Clancie than she does to or about you.

You are perfectly aware that the count is incredibly flawed. You simply ignore this, and perpetuate this lie of yours.

Originally posted by T'ai Chi
2) others are noticing that you post to or about her a lot, and they started a long tread about your behavior

Indeed. So? Should we also count the number of people who have pointed out that your statistics suck? Does that mean anything?

Originally posted by T'ai Chi
3) you cataloging entire fourms, and

That, I do. Many people have benefitted from that.

Originally posted by T'ai Chi
4) you posting to or about Clancie on other boards

Actually, Clancie and I first "met" on TVTalkshows, and she then followed me here. I was not under the impression that it is forbidden to post on other boards.

Originally posted by T'ai Chi
5) you doing similar things to some other posters, ie telling Lucianarchy that wherever he goes, you'll be there, etc.

Yeah. And they hate it. For obvious reasons...

Originally posted by T'ai Chi
That is convergence of evidence of some interesting behavior.

Perhaps. Why don't you tell us what this behavior is?

[T'ai Chi]

Originally posted by CFLarsen

But then, you must have decided what that number is. What is it?


Wrong, dear.

I am just counting. Why does that mean I need to set some level??


Don't change the topic?? You have been for a long time derailing threads. Hypocrite!


You're a hypocrite, hypocrite. You derail threads, and there is a lot of evidence for that...


Excuse me? You count numbers, claim that I have more posts to Clancie that she has to me, and you don't claim that it means anything?


Claim that you have more posts... LOL, you DO have more posts, dude.


You are perfectly aware that the count is incredibly flawed. You simply ignore this, and perpetuate this lie of yours.


You, or anyone, has failed to show that the estimate is "incredibly flawed". In fact, the only examples you have are artificially constructed ones, constructed for the purpose of showing it was flawed. Too bad for you those examples don't actually happen in reality.

Now tell us, Claud, what do YOU mean when you claimed Clancie was "obsessive"?

Is it any of these things:

1)over 4 months you have many more posts to or about Clancie than she does to or about you.

2) others are noticing that you post to or about her a lot, and they started a long tread about your behavior

3) you cataloging entire fourms, and

4) you posting to or about Clancie on other boards

5) you doing similar things to some other posters, ie telling Lucianarchy that wherever he goes, you'll be there, etc.

??

Let us know..


1)over 4 months you have many more posts to or about Clancie than she does to or about you.

You are perfectly aware that the count is incredibly flawed. You simply ignore this, and perpetuate this lie of yours.

Do you have the balls to say that I "lie" when I claim what I claimed in 1) above?? PLEASE let me know, Claud, that you are in fact saying I am lying when I say you have many more posts to or about Clancie than she does to or about you.

[jj]

Originally posted by T'ai Chi

I am just counting. Why does that mean I need to set some level??


You have, in the past, insisted that your count means something. Please show us the details of the burden you have assumed by suggesting that your counting has meaning. Please show us your validation experiments, your analysis of error and bias issues, and your concise summary of your validation experiments.

Please explain to us how you can claim anything before you establish the validity of your counting.



You, or anyone, has failed to show that the estimate is "incredibly flawed".


I've shown a convincing, telling bias mechanism. You have refused to address it in any substantive fashion, and have instead attempted to shift the burden of proof for your metric to the people who are pointing out the fallacious, quacked nature of it.


In fact, the only examples you have are artificially constructed ones, constructed for the purpose of showing it was flawed. Too bad for you those examples don't actually happen in reality.


Show your evidence, including your clear, incontrovertable evidence of intent and construction. Show that Claus was even aware of the terrible error mechanism I've pointed out when he posted the particular quote I've cited to you.

Show, also, that the other examples of bias someone else has cited to you are "artificially constructed". Again, you're making claims without evidence, and you're still trying to shift the burden of proof that your junk statistics mean anything to the people who have pointed out the flaws.

The burden of proof is completely, absolutely, and irrevocably on you, and not on anyone else. Until you show the error bars, the effects of the bias mechanisms, and how you can control for them, you have no standing to suggest that you have measured anything beyond a mistake, and no reason to conclude anything more than you've exposed the bias mechanisms in your junk statistics. It's not even clear that you have standing to even claim that much.

1)over 4 months you have many more posts to or about Clancie than she does to or about you.


Now, you go from claiming in one thread that "all you're doing is counting", to claiming that your metric is accurate.

