Moderation Bin

I don’t like the idea banning posters or deleting comment. I am loathe to do either. As such, I am going to try a new moderation option. From here on in, users who cannot follow basic guidelines will be forbidden from writing in any topic save ones I explicitly mark as bins for them.

If they wish to participate in other threads, they are allowed only to make comments with a single link to a comment they’ve made in such a moderation bin. This will allow them to still comment while not forcing readers to see what they say.

Advertisements

10 comments

  1. Kevin O’Neill made a comment in another topic despite having been told to he was under moderation. I’m fixing his mistake. His comment was:

    Carrick – I asked in # 6) Are you claiming that McIntyre’s code only produced upward-sloping ‘hockey sticks’?

    The answer is, “No, it produced both, but the archive only had upward-sloping HS”

    The whole devolution here is because both you and Brandon claimed that fact was insignificant. I said it underscored the fact Wegman didn’t do due diligence. Even the least curious of us would be intrigued by a coin that always came up heads.

    The whole how they’re positioned argument is a red herring. Wegman didn’t flip any results. We know that because we see the code. We know that every result in Fig. 4.4 comes from the archive. There were no negative results to flip.

    The question I asked earlier still remains. Wegman saw 12 upward sloping HS. What are the odds of that?

  2. Kevin O’Neill chose to disregard being under moderation, so I’ve deleted a couple of his comments on another thread. This comment is a note providing some information.

    Kevin O’Neill, I’ve told you you are under moderation. My moderation rules are simple, and you can get off moderation by addressing the issue I told you you are required to address.

    Until such a time, however, you know you are not allowed to write comments such as those you posted here. If you wish you contribute to topics, you may, but you must post in the manner you’ve been instructed to. In case there is any question about the process by which you may still comment, here is a link to the explanation again.

    I can’t prevent you from complaining my moderation is unfair, but for any audience there may be, let me make something simple. You can participate without any risk of your comments being deleted. You can do so while saying anything you want. All you have to do is follow the rules about where you place your comments. This is not censorship. It is not abusing my moderation powers. It’s simply enforcing a rule you are well aware of.

    For those who are curious about this rule, it is a rule at this site when challenged on any factual point, one must fully address the challenge. A refusal to do so will get you placed in moderation. I informed O’Neill of his situation in this comment. O’Neill claimed for “the purposes of reproduction/replication,” AR and ARFIMA noise models are “completely different.” That’s a bold (and ridiculous) statement which he currently refuses to even attempt to justify. That’s why he’s under moderation.

  3. Brandon, I already told you why the two models are completely different. i realize now that you are a bad faith actor. But just in case I wasn’t clear I will make the attempt one more time.

    In my world and line of work an AR1(.2) noise model is *completely different* from an AR1(.9) noise model – and that’s what we’re comparing. The implied memory is physically plausible in many instances for a lag coefficient of 0.2, but for many processes an implied memory of 19 years is completely ridiculous.

    Now, perhaps you’ve never done real-world failure analysis or root cause analysis. Trying to sift the wheat from the chaff – or more likely assigning probabilities to various types of chaff – is challenging enough without adding confounding factors like inappropriate statistical models. If I ever tried to claim a persistence of 19 years in my line of work I’d be laughed out of the room.

  4. Kevin O’Neill, you can consider (or “realize” about) me whatever you want, but what you say is not just wrong, it misrepresents what is being discussed. You say:

    In my world and line of work an AR1(.2) noise model is *completely different* from an AR1(.9) noise model – and that’s what we’re comparing.

    But this is not true. We were not comparing an AR1(.2) noise model to an AR1(.9) noise model We were comparing AR and ARFIMA models. That one can estimate a roughly equivlanet AR1 parameter for a given ARFIMA model does not mean one can simply ignore the fact we are examining an ARFIMA model. It certainly is not okay to do so after criticizing the Wegman Report for mixing up the two noise models.

    Moreover, nobody with the slightest sense would argue two AR models with different parameters are “completely different.” All AR models have many similarities given by the fact they are created via the same process. That one parameter of a model is tweaked between runs in no way justifies saying the models are “completely different.”

    In case the point isn’t obvious enough to someone, it’s easily demonstrated by asking, “How high can you get an AR coefficient before it is ‘completely different’ from AR(.2)?”

