Re: The disastrous ratings decline continues

[The Doctor]

In article <slrnkmrgu...@pjr.no-ip.org>,
Peter J Ross <peadar...@gmx.com> wrote:
>Series 7A, Episode 3 ("A Town Called Mercy") -
>
>6.6 million viewers
>
>
>
>Series 7B, Episode 3 ("A Disgusting Insult to the Memory of Brian
>Hayles and Bernard Bresslaw") -
>
>5.7 million viewers
>
>
>
>Surely cancellation can't be far away.
>
>
>--
>PJR :-)
>

Only in your riots PEter.
--
Member - Liberal International This is doc...@nl2k.ab.ca Ici doc...@nl2k.ab.ca
God,Queen and country!Never Satan President Republic!Beware AntiChrist rising!
http://www.fullyfollow.me/rootnl2k Look at Psalms 14 amnd 53 on Atheism
B.C. do not condemn your province - vote Liberal!

[Your Name]

In article <slrnkmrgu...@pjr.no-ip.org>, Peter J Ross
<peadar...@gmx.com> wrote:
>
> Series 7A, Episode 3 ("A Town Called Mercy") -
>
> 6.6 million viewers
>
>
>
> Series 7B, Episode 3 ("A Disgusting Insult to the Memory of Brian
> Hayles and Bernard Bresslaw") -
>
> 5.7 million viewers
>
>
>
> Surely cancellation can't be far away.

Unlike American TV networks, the BBC isn't driven solely by the idiotic
and inaccurate stupidity of the almighty ratings system.

[Tahiri]

> 5.7 million viewers
>
> Surely cancellation can't be far away.
>

> PJR :-)

Your wishful thinking will not make it so. The first figures
don't include iplayer do they? Since it is shown at a more
inconvenient time now more people, like me, have to
watch it later.

[Keith Cunningham]

"Your Name" <Your...@YourISP.com> wrote in message
news:YourName-170...@203-118-187-243.dsl.dyn.ihug.co.nz...

How can any ratings system be valid unless it can be proven that everyone
has been consulted/ Nobody has ever asked me what I've been watching so that
means that the figures are already out by 1. And how many other people
haven't been asked?

[Ross]

On Apr 16, 5:33 pm, Peter J Ross <p...@example.invalid> wrote:
> Series 7A, Episode 3 ("A Town Called Mercy") -
>
> 6.6 million viewers
>
> Series 7B, Episode 3 ("A Disgusting Insult to the Memory of Brian
> Hayles and Bernard Bresslaw") -
>
> 5.7 million viewers
>
> Surely cancellation can't be far away.
>

> --
> PJR :-)
>
> πολλοῖς δ' ἀντιλέγειν μὲν ἔθος περὶ παντὸς ὁμοίως,
>    ὀρθῶς δ' ἀντιλέγειν, οὐκέτι τοῦτ' ἐν ἔθει. (Euenus)

Is the reason that you flag all your posts No-Archive because you
don't want people to look back and see that you've been making the
exact same predictions for seven years now?

[The Doctor]

In article <YourName-170...@203-118-187-243.dsl.dyn.ihug.co.nz>,

That is good news.

[The Doctor]

In article <n7CdnfrktMoU2_PM...@brightview.co.uk>,

That's PJR for you.

[Stephen Wilson]

"Peter J Ross" <p...@example.invalid> wrote in message
news:slrnkmrgu...@pjr.no-ip.org...

> Series 7A, Episode 3 ("A Town Called Mercy") -
>
> 6.6 million viewers
>
>
>
> Series 7B, Episode 3 ("A Disgusting Insult to the Memory of Brian
> Hayles and Bernard Bresslaw") -
>

> 5.7 million viewers
>
>
>
> Surely cancellation can't be far away.

The lowest overnight figure for series 7 so far is 5.49 million viewers (The
Power of Three). At 5.7 million, there is no indication that Dt Who is on a
downward trend. And, of course, the consolidated figures add about another 2
million to the overnights.

But let's ignore the actual viewing numbers for a minute and look instead at
how Dr Who has performed against other programmes - in the same time slot
and throughout the week. The answer - it's in pretty much the same position
as it was 4 or 5 years ago.

