TCO,

The Ritalin comment was not so much targetting the AR1 coefficient estimation, but the fact that Ritson was demanding a response from Wegman within around four weeks or so when the hockey team were guilty of dragging out responses much longer than that.

The AR1 coefficient issue was covered in more depth here. The problem Steve identified with the Ritson method compared to the conventional method was that in the event of model misspecification (ARMA(1,1) vs. AR1) the Ritson method gave very wrong results, whereas the conventional test is more robust. The inconsistency between the Ritson method and the conventional method tie in almost exactly with this type of model misspecification.

Steve has talked about the danger of conditional defenses. Ross’s comment was a conditional defense!

In addition, Steve’s post talked about “I don’t concede it for a minute”. Well, does he contest it? Which is right and why?

In addition, Steve has a bunch of (correct) comments about “if we make the change, the effect still remains but at a lower level”. So fine. That is a conditional defense also. If Ritson is right, admit it. If not, show why he’s wrong. If you’re unsure if he is right or not, admit that (and discuss the issue itself).

On Ross’s statement, yes, I can read that in. But, as stated, it is a concern, not an accusation.

I read his post and thought he was employing a standard and often used phraseology to make his point. To be able to read something nefarious into such common language (and you do this often with Steve M) is a form of innuendo, and vague at that, that serves no useful purpose in furthering the discussion. Part of my problem with your critiques is that they are not sharp and clear, to me at least, and so if I am looking for criticism of Steve’s and Ross’ work you would not be my first choice.

Group think is to be avoided and, if that is what you are about then I say you have good intentions, but that is best accomplished in my view by clear and thoughtful questions without a lot of inferences attached.

re: #23

P.s. I bet if Wegman were here, he would back me up. And Martin Ringo who knows vector algebra out the ying yang, has been very kind and responsive to my questions, to places where I have pushed Steve for logical defense or for numerical description.

It is statements like the above that make me perceive that you are very much full of yourself and this seems to guide your responses at this blog. I often hear from you that you pushed Steve M to do this or that and that TCO is seldom wrong. Wegman would back you up on what — not appealling to authority? What are these places specifically where you have pushed Steve for logical defense or for numerical description? And how did all of this add to or influence the discussion? If you could be more specific and clear in your responses perhaps I could take your efforts more seriously.