probability estimates

[gsgs]

OK, here is my probability-estimates thread at flutrackers !


I always considered and still do consider this the main question
in the whole context, yet never saw this asked by others :
(strange !)

=================================================
what's the expert's estimates on the probability of a severe
pandemic with >100million H5N1-deaths in the next (-say-) 5 years
=================================================

are there experts out there, who would answer this question ??
I was not very successful to find some in the past.
See my threads "probability estimate"
at fluwiki http://www.fluwikie.com/index.php?n=Forum.ProbabilityEstimate
and this webpage: http://www.psandman.com/gst2005.htm#guenter

what do you think, will be the average answer experts will give to
the above question ? (_if_ they answer)

why do experts usually refuse to answer this ?
do you think they would answer, if they could stay anonymous ?

another question which I'm even more interested in:

=================================================
what's your expectation value of the number of H5N1-deaths
in the next 5 years ?
=================================================


I'm looking for a list of the most qualified experts who are also
likely to answer one of these questions.
And a referrence, if someone ever has seen an expert
answering this or a similar question.

 

[Toaster2]

Re: probability estimates

Gsgs I am afraid that by definition, none of the experts will give you a (gu)es(s)timate.

 

[gsgs]

Re: probability estimates

> Gsgs I am afraid that by definition, none of the experts will not
>give you (gu)es(s)timate.

what do you mean with "by definition" ?

Yes, replace "estimate" with "guesstimate" it you prefer.
Do you think, I should use "guesstimate" instead ? It's not
in my dictionary.

And why won't they answer "to me" ? Would they more likely answer to
someone else ? Someone whom I can hire for that task ?

The double negation, I assume it is a mistake and not a
positive statement.

 

[Toaster2]

Re: probability estimates

gsgs, the not was indeed a typo, I corrected that. What I meant is that the experts will not make guesses as to how many people will die on an internet forum (regardless of who is asking). That is because in order for them to be taken seriously by their community, they should stick to the facts. And the fact is that no-one really knows - it all depends on many factors as you know.

Any expert that will come out and say: I guess there will be 1 billion dead will be ridiculed. Any expert who says, only 3 million will also be ridiculed. So basically it is smarter not to give a quote, as it most likely is a lose-lose situation for a scientist to make a (gu)es(s)timate on a sensitive issue like this.

> Gsgs I am afraid that by definition, none of the experts will not
>give you (gu)es(s)timate.

what do you mean with "by definition" ?

Yes, replace "estimate" with "guesstimate" it you prefer.
Do you think, I should use "guesstimate" instead ? It's not
in my dictionary.

And why won't they answer "to me" ? Would they more likely answer to
someone else ? Someone whom I can hire for that task ?

The double negation, I assume it is a mistake and not a
positive statement.

 

[gsgs]

Re: probability estimates

I meant is that the experts will not make guesses as to how many people will die on an internet forum (regardless of who is asking). That is because in order for them to be taken seriously by their community, they should stick to the facts. And the fact is that no-one really knows - it all depends on many factors as you know.

Any expert that will come out and say: I guess there will be 1 billion dead will be ridiculed. Any expert who says, only 3 million will also be ridiculed. So basically it is smarter not to give a quote, as it most likely is a lose-lose situation for a scientist to make a (gu)es(s)timate on a sensitive issue like this.

that doesn't make sense. Imagine a superexpert who has all available
data and very intensive studies and his guesstimate is SE.
Other experts who ridicule about SE were clearly wrong.

So, ridiculing others' estimates is as "dangerous" as giving an own
guesstimate when you don't know what SE really is.

And also, this could be prevented by allowing the experts to stay
anonymous.

Not giving probability estimates is just being
unprecise/noninformative/ambiguous deliberately.

We should insist on these estimates. That's what we pay the experts for.
You won't accept them withholding all information about their
research but you do accept them hiding their conclusions
in ambiguous statements instead of giving guesstimates ?

 

[Toaster2]

Re: probability estimates

gsgs, please note, I am not trying to discourage you. But we all know the possible scenarios - these numbers have been going around for quite awhile.

The options are:

1) no pandemic will occur
2) only a very mild version will achieve efficient H2H (result +/- 3 million deaths)
3) an intermediate version will go H2H (result 3-30 million)
4) a severe as in 1918 will do it (30-180 million)
5) the indonesion goes H2H witouht losing its death rate (180 million-1 billion or more)

Everyone knows these options, and as soon as a pandemic H5N1 arises one can start making predictions because by then the most likely scenario is known. Until then it is just a person's gut feeling that determines what he thinks is likely or not.

Although it would be interesting to know what people's gut feelings are I just do not think that people working in the field of influenza will step forward and present us with their personal opinion - which as they are experts may cause panic and may have an adverse effect on their credibility if their guess is way off.

You are right to say ridiculing such estimates would be wrong - I should have written that these people would lose their credibility rather than be ridiculed.

that doesn't make sense. Imagine a superexpert who has all available
data and very intensive studies and his guesstimate is SE.
Other experts who ridicule about SE were clearly wrong.