Even if that was not the case, the BBC are not going to be in a hurry to axe
what is one of their biggest selling programmes world-wide. Perhaps I should
rephrase that. It's not one of their biggest selling programmes. It IS their
biggest selling programme. In 2010, Dr Who was top, followed by Top Gear,
Lonely Planet, Strictly Come Dancing and BBC Earth. In 2011, Dr Who was top,
followed by Sherlock, Top Gear and Human Planet. Dr Who was also the 3rd
biggest seller on i-tunes in the USA - behind Mad Men and Glee.

Surely cancellation is far away.

[Your Name]

In article <l9mdnR1NrdZP9fPM...@brightview.co.uk>, "Keith

Bazillions. Out of that supposed 5.7 million viewers, you'll be lucky if
they actually asked more than about 1000.

But that is the only fault with the system. They also pick and choose who
they include, they ignore such things as people recording to view later,
people downloading, etc.

[The Doctor]

In article <iUAbt.8027$FR3....@fx30.fr7>,

10/10 SW.

[Tim Roll-Pickering]

Your Name wrote:

> Bazillions. Out of that supposed 5.7 million viewers, you'll be lucky if
> they actually asked more than about 1000.

That's a standard sampling technique as practised by pollsters fo decades.

> But that is the only fault with the system. They also pick and choose who
> they include, they ignore such things as people recording to view later,
> people downloading, etc.

Timeshifting has been included in the final ratings for many years now.
However the overnight ratings only include VOSDAL timeshifting so you often
get inaccurate comparisons acoss years because people either overlook that
they're not comparing like with like, or just pick the figures that best
suit their argument.

--
My blog: http://adf.ly/4hi4c

[The Doctor]

In article <slrnkn08u...@pjr.no-ip.org>,

Peter J Ross <peadar...@gmx.com> wrote:

>In rec.arts.drwho on Wed, 17 Apr 2013 04:18:24 -0700 (PDT), Ross
>wrote:

>No, it's because I don't want a dodgy multinational corporation to
>profit from the obsessive interest some people have in looking up my
>posts from seven years ago.
>
>If you gave up being a Google Groper, you'd find that real newsservers
>keep my posts practically for ever. You'd also see that I first
>started talking about ratings decline when the ratings started to
>decline, which was when Stephen "Moronic Twat" Moffat's first episode
>as show-runner was transmitted.

>
>
>--
>PJR :-)
>
>πολλοῖς δ' ἀντιλέγειν μὲν ἔθος περὶ παντὸς ὁμοίως,
> ὀρθῶς δ' ἀντιλέγειν, οὐκέτι τοῦτ' ἐν ἔθει. (Euenus)

What decline.

[The Doctor]

In article <slrnkn09t...@pjr.no-ip.org>,

Peter J Ross <peadar...@gmx.com> wrote:

>In rec.arts.drwho on Wed, 17 Apr 2013 18:23:18 +0100, Stephen Wilson
>wrote:

>
>>
>> "Peter J Ross" <p...@example.invalid> wrote in message
>> news:slrnkmrgu...@pjr.no-ip.org...
>>> Series 7A, Episode 3 ("A Town Called Mercy") -
>>>
>>> 6.6 million viewers
>>>
>>>
>>>
>>> Series 7B, Episode 3 ("A Disgusting Insult to the Memory of Brian
>>> Hayles and Bernard Bresslaw") -
>>>
>>> 5.7 million viewers
>>>
>>>
>>>
>>> Surely cancellation can't be far away.
>>
>> The lowest overnight figure for series 7 so far is 5.49 million viewers (The
>> Power of Three). At 5.7 million, there is no indication that Dt Who is on a
>> downward trend. And, of course, the consolidated figures add about another 2
>> million to the overnights.
>

>They need to work out a way of adding 3 million instead of 2 million
>when the imaginary figures are cobbled together. Nothing else will
>keep the Moffat Mafia happy.
>
>Viewing figures fluctuate, but the overall trend seems to me to
>downward.

>
>> But let's ignore the actual viewing numbers for a minute and look instead at
>> how Dr Who has performed against other programmes - in the same time slot
>> and throughout the week. The answer - it's in pretty much the same position
>> as it was 4 or 5 years ago.
>

>The ITV opposition consists of an elderly game show. What ITV needs is
>a "Moffat Killer" - a potentially popular show that they're willing to
>plug the way the BBC plugs Moffat's risible imitation of RTD's Doctor
>Who.

