I dunno, you started with the name calling and changed the topic of discussion.

Thanks for the permission to continue!

Not the lower bound of the 95% CI

I'm happy to discuss, but based on the giant, yelling, font I imagine it'll be the usual routine.

I posted an article that evaluated viral escape.

The authors reported a point estimate (of hazard ratio) for protection afforded by previous infection: "0.81 [95%CI: 0.73-1.00] (i.e. remaining protection of 19% [95%CI: 0-27%]) against Omicron."

The authors also reported the same metrics for 4 different vaccination conditions:

"AZ 2+weeks post-dose 2 (PD2) 1.86 (95%CI: 1.67-2.08) ,

Pfizer 2+w PD2 2.68 (95%CI: 2.54-2.83),

AZ 2+w post-dose 3 (PD3) 4.32 (95%CI: 3.84-4.85)

PF 2+w PD3 4.07 (95%CI: 3.66-4.51) PD3 states are categorised by the dose 1/2 vaccine used).

These estimates translate into Omicron VE estimates of between 0% and 20% PD2 and between 55% and 80% PD3 against Omicron"

You took issue with the article because they authors did not provide a single point-estimate for protection afforded by the 4 discrete vaccination conditions.

After much trial and tribulation, you asserted that the point estimate for overall/pooled vaccination should be 0% because that is the lowest bound of the estimates. I asked if you were asserting that 0% should also be the estimate for previous infection and you agreed.

By your logic, the reported estimate of 19% from previous infection is too high and should be 0%. The lowest boundary of the condition with the least protection should be reported as a point estimate for vaccines across 4 discrete conditions. With this you agreed.

Thus, your argument is that protection from omicron infection is equivalent between previous infection and any of the 4 discrete vaccination condition - 0%.

My response was that this logic seems flawed for two reasons.

The first is in a common misunderstanding of what frequentist confidence intervals represent.

A frequentist 95% confidence interval means that with a large number of repeated samples, 95% of such calculated confidence intervals would include the true value of the parameter. In frequentist terms, the parameter is fixed (cannot be considered to have a distribution of possible values) and the confidence interval is random (as it depends on the random sample).

Second, you're assuming the lower bound of the CI from pooled estimates would be equal to the lowest CI boundary of the unpooled estimates. That's unlikely to be the case as there would be regression toward the pooled mean.

Then we went digressed into the usual name calling, at least this time we had a new technique - the giant fonts.

Baling on the discussion. While expected, it's still disappointing.

Scroll.

You asserted the lower bound should be reported as the point estimate.

Sorry man, I thought you might be a fun sparring partner if I could ever get you to engage in a discussion.

I am disappoint.

After seeing how you react when your argument falls apart, I can understand why you generally stick to trolling.

This is your original post in this thread.Nothing to do with the topic. It looks like trolling.

I'll do this in parts to get past the server errors:

RepostingThen the conversation degrades in the usual way:

Is it not obvious that's my post? My name is at the top of the quote. I had to break the response up due to server errors. Posts require text in addition to quotes.

Regardless, it doesn't appear we'll be discussing anything.

It does appear you enjoy trolling me, so have at it.

Yesterday I knew 9 people who die of comicon.

Hey OP, suck a long fat dink ok? Ph4ggot

I guess all arguments were decimated! A sign of a great debate. Let's do it again some time

Long AND fat?!

If your argument isn't clear, that suggests re-framing might be necessary. "Guess my argument" is an odd approach to debate, but a reliable approach to trolling.

A 0% estimate is not supported by this paper.

You never state a position, you try to coerce people into guessing your position by asking vague questions. Position statements don't end with a question mark nor do they take the form of a demand.

When people don't follow your orders, you start yelling and name calling, now in giant colourful text. Funny that you keep mentioning pre-school. Tantrums are funny.

Where?

You yell. You call names. But never really say anything. Tantrum style

There we go. A non answer.

Up next, bigger fonts and bigger tantrums.

Really?

You're gonna go with "I know you are, but what am I?"

Son, I am disappoint.

I wonder what variant we will be dealing with a year from now that will be "100 times worse then the omicron variant"

Nothing like flipping the world upside down for a disease with a 99.97% survival rate.

Honk honk, as they say.

Eat my ass OP

“Just as deadly” they said

Did they say "deadly"? That word doesn't appear with a search in the article.