Please, before you make any more claims, show that your metric is in fact a valid metric. You've just attempted to make a claim using your metric while still failing to show any validation whatsoever.

You owe us all complete, clear evidence that your metric is VALID, and that you can account fully and systematically for bias mechanisms, and you've provided none of that.



2) others are noticing that you post to or about her a lot, and they started a long tread about your behavior


Argumentum ad populum.


5Do you have the balls to say that I "lie" when I claim what I claimed in 1) above??


He doesn't need to say that. You have to show that your metric measures anything more than bias mechanisms before he should even deign to reply to your junk statistics.


PLEASE let me know, Claud, that you are in fact saying I am lying when I say you have many more posts to or about Clancie than she does to or about you.

You have no reason or authority to make that claim until you present a full evidentary analysis, showing not only your negligently biased measurement, but also your measurement and correction for the egregous bias metric that you won't address substantively.

[CFLarsen]

Originally posted by T'ai Chi
I am just counting. Why does that mean I need to set some level??

But how can you do my "homework" for me, quantifying it, if you haven't set the level?

Originally posted by T'ai Chi
You're a hypocrite, hypocrite. You derail threads, and there is a lot of evidence for that...

Sure, fine. Whatever. Is there plenty of evidence that you derail threads, yes or no?

Originally posted by T'ai Chi
Claim that you have more posts... LOL, you DO have more posts, dude.

Perhaps. But how many? That is what your number says nothing about.

Originally posted by T'ai Chi
You, or anyone, has failed to show that the estimate is "incredibly flawed". In fact, the only examples you have are artificially constructed ones, constructed for the purpose of showing it was flawed. Too bad for you those examples don't actually happen in reality.

What the hell are you talking about? Have you not been shown concrete examples of me replying to a poster mentioning Clancie, but I did not talk about or to Clancie? Have you or have you not?

Originally posted by T'ai Chi
Now tell us, Claud, what do YOU mean when you claimed Clancie was "obsessive"?

She claimed to have me on ignore, yet read my posts.

Originally posted by T'ai Chi
Is it any of these things:

None.

Originally posted by T'ai Chi
Let us know..

I did. Are you going to address my points?

Originally posted by T'ai Chi
Do you have the balls to say that I "lie" when I claim what I claimed in 1) above?? PLEASE let me know, Claud, that you are in fact saying I am lying when I say you have many more posts to or about Clancie than she does to or about you.

Yes, I have the balls: You lie when you claim that your count is not flawed.

[T'ai Chi]

Originally posted by CFLarsen

But how can you do my "homework" for me, quantifying it, if you haven't set the level?


You quantify it by counting it. Where do you get this "set the level" bs? Just where the heck did you dream that up from?


Sure, fine. Whatever. Is there plenty of evidence that you derail threads, yes or no?


Is there evidence that you derail threads, yes or no?


Perhaps. But how many? That is what your number says nothing about.


It is an estimate, as I've repeatedly said, Claud. Moreover, your estimate of N for each month, so far for 3 months straight, is over 4 times the estimate of Clancies'.. remember and that is with counting N(Clancie, Claus) added to Clancie's count.


Have you not been shown concrete examples of me replying to a poster mentioning Clancie, but I did not talk about or to Clancie? Have you or have you not?


You never see their reasoning that N(x,y) isn't close to N however for some reason... Ask them, Clauzzz.

Moreover, as I've already said, these examples have been artifically constructed; been in threads where it was talked about how to 'fool' being picked up by N(x,y). Those types of posts don't seem to occur in reality however. That is, when you respond to Clancie, you have historically mentioned her name in your post. Therefore, the "criticism" is seriously in doubt.


She claimed to have me on ignore, yet read my posts.


As I've mentioned before, one can have a user on their ignore list, and not ignore them. I have you and some others on my ignore list because it makes it easier to see where your posts occur when I scroll through a page.


None.


Really? You don't claim to have done ANY of those things, Claud?


Yes, I have the balls: You lie when you claim that your count is not flawed.

I never claimed it wasn't flawed. N(x,y) is an estimate, dear, it will naturally be off. What I have responded to, and deny, are the claims that it is "seriously" flawed as some have levelled, but have failed to present good evidence for. Their "criticism" amounts to saying census bureau personnell will be off when they carry out a population census.

What is clear, in any case, is that you post more to or about Clancie than she does to or about you.. I really don't care if you don't like that fact. :)

[T'ai Chi]

jj, I'll be happy if you can remember how to use the quote function.


You have, in the past, insisted that your count means something.