>
>>
>> Even if that was not the case, the BBC are not going to be in a hurry to axe
>> what is one of their biggest selling programmes world-wide. Perhaps I should
>> rephrase that. It's not one of their biggest selling programmes. It IS their
>> biggest selling programme. In 2010, Dr Who was top, followed by Top Gear,
>> Lonely Planet, Strictly Come Dancing and BBC Earth. In 2011, Dr Who was top,
>> followed by Sherlock, Top Gear and Human Planet. Dr Who was also the 3rd
>> biggest seller on i-tunes in the USA - behind Mad Men and Glee.
>>
>> Surely cancellation is far away.
>

>You're probably right. I find it irresistible to predict cancellation,
>just so I can watch the silly fans splutter and whinge. In reality, I
>think (as I've said before) that (as I've said before) that (as I've
>said before) that (as I've said before) that (as I've said before)
>that (as I've said before) that (as I've said before) that (as I've
>said before) that (as I've said before) that a move away from Saturday
>nights and BBC 1 is more likely. Again, that depends on whether or
>not ITV chooses to compete /mano a mano/.

>
>
>--
>PJR :-)
>
>πολλοῖς δ' ἀντιλέγειν μὲν ἔθος περὶ παντὸς ὁμοίως,
> ὀρθῶς δ' ἀντιλέγειν, οὐκέτι τοῦτ' ἐν ἔθει. (Euenus)

Far away as in 50 years from now.

[The Doctor]

In article <slrnkn0b4...@pjr.no-ip.org>,

Peter J Ross <peadar...@gmx.com> wrote:

>In rec.arts.drwho on Thu, 18 Apr 2013 18:04:57 +0100, Peter J Ross
>wrote:

>
>> (as I've said before) that (as I've said before) that (as I've
>> said before) that (as I've said before) that (as I've said before)
>> that (as I've said before) that (as I've said before) that (as I've
>> said before) that (as I've said before) that
>

>Sorry about that. My mouse buttons are sticking.

>
>
>
>
>--
>PJR :-)
>
>πολλοῖς δ' ἀντιλέγειν μὲν ἔθος περὶ παντὸς ὁμοίως,
> ὀρθῶς δ' ἀντιλέγειν, οὐκέτι τοῦτ' ἐν ἔθει. (Euenus)

KBs still rule!

[The Doctor]

In article <slrnkn0en...@pjr.no-ip.org>,

Peter J Ross <peadar...@gmx.com> wrote:

>In rec.arts.drwho on Thu, 18 Apr 2013 18:09:01 +0100, Tim

>Roll-Pickering wrote:
>
>> Your Name wrote:
>>
>>> Bazillions. Out of that supposed 5.7 million viewers, you'll be lucky if
>>> they actually asked more than about 1000.
>>
>> That's a standard sampling technique as practised by pollsters fo decades.
>>
>>> But that is the only fault with the system. They also pick and choose who
>>> they include, they ignore such things as people recording to view later,
>>> people downloading, etc.
>>
>> Timeshifting has been included in the final ratings for many years now.
>

>"has been for many years" doesn't imply "ought to be". Both BBC and
>ITV find it advantageous to pretend that their programmes are more
>popular than they really are. Similarly, compare the government's
>"official" figures for unemployment, inflation and immigration with
>the reality.

>
>> However the overnight ratings only include VOSDAL timeshifting so you often
>> get inaccurate comparisons acoss years because people either overlook that
>> they're not comparing like with like, or just pick the figures that best
>> suit their argument.
>

>Is there a technical explanation for the 2 million drop in ratings
>when Tennant + RTD were replaced by Smith + Moffat? No. Ratings fell
>because an attractive star and a competent showrunner were replaced by
>Frankenstein's monster and a pompous twit.

>
>--
>PJR :-)
>
>πολλοῖς δ' ἀντιλέγειν μὲν ἔθος περὶ παντὸς ὁμοίως,
> ὀρθῶς δ' ἀντιλέγειν, οὐκέτι τοῦτ' ἐν ἔθει. (Euenus)

Smith reminds me too luch of Tennant.