  5. Brandon, as the NRC/NAS study showed, they had to use AR1(.9) to achieve the result M&M got with their ‘persistent red noise’ model or that Wegman got with his AR1(.2). So we’re comparing AR1(.2) to AR1(.9) The implied persistence is 1.5 years versus 19 years. In my line of work physical plausibility is necessary to ‘explain’ anything – otherwise I might as well attribute it to fairies, gnomes or gremlins. Well, assigning problems to gremlins is actually pretty common – we just word it differently 🙂

    You choose to believe that because both models are AR1 or red noise in general that they can’t be completely different in the effects or results. You would have a hard time finding a believable result with that attitude in my field.

    It’s a common misperception that what I *know* to be true in *my* field must be true in all other fields. Generally this is a just a lack of exposure to other fields and the methods they’ve developed. The most common form of this I’ve seen is the sampling frequency relationship; most electrical/electronic engineers have a dogmatic belief in the Nyquist Theorem. They’ve never learned Time Series Analysis where uneven sampling negates many of the limitations imposed by the Nyquist Theorem.

    In any event – the AR1(.2) versus AR1(.9) relationship vis a vis what Wegman stated and what the results actually mimicked has been known for years. That you pretend(?) ignorance of this fact makes you either far less knowledgeable about the matter than you like to present – or you are a bad faith actor. I tend to believe the latter.

  6. Kevin O’Neill, given you were placed on moderation for making untrue factual claims, it would behoove you not to make further obviously untrue factual claims like:

    Brandon, as the NRC/NAS study showed, they had to use AR1(.9) to achieve the result M&M got with their ‘persistent red noise’ model or that Wegman got with his AR1(.2). So we’re comparing AR1(.2) to AR1(.9) The implied persistence is 1.5 years versus 19 years.

    This is complete BS as anyone who reads our conversation knows. Leaving aside the fact you are simply making things up when you say:

    the NRC/NAS study showed, they had to use AR1(.9) to achieve the result M&M got

    We were not comparing what you claim. We were specifically discussing the use of an ARFIMA noise model as opposed to an AR model. You discussed how one could try to translate an aspect of an ARFIMA model into that of an AR model, but that doesn’t change the fact we were comparing ARFIMA vs. AR.

    On the subject of making things up:

    You choose to believe that because both models are AR1 or red noise in general that they can’t be completely different in the effects or results. You would have a hard time finding a believable result with that attitude in my field.

    I have never expressed any such view. You have no basis for claiming I believe such. I don’t believe what you claim I “choose to believe.”

    But in the end, it all comes back to a simple thing. You claimed AR and ARFIMA models are “completely different,” and you refuse to address that issue. Your discussion of how different parameters within a model does nothing to address the point at hand. Even worse, if we accept your change of subject from ARFIMA vs AR to AR(.2) vs AR(.9), your claim makes no sense. Two noise models cannot be “completely different” simply because you fed a different value into them for a particular parameter.

    Address the point which has gotten you placed in moderation or not, but I’m not going to continue responding to your red herrings.

  7. Kevin O’Neill, your response is stupid. That the NAS used a particular parameter does not mean they showed you must use that parameter to get particular results. The only way they could have showed one must use that parameter is if they tried lower AR values and found they didn’t get similar results. They did not.

    Beyond which, my moderation policy does not require a person be put on moderation for being wrong. They only get placed in moderation when they, like you, refuse to even attempt to address the wrong things they say. I’ll note at this point you aren’t pretending to address the issue which got you placed in moderation.

    Unless you go back to addressing that issue, I’m not going to bother responding to the stupid things you say.

  8. —————-
    This is complete BS as anyone who reads our conversation knows. Leaving aside the fact you are simply making things up when you say:

    “the NRC/NAS study showed, they had to use AR1(.9) to achieve the result M&M got”
    —————–

    You claimed I was making things up. I gave you the proof. Again, you show you are a bad faith actor.

  9. Kevin O’Neill,even Nick Stokes disagrees with your current argument when you made it at his site. He made the same point I made, that showing results for one parameter value does not indicate that parameter value is the only one which would produce such results.

    You can call me a bad faith actor all you want, but it’s so obvious you’re full of it not even Stokes will support you on this point.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s