It means that Claus posts more to or about Clancie that she does to or about him, as estimated by the forum's search feature. By defintion.

I estimated the trees outside my window. Would you like my estimate of bias, my "validation experiments", and so on? Do you still believe I am doing an experiment when I count?


Please explain to us how you can claim anything before you establish the validity of your counting.


Please explain how Clauzz can claim anything when he hasn't even done a count. Why aren't you getting on his claim, on his case for not doing science, dude?


I've shown a convincing, telling bias mechanism.


This again?

Your "mechanism", only seems to crop up in threads where you or someone else talk about how to 'fool' N(x,y). Do your examples really happen in reality? Has Claus replied to Clancie and then deleted her name from his post? Can you actually find examples to back up your claim of flawed, serious error, negligence, etc etc etc etc etc.

Show me that my estimate is seriously off, as you have claimed, by showing that N(x,y) drastically overestimates N. Until then, you can tell the kind census bureau people that their population counts are serious flawed! Im sure they'd love to hear your comments, just as much as I do.


The burden of proof is completely, absolutely, and irrevocably on you, and not on anyone else.


You made a claim of serious error; have the balls to show that it is serious as you claimed. I already know that it is an estimate, dear.


You have no reason or authority to make that claim until you present a full evidentary analysis, showing not only your negligently biased measurement, but also your measurement and correction for the egregous bias metric that you won't address substantively.


Nope.

It is clear that Claus has many more posts to or about Clancie than she does to or about him as gauged by N(x,y).

[CFLarsen]

Originally posted by T'ai Chi
It means that Claus posts more to or about Clancie that she does to or about him, as estimated by the forum's search feature. By defintion.

No, quite contrary. It means that the method you use inflates the count, by counting other people's mentions of Clancie as well.

We have seen examples of this. You ignore those, and continue to lie. Do go on, liar.

Originally posted by T'ai Chi
It is clear that Claus has many more posts to or about Clancie than she does to or about him as gauged by N(x,y).

Do go on, liar.

[T'ai Chi]

Originally posted by CFLarsen


No, quite contrary. It means that the method you use inflates the count, by counting other people's mentions of Clancie as well.

We have seen examples of this. You ignore those, and continue to lie. Do go on, liar.

Do go on, liar.

Ahh, your repetion tactic. :rolleyes:

Think about it, Claud. If you responding to someone who mentioned Clancie to talk about Clancie, you are talking about Clancie, which is what N(x,y) counts, by definition.

Can you show any posts that aren't constructed just to beat N(x,y), where you respond to someone who talked about Clancie but you yourself don't talk about Clancie in that post?

Let us know, dear.

[CFLarsen]

Originally posted by T'ai Chi
Think about it, Claud. If you responding to someone who mentioned Clancie to talk about Clancie, you are talking about Clancie, which is what N(x,y) counts, by definition.

Can you show any posts that aren't constructed just to beat N(x,y), where you respond to someone who talked about Clancie but you yourself don't talk about Clancie in that post?

Let us know, dear.

When I find them, then what are you going to do? Admit that you are wrong?

[T'ai Chi]

Originally posted by CFLarsen

When I find them, then what are you going to do?

Just actually find them first, please. You claim they exist outside of threads where it is brought up how to 'beat' N(x,y), so YOU find them.

Then if/when you do, I'd like to see how many you find, say M of them. Then ask yourself, is M so large that N(x,y) is a poor estimate of N, or does the size of M not matter, since it is obvious that you post more to or about Clancie anyway..

I mean, you'll have to try and convince me that N(Claus, Clancie)-M is near N(Clancie, Claus) + N(Clancie, CFlarsen), which I am STRONGLY skeptical of. I'm willing to see your evidence however..

[CFLarsen]

Originally posted by T'ai Chi
Just actually find them first, please. You claim they exist outside of threads where it is brought up how to 'beat' N(x,y), so YOU find them.