[Your Name]

In article <atanqj...@mid.individual.net>, "Tim Roll-Pickering"

<T.C.Roll-...@qmul.ac.uk> wrote:
> Your Name wrote:
>
> > Bazillions. Out of that supposed 5.7 million viewers, you'll be lucky if
> > they actually asked more than about 1000.
>
> That's a standard sampling technique as practised by pollsters fo decades.

Yep, and why all polls / surveys / studies are completely useless (the
only exception is that they do give results for the people actually in the
survey).

It's part of the reason why the idiot "experts" keep changing their minds
why some particular food is good for you. They "study" half a dozen people
and then claim the results are true for everyone ... then another "expert"
studies a different half dozen people and, surprise, comes up with
different results.

Only absolutely brainless morons blindly believe the results of any poll /
survey / study ... which is of course why management uses them to decide
which shows to keep going. :-\

> because people either overlook that they're not comparing like with like,
> or just pick the figures that best suit their argument.

And here's another reason silly polls are useless ... they are usually
manipulated to "prove" whatever the person paying for or running the
survey wants it to "prove". :-\

[Tim Roll-Pickering]

Your Name wrote:

>> > Bazillions. Out of that supposed 5.7 million viewers, you'll be lucky
>> > if
>> > they actually asked more than about 1000.

>> That's a standard sampling technique as practised by pollsters fo
>> decades.

> Yep, and why all polls / surveys / studies are completely useless (the
> only exception is that they do give results for the people actually in the
> survey).

A chef doesn't need to consume an entire saucepan of soup to determine if
it's right or not. They merely need to taste a representative sample. The
principle is the same here.

> It's part of the reason why the idiot "experts" keep changing their minds
> why some particular food is good for you. They "study" half a dozen people
> and then claim the results are true for everyone ... then another "expert"
> studies a different half dozen people and, surprise, comes up with
> different results.

That's a different field of scientific analysis as opposed to opinion
analysis.

> Only absolutely brainless morons blindly believe the results of any poll /
> survey / study ... which is of course why management uses them to decide
> which shows to keep going. :-\

How else should they decide? By weight of letters?

>> because people either overlook that they're not comparing like with like,
>> or just pick the figures that best suit their argument.

> And here's another reason silly polls are useless ... they are usually
> manipulated to "prove" whatever the person paying for or running the
> survey wants it to "prove". :-\

No. Reputable firms can and do get pulled apart for that. The media often
misreport polls but the industry itself is generally able to differentiate
between metholodgical disagreement and cowboys trying to pull a fast one.

[Your Name]

In article <ato9em...@mid.individual.net>, "Tim Roll-Pickering"

<T.C.Roll-...@qmul.ac.uk> wrote:
> Your Name wrote:
>
> >> > Bazillions. Out of that supposed 5.7 million viewers, you'll be lucky
> >> > if
> >> > they actually asked more than about 1000.
>
> >> That's a standard sampling technique as practised by pollsters fo
> >> decades.
>
> > Yep, and why all polls / surveys / studies are completely useless (the
> > only exception is that they do give results for the people actually in the
> > survey).
>
> A chef doesn't need to consume an entire saucepan of soup to determine if
> it's right or not. They merely need to taste a representative sample. The
> principle is the same here.

A chef also doesn't usually claim that the entire saucepan contents will
be exactly the same, onr claim that ALL the saucepans in the entire world
are the same as his tiny sample.

> > It's part of the reason why the idiot "experts" keep changing their minds
> > why some particular food is good for you. They "study" half a dozen people
> > and then claim the results are true for everyone ... then another "expert"
> > studies a different half dozen people and, surprise, comes up with
> > different results.
>
> That's a different field of scientific analysis as opposed to opinion
> analysis.

It's the exact same process - survey a tiny percentage of people and
pretend the results are the same for everyone. It's complete and utter
nonsense and garbage believed by blinkered morons (hence why network
management use it).

[The Doctor]

In article <YourName-240...@203-118-187-88.dsl.dyn.ihug.co.nz>,

Hence why J Powell is no longer popular at the BBC.

[Stephen Wilson]

"Your Name" <Your...@YourISP.com> wrote in message

news:YourName-240...@203-118-187-88.dsl.dyn.ihug.co.nz...