So, ridiculing others' estimates is as "dangerous" as giving an own
guesstimate when you don't know what SE really is.

And also, this could be prevented by allowing the experts to stay
anonymous.

Not giving probability estimates is just being
unprecise/noninformative/ambiguous deliberately.

We should insist on these estimates. That's what we pay the experts for.
You won't accept them withholding all information about their
research but you do accept them hiding their conclusions
in ambiguous statements instead of giving guesstimates ?

 

[gsgs]

Re: probability estimates

>gsgs, please note, I am not trying to discourage you.

thanks for your feedback.

>But we all know the possible scenarios - these numbers
>have been going around for quite awhile.
>
>The options are:
>
>1) no pandemic will occur
>2) only a very mild version will achieve efficient
> H2H (result +/- 3 million deaths)
>3) an intermediate version will go H2H (result 3-30 million)
>4) a severe as in 1918 will do it (30-180 million)
>5) the indonesion goes H2H witouht losing its death rate
> (180 million-1 billion or more)

splitting it into 5 or more or less scenarios doesn't change anything.

>Everyone knows these options, and as soon as a pandemic H5N1 arises
>one can start making predictions because by then the most likely
>scenario is known. Until then it is just a person's gut feeling
>that determines what he thinks is likely or not.

...but using the data in the available papers and own analysis on
the likelyhood of virus-mutations etc.
Clearly the experts should have a better "gut feeling" here
than the other people.

>Although it would be interesting to know what people's
>gut feelings are I just do not think that people working
>in the field of influenza will step forward and present
>us with their personal opinion - which as they are experts
>may cause panic and may have an adverse effect on their
>credibility if their guess is way off.

or may have a positive effect on their credibility if it isn't.
Note, that such estimates are already being done, just that
they are more or less inexact, are obfuscated or kept ambiguous.
No numbers are given to make it precise, but the experts do
give interviews - and are frequently misunderstood.

>You are right to say ridiculing such estimates would be wrong
> - I should have written that these people would lose their
>credibility rather than be ridiculed.


would you recomment to replace "guesstimate" by "gut feeling"
in the question now ?

 

[Toaster2]

Re: probability estimates

I hpe there will be reactions - the info would be nice - but still I think the experts won't react.

 

[gsgs]

Re: probability estimates

see here, how easy it is to get estimates
http://drbobgleeson.typepad.com/bird_flu/2006/02/your_h5n1_predi.html

I must have been doing something wrong all the years.

 

[gsgs]

Re: probability estimates

time to bump this.
time to ask experts for probability estimates when there
are not so much news to talk about.






"numbers,statistics,probabilities" is not just a passion or kink of me,
in fact it's exact science, generally accepted and supported by experts who
wrote book s about it also a blog entry by Michael Steele

Don't get distracted by the fact that you meet other gsgs's here rarely.

Neustadt,May identified the lack of probability-estimates discussion
as the major mistake in the swine flu-non-pandemic aftermath.
The same could happen in the H5N1-panflu aftermath.

The problem is just that medical people don't want to realize it.
I guess, those Osterholms,Websters,Gregers have no education in
statistics,probability theory.(the reveres have, but still won't give numbers !)
They can't so well give numbers so they claim
it's impossible. But those actuaries _can_ assign numbers to the panflu-threat
and they do it and this is all what we and the politicians and economists have,
since these are not being challenged by medical experts, who retract from
interdisciplinary discussion about numbers.
They have their own agenda for getting funding and politicians might reduce it to that,
while not taking the claims very serious.

The whole organisation of medical scientice is not so supportative
for interdisciplinary truth. Influenced by big pharma,economics,University-rivalry
resulting in secret databases, programs without code, papers being published
only years after the outbreaks, contradicting researchers, no public expert forums or mailing lists.
Not what I'm used to see within mathematics.


See here for a collection of statements and polls about estimating the panflu-thread:
http://www.setbb.com/fluwiki2/viewf...d0f10da2f14acbedf5c3ae59343aa&mforum=fluwiki2

 

[Jonesie]

Re: probability estimates

would you recomment to replace "guesstimate" by "gut feeling"
in the question now ?

gsgs ... Since you are having so much difficulty with unclear words and phrases, why don't you take your questions to the WordReference Forums, where they just might have the patience and expertise to help you out:

http://forum.wordreference.com/showthread.php?t=162476

They do have a sub-forum: German For questions about German, or translations between German and any other language:

http://forum.wordreference.com/forumdisplay.php?f=18


And they do discuss pandemics:

http://forum.wordreference.com/search.php?searchid=2481846

:magnify:

 

[gsgs]

Re: probability estimates

OK, it's here:
http://forum.wordreference.com/showthread.php?p=2782041
let's see whether there will be some feedback

 

[Treyfish]

Re: probability estimates

I have a gut feeling,that a pandemic is extremely likely.Most scientist estimate it is certainly probable.It is impossible to accuratly predict the start of a pandemic.The worst guessimate anyone can make, would be to assume it was unlikely.Maybe i'm wrong ,but from news i read,that is my guess!