Here's ten, found in minutes:

Example 1 (http://www.randi.org/vbulletin/showthread.php?s=&postid=1869976328#post1869976328)
Example 2 (http://www.randi.org/vbulletin/showthread.php?s=&postid=1869993457#post1869993457)
Example 3 (http://www.randi.org/vbulletin/showthread.php?s=&postid=1870017352#post1870017352)
Example 4 (http://www.randi.org/vbulletin/showthread.php?s=&postid=1870050006#post1870050006)
Example 5 (http://www.randi.org/vbulletin/showthread.php?s=&postid=1870062089#post1870062089)
Example 6 (http://www.randi.org/vbulletin/showthread.php?s=&postid=1870068357#post1870068357)
Example 7 (http://www.randi.org/vbulletin/showthread.php?s=&postid=1870074313#post1870074313)
Example 8 (http://www.randi.org/vbulletin/showthread.php?s=&postid=1870144226#post1870144226)
Example 9 (http://www.randi.org/vbulletin/showthread.php?s=&postid=1870157084#post1870157084)
Example 10 (http://www.randi.org/vbulletin/showthread.php?s=&postid=1870157479#post1870157479)

Originally posted by T'ai Chi
Then if/when you do, I'd like to see how many you find, say M of them. Then ask yourself, is M so large that N(x,y) is a poor estimate of N, or does the size of M not matter, since it is obvious that you post more to or about Clancie anyway..

That is for you to determine, as well as how much more.

Originally posted by T'ai Chi
I mean, you'll have to try and convince me that N(Claus, Clancie)-M is near N(Clancie, Claus) + N(Clancie, CFlarsen), which I am STRONGLY skeptical of. I'm willing to see your evidence however..

I am not here to convince you. You have been shown evidence that your count is deeply flawed. The rest is up to you.

[T'ai Chi]

Originally posted by CFLarsen


Here's ten, found in minutes:

Example 1 (http://www.randi.org/vbulletin/showthread.php?s=&postid=1869976328#post1869976328)
Example 2 (http://www.randi.org/vbulletin/showthread.php?s=&postid=1869993457#post1869993457)
Example 3 (http://www.randi.org/vbulletin/showthread.php?s=&postid=1870017352#post1870017352)
Example 4 (http://www.randi.org/vbulletin/showthread.php?s=&postid=1870050006#post1870050006)
Example 5 (http://www.randi.org/vbulletin/showthread.php?s=&postid=1870062089#post1870062089)
Example 6 (http://www.randi.org/vbulletin/showthread.php?s=&postid=1870068357#post1870068357)
Example 7 (http://www.randi.org/vbulletin/showthread.php?s=&postid=1870074313#post1870074313)
Example 8 (http://www.randi.org/vbulletin/showthread.php?s=&postid=1870144226#post1870144226)
Example 9 (http://www.randi.org/vbulletin/showthread.php?s=&postid=1870157084#post1870157084)
Example 10 (http://www.randi.org/vbulletin/showthread.php?s=&postid=1870157479#post1870157479)



Some of those are valid. What is interesting, is for a lot of them, the post immediately before or after the link above you are talking to or about Clancie..


That is for you to determine, as well as how much more.


NO, dear, that is for YOU to determine, since you are making the claim of serious error, fundamentally flawed, etc. I am saying it is an estimate. Being off by some is fine. The point is, that when compared to Clancie's, yours is much higher, dear.

Your homework, to demonstrate evidence for your claim, is to try and convince me that N(Claus, Clancie)-M is near N(Clancie, Claus) + N(Clancie, CFlarsen), which I am STRONGLY skeptical of. That is what you are trying to convince me of here.


I am not here to convince you. You have been shown evidence that your count is deeply flawed. The rest is up to you.

Just provide evidence for your claim, dear, not evidence that N(x,y) is an estimate, which is something we already know.

[CFLarsen]

Originally posted by T'ai Chi
Some of those are valid.

Please point out those that are not valid.

Originally posted by T'ai Chi
What is interesting, is for a lot of them, the post immediately before or after the link above you are talking to or about Clancie..

That is completely beside the point. Nice attempt at changing the focus, though...

Originally posted by T'ai Chi
NO, dear, that is for YOU to determine, since you are making the claim of serious error, fundamentally flawed, etc. I am saying it is an estimate. Being off by some is fine. The point is, that when compared to Clancie's, yours is much higher, dear.

How much? Is it one or a gazillion? You don't know, do you?

Originally posted by T'ai Chi
Your homework, to demonstrate evidence for your claim, is to try and convince me that N(Claus, Clancie)-M is near N(Clancie, Claus) + N(Clancie, CFlarsen), which I am STRONGLY skeptical of. That is what you are trying to convince me of here.

Again, I am not trying to convince you. Your count is severely flawed. If you want to point to a number that is flawed, go ahead.

Originally posted by T'ai Chi
Just provide evidence for your claim, dear, not evidence that N(x,y) is an estimate, which is something we already know.

Huh?