>>
>> That's a different field of scientific analysis as opposed to opinion
>> analysis.
>
> It's the exact same process - survey a tiny percentage of people and
> pretend the results are the same for everyone. It's complete and utter
> nonsense and garbage believed by blinkered morons (hence why network
> management use it).

Sampling can cause problems - mainly when the sample population is heavily
biased one way or another. But to claim it's complete and utter nonsense is
ridiculous. Sampling techniques have been around for centuries. Statiticians
are aware of the pros and cons. Whilst it can never be 100% accurate, it's
fine for producing estimates. And those estimates are often reasonably
accurate.

[Your Name]

In article <RdEdt.39002$Hx7....@fx11.fr7>, "Stephen Wilson"

That depends on your defiition of "reasonably accurate" ... out by
millions of people is not "reasonable", supposed results that keep chaging
every other week are not "reasonable".

ALL polls, surveys, studies, etc., including the ridiculous rating system,
the box office counts, and even the useless Census government folls force
people to fill in, are only useful as extra paper in the recycling bin.
They're all falwed and faulty, and the so-called results mis-represented
and manipulated. Any results from them are only accurate for the people
actually surveyed. Claiming those results are true for everyone is
ridiculous nonsense .. and damn dangerous when it comes to medical
"studies".

[Ross]

On Apr 23, 4:32 pm, "Tim Roll-Pickering" <T.C.Roll-

Picker...@qmul.ac.uk> wrote:
> Your Name wrote:
> >> > Bazillions. Out of that supposed 5.7 million viewers, you'll be lucky
> >> > if
> >> > they actually asked more than about 1000.
> >> That's a standard sampling technique as practised by pollsters fo
> >> decades.
> > Yep, and why all polls / surveys / studies are completely useless (the
> > only exception is that they do give results for the people actually in the
> > survey).
>
> A chef doesn't need to consume an entire saucepan of soup to determine if
> it's right or not. They merely need to taste a representative sample. The
> principle is the same here.

The trick is to actually know when you've got a represenative sample.
Because if you just yank out a bay leaf and chew on it, you're going
to think that the soup is mildly toxic

[The Doctor]

In article <YourName-240...@203-118-187-8.dsl.dyn.ihug.co.nz>,

Census useless? Unless your name is Stephen Haper.

[Tim Roll-Pickering]

Peter J Ross wrote:

>> Timeshifting has been included in the final ratings for many years now.

> "has been for many years" doesn't imply "ought to be". Both BBC and
> ITV find it advantageous to pretend that their programmes are more
> popular than they really are.

They're reflecting a growing change in the nature of watching television
whereby many viewers are using technology to bypass the schedule altogether.
There are already many shows that have far more people watching on timeshift
than during the announced broadcast.

And technology has already got to the point where it's not only possible but
it's available in many homes to combine viewing both live and timeshifting -
either by pausing the live programme or by recording and catching up. (I
often use the latter method to bypass adverts.) So you can't even filter out
live any more.

> Similarly, compare the government's
> "official" figures for unemployment, inflation and immigration with
> the reality.

The reality being what? What someone down the pub thinks is the truth
contrary to all the available data?

> Is there a technical explanation for the 2 million drop in ratings
> when Tennant + RTD were replaced by Smith + Moffat?

Unfortunately I can't currently access the full ratings list or all past
posts here so please can you state the episodes and figures your comparison
refers to.

[Your Name]

In article
<a99c3a9d-e35b-494e...@n5g2000yqg.googlegroups.com>, Ross
<rras...@trenchcoatsoft.com> wrote:
> On Apr 23, 4:32=A0pm, "Tim Roll-Pickering" <T.C.Roll-

Judging by most high-price restaurants and the miserly single spoon-dollop
they give you, it probably is toxic in anything but tiny amounts. ;-)

But you've proven the point. If you take a small sample from one place, it
will not be the same as a small sample from another place, nor the same as
the whole thing ... which is why polls / surveys / studies are
ridiculously useless when used to pretend they remotely mean anything for
everyone, and only incompetent idiots (i.e. management) blindly believe
the idiotic "results" of such things.

[Your Name]

In article <atr1gj...@mid.individual.net>, "Tim Roll-Pickering"
<T.C.Roll-...@qmul.ac.uk> wrote:

> Peter J Ross wrote:
> >
> > Similarly, compare the government's
> > "official" figures for unemployment, inflation and immigration with
> > the reality.
>
> The reality being what? What someone down the pub thinks is the truth
> contrary to all the available data?

