recomendations on TVs

[NSM]

| CR just reports info from their survey. They're just reporting
| what maybe 100,000 survey respondants said -- whether true or
| not. Asking on this newsgroup gives the opinion of maybe three
| people (instead of 100,000). FWIW.
|
| Mike
|
| P.S. - I agree they often don't know what they're talking about.
| Biggest problem is that they don't have the "tastes" of
| an enthusiast of whatever the subject is. They have
| the mentality that a Big Mac (tm) is a gourmet meal (which
| unfortunately is probably a good fit for most subscribers
| who are looking for low-end cheapie stuff that isn't
| outright bad).

IME when you know the subject of their investigation you often know how
wrong they are. They concentrate on unimportant aspects of the item and
ignore the really important good and bad points of whatever it is.

N

 

[LASERandDVDfan]

CR just reports info from their survey. They're just reporting

what maybe 100,000 survey respondants said -- whether true or
not. Asking on this newsgroup gives the opinion of maybe three
people (instead of 100,000). FWIW.

The problem with CR surveys is that they are made by consumers who may make the
biggest stink over what may have been due to something as simple as a loose
cable. There is a generality that is consistent, such as RCA sets being
horrible about reliability, but other claims, like Sanyo sets being the most
reliable, is simply not true.

Plus, a lot of the so-called "Consumer Reports Scientists" have problems
hooking up home theater receivers even with the aid of manuals.

Scientists, huh. Makes me want to go over there and smack each and every one
of those morons with a tack hammer.

Hooking up is much more simple than most people think. The biggest problem is
that they look and think about the whole thing instead of breaking it down
logically, paying attention only to the connections that are needed, and doing
one thing at a time. - Reinhart

 

[[email protected]]

There is a generality that is consistent, such as RCA sets being
horrible about reliability, but other claims, like Sanyo sets being the most
reliable, is simply not true.

Their report is based on a survey of over 100,000 owners of said
equipment. Out of curiosity, what is your judgement based upon?
Calculated theoretical MTBF's of those particular years worth
of product?

Hooking up is much more simple than most people think. The biggest problem is
that they look and think about the whole thing instead of breaking it down
logically, paying attention only to the connections that are needed, and doing
one thing at a time. - Reinhart

CR also tests only using the out-of-the-box settings for whatever it
is. Which
is really stupid a lot of the time. Only problem with calling them
stupid
is that I suspect the vast majority of those who subscribe to CR will
do
exactly that when they buy one. So unfortunately, it's probably the
proper
thing to do in their testing. :-(


Mike

 

[Leonard Caillouet]

Think about it. Do you really believe that they surveyed 100,000 Sanyo
owners. 100,000 of every product that they rate? Maybe 100,000 total for a
category, but more likely 100,000 in a year.

I find Sanyo products to be pretty good, in their price class, but to
consider them the most reliable TV is pretty far fetched.

Leonard

 

[James Sweet]

Think about it. Do you really believe that they surveyed 100,000 Sanyo
owners. 100,000 of every product that they rate? Maybe 100,000 total for a
category, but more likely 100,000 in a year.

I find Sanyo products to be pretty good, in their price class, but to
consider them the most reliable TV is pretty far fetched.

Statistics in general are pretty worthless, they can show trends, but
they're by no means the final word, it's true in any area.

 

[Leonard Caillouet]

for


Statistics in general are pretty worthless, they can show trends, but
they're by no means the final word, it's true in any area.

You are correct, but I'd put it slightly differently. Statistics are very
useful, when you understand the population, the assumptions about the
population, the sample, the process for selecting the sample, and the
statistics and calculations themselves. The problem with statistics is that
few of these are understood by the people using them, which makes them
useless, as you said, or worse, abused. This is Usenet, the realm of the
out-of-context and poorly informed, which makes it very likely that they are
useless.

Now that I think about it, CR fits in perfectly with Usenet...

Leonard

 

[JW]

Statistics in general are pretty worthless, they can show trends, but
they're by no means the final word, it's true in any area.

^^^^^^^^^^^^^^^^^^^^^
Well, 95% of the time anyway

 

[[email protected]]

Think about it. Do you really believe that they surveyed 100,000 Sanyo
owners. 100,000 of every product that they rate? Maybe 100,000 total for a
category, but more likely 100,000 in a year.

I find Sanyo products to be pretty good, in their price class, but to
consider them the most reliable TV is pretty far fetched.

I haven't subscribed in a while (go look at library copies when I
have a need), but they at least used to send surveys out yearly
to something like 500,000 (or more) households. So number of
responses for any give product varies with popularity, but still
quite a few.

Mike

 

[[email protected]]

total for
a


Statistics in general are pretty worthless, they can show trends, but
they're by no means the final word, it's true in any area.

It depends, some stats are VERY definitive (say, my age, salary, etc
are very definitive stats within my household). Poll stats that
come from the election process usually are pretty definitive
final words as to who gets the job(s). CR polls have their
weaknesses, but they have strength in that they have a LOT of
samples for most products reported and that those samples don't
seem to have any strong biases inherent (polling salesmen for products
made by them and their competitors might not be a useful one,
for instance). USENET information is purely anecdotal in
comparison. Instead of hundreds of thousands of opinions boiled
down into a stat, you have maybe three opinions, and by those who
self-choose to participate. Not strong on size nor randomness.

The only stats for consumer goods that might be better for
reliability might be return rates (or warantee stats) for
each model/mfgr gotten from the mfgrs. I don't think that
kind of info is released, however!

Note that something that is very crummy can be very reliable
if it is able to stay that same crumminess for a long time
and does so for most all instances of the thing. Doesn't
mean it performs well!

Mike

 

[LASERandDVDfan]

Their report is based on a survey of over 100,000 owners of said

equipment.

Yeah, but those 100,000 would be random elements of any brand from a pool of
several million. There is no way to really make a truly accurate statistic
with such a disparity in numbers.

Out of curiosity, what is your judgement based upon?

Calculated theoretical MTBF's of those particular years worth
of product?

No. Looking over a product and observing what is good engineering sense and
what isn't and recognizing other factors such as the quality of assembly and
parts.

Bad engineering sense can include an overly complex circuit design for a simple
function, taking too many shortcuts on a circuit design to try and save
manufacturing costs, and badly executed ways of securing the boards (or a
single mainboard) to the chassis.

Not one is immune to the mentioned problems, but there are brands that are
better about it than others, as well as some that are worse.

For instance, Sanyo/Fisher tends to design things simple, which may help in the
reliability front (and certainly makes things cheap) but could hamper
performance as adding things to refine performance tends to increase
complexity. But, Sanyo are far from the worst.

I used to think RCA was the worst. But, they seem to have gotten better
somewhat, and there are RCA chassis that are actually decent. Some of the
chassis that are prone to problems, like the ones that used the inline tuner
modules (the CTC numbers escape me at this moment, cracked solder ground path
problems), can be repaired to the point where the problems may not happen again
until years later.

The worst I've seen so far is stuff from Funai and Apex. Cheap designs and
cheap parts along with a cheap way of putting everything together. - Reinhart