[jj]

Originally posted by T'ai Chi

Just provide evidence for your claim, dear, not evidence that N(x,y) is an estimate, which is something we already know.
You still owe us all a full and complete validation for your "metric", one that accounts for all known bias mechanisms (including this one) and that shows their influence, convincingly, as supported by measured, verifiable, confirmable evidence.

You have done none of that.

Until you do so, every reference implying substance to or about your metric is deceit.

[T'ai Chi]

Originally posted by jj

Please explain to us how you can claim anything before you establish the validity of your counting.


Please see: http://www.randi.org/vbulletin/showthread.php?s=&postid=1870509382#post1870509382.


aware of the terrible error mechanism I've pointed out when he posted the particular quote I've cited to you.


Basically 8 misidentified out of 100 is "terrible"? Shirley you must be joking.

[CFLarsen]

T'ai Chi,

Please point out those of my references that are not valid.

[T'ai Chi]

See http://www.randi.org/vbulletin/showthread.php?s=&postid=1870509382#post1870509382.

Claus, as I've explained, estimates are naturally "off". The question is, how far off is it? Being off by 80 out of 1000 ain't bad. :)

[CFLarsen]

That is not what it's about. I presented 10 references, where you count me talking to/about Clancie, where I clearly don't.

You claimed that some were invalid. Please point out those, and explain why.

[T'ai Chi]

I retract my comment. All 10 of your examples are valid.

I will repeat, however, that for a lot of them, the post immediately before or after the linked post you are talking to or about Clancie.

Would you like to address the 8% error? In other words, for those 10 you presented, there are many more where you do talk to or about Clancie.

[jj]

Originally posted by T'ai Chi
I retract my comment. All 10 of your examples are valid.

I will repeat, however, that for a lot of them, the post immediately before or after the linked post you are talking to or about Clancie.

Would you like to address the 8% error? In other words, for those 10 you presented, there are many more where you do talk to or about Clancie.

And that's outright fraud. You count ONLY the examples he shows you as the error, you have done no work to show that there are no more obvious bias errors.

Do your work, or stop claiming your quackery means anything.

[CFLarsen]

Originally posted by T'ai Chi
I retract my comment. All 10 of your examples are valid.

Thank you. You obviously did not even bother to look at them.

Originally posted by T'ai Chi
I will repeat, however, that for a lot of them, the post immediately before or after the linked post you are talking to or about Clancie.

Red herring. Irrelevant.

Originally posted by T'ai Chi
Would you like to address the 8% error? In other words, for those 10 you presented, there are many more where you do talk to or about Clancie.

Sorry, you have just in another thread thrown in a condition like "meaningful". Address that first.

[T'ai Chi]

Originally posted by jj

And that's outright fraud. You count ONLY the examples he shows you as the error, you have done no work to show that there are no more obvious bias errors.


See my sig file.

8% error apparently.

Since we are counting approaching 2000 posts, that ain't bad.

[T'ai Chi]

Originally posted by CFLarsen


Thank you. You obviously did not even bother to look at them.


Oh, I read them. I probably just thought a couple of them were talking about Clancie when they weren't. No biggie.


Red herring. Irrelevant.


Very relevant, considering that is why N(CFlarsen, Clancie) is sooo large...

[CFLarsen]

Originally posted by T'ai Chi
Oh, I read them. I probably just thought a couple of them were talking about Clancie when they weren't. No biggie.
"No biggie"? You try to dismiss the evidence, but can't be bothered to look at them?

Fool.

[CFLarsen]

Bump.

[Complexity]

Let me add that I know the answer and am just taking a poll for those on this board who call things I've counted 'experiments'.
No, T'ai Chi - You think you know the answer.

The real question is

Why should anyone care about what you post?

[Complexity]

T'ai Chi seems proud of his post count.

If I were him, I'd be concerned about his signal / noise ratio.

Some of us aim for quality; others have to settle for quantity.

[Soapy Sam]

Can't see why anyone would revive this old corpse, but since they have, I feel the following are relevant.

1. In general, you should not put fruit in the fridge.
2. "Criteria" is a plural. The singular is "criterion".

Thank you.

[Morwen]

Darn! And I kinda wanted to see Tai's answer to The Don's very nice post about what does constitute a positive result and a good counting methodology...

[CFLarsen]

Darn! And I kinda wanted to see Tai's answer to The Don's very nice post about what does constitute a positive result and a good counting methodology...
After almost two years of evading, why do you think T'ai Chi will answer it now?