The "available data" relies on things like people signing up for
unemployment assistance, Census data, etc. Not everyone does that, so
unemployment figures are simply another stated "fact" from polls and
surveys that is never actually accurate.

[Tim Roll-Pickering]

Your Name wrote:

>> > Similarly, compare the government's
>> > "official" figures for unemployment, inflation and immigration with
>> > the reality.

>> The reality being what? What someone down the pub thinks is the truth
>> contrary to all the available data?

> The "available data" relies on things like people signing up for
> unemployment assistance, Census data, etc. Not everyone does that, so
> unemployment figures are simply another stated "fact" from polls and
> surveys that is never actually accurate.

Yes all data collection has some gaps. But "the reality" claimed the
contrary is invariably on far less firm ground.

[Tim Roll-Pickering]

Ross wrote:

> > A chef doesn't need to consume an entire saucepan of soup to determine
> > if
> > it's right or not. They merely need to taste a representative sample.
> > The
> > principle is the same here.

> The trick is to actually know when you've got a represenative sample.
> Because if you just yank out a bay leaf and chew on it, you're going
> to think that the soup is mildly toxic

Yes and chefs will resort to methods such as stirring to mix it through when
they take the sample or else using a tool to extract samples from multiple
levels. Polling works on a similar rationale by demographically balancing
and weighting the sample.

[Tim Roll-Pickering]

Your Name wrote:

>> > Yep, and why all polls / surveys / studies are completely useless (the
>> > only exception is that they do give results for the people actually in
>> > the
>> > survey).

>> A chef doesn't need to consume an entire saucepan of soup to determine if
>> it's right or not. They merely need to taste a representative sample. The
>> principle is the same here.

> A chef also doesn't usually claim that the entire saucepan contents will
> be exactly the same,

That's called the margin of error.

> onr claim that ALL the saucepans in the entire world
> are the same as his tiny sample.

He's testing his own saucepan but the principle is the same.

[Tim Roll-Pickering]

Your Name wrote:

>> > A chef doesn't need to consume an entire saucepan of soup to determine
>> > if
>> > it's right or not. They merely need to taste a representative sample.
>> > The
>> > principle is the same here.

>> The trick is to actually know when you've got a represenative sample.
>> Because if you just yank out a bay leaf and chew on it, you're going
>> to think that the soup is mildly toxic

> Judging by most high-price restaurants and the miserly single spoon-dollop
> they give you, it probably is toxic in anything but tiny amounts. ;-)

That's why I never order soup in restaurants. ;)

> But you've proven the point. If you take a small sample from one place, it
> will not be the same as a small sample from another place, nor the same as
> the whole thing ...

But the sample is not taken from one place alone. The sample as a whole is
taken in the best way to draw from the whole - in the case of soup stirring
and digging deep, in the case of polls through ensuring the sample is both
sufficiently large and demographically mixed.

> which is why polls / surveys / studies are
> ridiculously useless when used to pretend they remotely mean anything for
> everyone, and only incompetent idiots (i.e. management) blindly believe
> the idiotic "results" of such things.

The polls/surveys/studies are based on rather more sophisticated sampling
than you assume.

[Your Name]

In article <atrhu1...@mid.individual.net>, "Tim Roll-Pickering"

By the time morons in Government and the media are talking about those
figures, they're also already months out of date.

[Your Name]

In article <atrhu2...@mid.individual.net>, "Tim Roll-Pickering"

<T.C.Roll-...@qmul.ac.uk> wrote:
> Your Name wrote:
>
> >> > Yep, and why all polls / surveys / studies are completely useless (the
> >> > only exception is that they do give results for the people actually in
> >> > the
> >> > survey).
>
> >> A chef doesn't need to consume an entire saucepan of soup to determine if
> >> it's right or not. They merely need to taste a representative sample. The
> >> principle is the same here.
>
> > A chef also doesn't usually claim that the entire saucepan contents will
> > be exactly the same,
>
> That's called the margin of error.

The bit the chef tastes is fine, the bit he serves you in fatally toxic
... that's some "error". ;-)

> > onr claim that ALL the saucepans in the entire world
> > are the same as his tiny sample.
>
> He's testing his own saucepan but the principle is the same.

