Author: Dann Corbit
Date: 07:58:01 04/05/99
I have some thoughts about opinion poll questions. The first thought is that
frequently they serve little purpose. Here is what I mean by that:
Q: Is Kasparov Stronger than Deep Blue?
Well, certainly it is an interesting question and a lot of people will have
opinions about it one way or another. But asking a poll for a question like this
does not produce a valuable output. It is like asking a question about moon
rocks, and none of us have any moon rocks to look at. We may be able to venture
an opinion, but without real data, the opinions do not have value. Furthermore,
it is possible to produce reliable data since moon rocks exist and are available
to scientists, but quite likely none of the poll-takers can get access.
Now, the reason I say that questions like this serve little purpose is that they
do not answer the question they set out to ask. Even if there is a consensus, it
means little. It could be just like the famous Indiana State legislature voting
the wrong value for pi. They did not change the constant (nature gets to decide,
not people) and only succeeded in making themselved into world famous buffoons
for choosing the wrong answer.
For this reason, I suggest that poll questions are more valuable if they can
produce data of value with a purpose. For instance, Ed could ask what kind of
features people want in a chess program. The current poll has some of that sort
of thing couched in crevices. He could ask about various categories of things
like interface, and strength in different areas, and many sorts of features.
This data has a purpose, because Ed can use the data to make future plans for
his product.
The questions do not have to be about chess programs. They could be about other
things of interest. What kind of contests would Rebel users be interested in?
Which famous chess players would they like to hear an interview from (and which
would they like to have tape over their mouth {rjf springs to mind} ;-))
Questions like that have utility because they can capture the interest of the
audience and lead to useful *actions* whereas silly opinions about how many
angels can dance on the head of a pin can only lead to more debate and no
answers.
Just my two pennies.
This page took 0 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.