Yes, but the idiotic polls and surveys claim their "results" are for everyone.

[Your Name]

In article <atrhu2...@mid.individual.net>, "Tim Roll-Pickering"

Also known as manipulating the results by picking and choosing who can /
cannot be surveyed ... making the so-called results are even less useful
than a truly random sample.

[Your Name]

In article <atrhu2...@mid.individual.net>, "Tim Roll-Pickering"

<T.C.Roll-...@qmul.ac.uk> wrote:
> Your Name wrote:
> >> >
> >> > A chef doesn't need to consume an entire saucepan of soup to determine
> >> > if it's right or not. They merely need to taste a representative sample.
> >> > The principle is the same here.
>
> >> The trick is to actually know when you've got a represenative sample.
> >> Because if you just yank out a bay leaf and chew on it, you're going
> >> to think that the soup is mildly toxic
>
> > Judging by most high-price restaurants and the miserly single spoon-dollop
> > they give you, it probably is toxic in anything but tiny amounts. ;-)
>
> That's why I never order soup in restaurants. ;)

That's why I never go to a restaurant. :-p

> > But you've proven the point. If you take a small sample from one place, it
> > will not be the same as a small sample from another place, nor the same as
> > the whole thing ...
>
> But the sample is not taken from one place alone. The sample as a whole is
> taken in the best way to draw from the whole - in the case of soup stirring
> and digging deep, in the case of polls through ensuring the sample is both
> sufficiently large and demographically mixed.

Except that stirring it only moves the poison elsewhere. ;-)

> > which is why polls / surveys / studies are
> > ridiculously useless when used to pretend they remotely mean anything for
> > everyone, and only incompetent idiots (i.e. management) blindly believe
> > the idiotic "results" of such things.
>
> The polls/surveys/studies are based on rather more sophisticated sampling
> than you assume.

No they aren't - they're done exactly how I know they are done. The
results are reported in ways to purposely fool the naive. I've done
mathematical statistics through university level and run many proper
surveys that don't claim ridiculous guesswork results for everyone.

[Your Name]

In article <atrhu2...@mid.individual.net>, "Tim Roll-Pickering"

<T.C.Roll-...@qmul.ac.uk> wrote:
>
> But the sample is not taken from one place alone. The sample as a whole is
> taken in the best way to draw from the whole - in the case of soup stirring
> and digging deep, in the case of polls through ensuring the sample is both
> sufficiently large and demographically mixed.

The fact is that most of these surveys use ridiculously small sample sizes
and then simply mathematically multiply the real results out and claim it
means something for everyone. Bzzzzt ... wrong.

Unless those samples are large enough to include EVERYONE (and you can
guarantee nodoby lied), claiming the results aree for everyone is simply
ridiculous. It's nothing better than guesswork and manipulation ...
they're are not and cannot be the accurate, true results they claim to be.

[The Doctor]

In article <atrhu2...@mid.individual.net>,

Pter J Ross should go to imdb.com and talk about decline.

He will be wearing mud!

[Stephen Wilson]

"Your Name" <Your...@YourISP.com> wrote in message

news:YourName-250...@203-118-187-90.dsl.dyn.ihug.co.nz...

OK then - so how do you propose to include EVERYONE, make sure they all fill
in their forms, check they do it correctly, on time and without telling any
lies, then analyse all the results while they're still valid?

Answer - you can't. In the case of viewing figures, you could insist that
every piece of equipment capable of displaying a broadcast is connected to
the internet so you could log every single programme that is shown. Even if
you could get past all the technical problems, cost, privacy objections, etc
you still would't know how many people were watching, what they thought of
the programme, etc.

In the meantime, statistical sampling represents the most economical way to
get meaningful results.

[The Doctor]

In article <ksfet.6925$86....@fx09.fr7>,

Check imdb.com

[Stephen Wilson]

"The Doctor" <doc...@doctor.nl2k.ab.ca> wrote in message
news:klc0kg$cpv$1...@gallifrey.nk.ca...

What do you think a movie database has got to do with this thread, Yads?

[Your Name]

In article <ksfet.6925$86....@fx09.fr7>, "Stephen Wilson"
<stephen.w...@nospam.ntlworld.com> wrote:
>

<snip>

>
> In the meantime, statistical sampling represents the most economical way to

> get meaningless results peddled as "fact" for everyone and beleived by
> dumasses.

Corrected that last sentence for you.

The results of any poll / survey / study are ONLY meaningful for those who
were surveyed (and assuming they didn't lie). They are not, cannot, and
never will be meaningful for the millions of other people the idiots
running the survey GUESSED the results for ... if they actually did
survery everyone they would get different results - FACT.

End of me wasting time on this topic.

[The Doctor]

In article <eKfet.50796$xs.1...@fx28.fr7>,

Demonstrate the show popularity if you check it out.

[The Doctor]

In article <YourName-260...@203-118-187-62.dsl.dyn.ihug.co.nz>,

REcall Dewey was porjected the winner over Truman in the US elections based
on polling.

[Stephen Wilson]

"The Doctor" <doc...@doctor.nl2k.ab.ca> wrote in message

news:klcb4c$moa$1...@gallifrey.nk.ca...

OK, let's check the last episode. Hide. Checking it right now shows that 359
people have rated it. Not 359 people from the UK - that's 359 people from
around the world. Are you saying that only 359 people have watched the
story? Are you saying that 359 is a sample? If so, do you believe it to be a
reasonablr sample featuring a good cross-section of the general public? How
would you work out from that sample how many people actually watched it? Or
the age ranges it appealed to?

[The Doctor]

In article <zRDet.122862$Pj4....@fx14.fr7>,

iWhat breakdown do you have?

[Your Name]

In article <zRDet.122862$Pj4....@fx14.fr7>, "Stephen Wilson"

<stephen.w...@nospam.ntlworld.com> wrote:
> "The Doctor" <doc...@doctor.nl2k.ab.ca> wrote in message
> news:klcb4c$moa$1...@gallifrey.nk.ca...
> >>>

> >>> Check imdb.com
> >
> >>What do you think a movie database has got to do with this thread, Yads?
> >
> > Demonstrate the show popularity if you check it out.
>
> OK, let's check the last episode. Hide. Checking it right now shows that 359
> people have rated it. Not 359 people from the UK - that's 359 people from
> around the world. Are you saying that only 359 people have watched the
> story? Are you saying that 359 is a sample? If so, do you believe it to be a
> reasonablr sample featuring a good cross-section of the general public? How
> would you work out from that sample how many people actually watched it? Or
> the age ranges it appealed to?

The same way Nielsen and others do ... you guess and then pretend it means
something. :-(

[Stephen Wilson]

"The Doctor" <doc...@doctor.nl2k.ab.ca> wrote in message

news:klf4ai$842$1...@gallifrey.nk.ca...

I don't have a breakdown. It was you who said I should check imdb. So
presumably you know where the breakdowns are.

[Tim Roll-Pickering]

The Doctor wrote:

> REcall Dewey was porjected the winner over Truman in the US elections
> based
> on polling.

If you've got to go back 65 years then you're overlooking the changes in
methodology since then and also the cause of the problem.

Firstly everyone remembers the Chicago Tribune headline but it's forgotten
that in 1948 that paper was suffering a printers' strike that forced it to
use an alternate method that required it to go to press hours earlier than
usual before a lot of polling stations had reported. But it was far from
alone, and is only remembered because Truman himself posed with a copy.

The main problem was the belief at the time that voters made up their mind
at least a month before polling day. There had been some then-recent
elections which supported this view, and so many firms stopped polling in
the last weeks of a campaign. Virtually all the conventional wisdom was that
Dewey would win - it's said that Truman couldn't convince even his own wife
he had a chance and his campaign staff were humouring him with a last
hurrah - and a lot of the polls actually stopped early.

Most of the media starting preparing coverage of a Dewey victory (for
instance NBC prepared for elephants to come out of a model of the White
House but didn't bother with donkeys) without waiting for the formality,
rather than paying attention to the few polls that were still being
conducted, which showed that Truman was closing the gap with his ferocious
underdog campaign. Had the polls carried on even further - remember the
technology restrictions of the time - they may well have shown Truman in the
lead by election day. Also it was a closer result in the electoral college
because Truman oly narrowly took each of Ohio, Illinois and California, with
78 college votes between them.