 |
|
|
|
|
|
| |
| |
|
|
|
|
| |
| |
|
|
John VanSickle <evi### [at] hotmail com> wrote:
> Another idea occurs to me:
>
> Allow voting by the general public. Average all of these votes
> together. This becomes one vote, which has weight equal to the entrant
> judges and the permanent judges.
>
> This allows the public to vote, but minimizes the harm that a spam
> campaign can do.
>
> Regards,
> John
Although a good idea, couldn't someone just submit like, 100 perfect votes for
themself, and improve odds.
Personally, I think voting should be kept to just inside entrants. Even there,
an entrant can give a 0 score to everyone else, but everyone is at a certain
level of fair play, and would therefore not submit to such low values (I
hope!).
I'm not making the choice however. The idea is good, but it just doesn't seem
to work too well.
But maybe, you could use ip addresses to back check spam votes.???
-Nimish
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
David Buck wrote:
> Your thoughts would be appreciated.
Might be too late to mention this, but I've always felt that a 5-point
score system is easier for users :)
After all, pretty much everyone is familiar with it, with the values
being "Horrible, poor, average, good, great". Rarely do you need more
granularity over the actual scores, as well.
With a ten point system, some people might feel that "5" is average,
while other people feel that anything less than "9.0" stinks. Of
course, individual voter tendencies are averaged out over the group, but
its still nice to have as much consistency as possible.
As far as voting goes... I think only entrants and panel judges should
be considered "official" voters, with the caveat that entering one round
should qualify you to vote for the next 12 months. Partial votes (where
not every images was rated), and votes from non-participants should be
tracked separately, perhaps as a "populist" vote, for curiosity's sake,
but should not affect the final score.
But then, I haven't entered since around 2001, so what do I know? :)
--
...Ben Chambers
www.pacificwebguy.com
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Chambers wrote:
> David Buck wrote:
>> Your thoughts would be appreciated.
>
> Might be too late to mention this, but I've always felt that a 5-point
> score system is easier for users :)
A 5 point system is just as easy to implement as a 10 point system. I'm
open to this suggestion if people think it's the best way.
David Buck
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Nimish Ajmani wrote:
> But maybe, you could use ip addresses to back check spam votes.???
>
> -Nimish
>
The problem with IP addresses is that many people use DHCP which
dynamically allocates IP addresses. The same person may get different
IP addresses each time they logon and the IP address used by a spammer
may later be used by a legitimate voter. Besides, I would like the
ability to vote from whatever IP address I happen to be using at the time.
David Buck
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
David Buck <dav### [at] simberoncom> wrote:
> Nimish Ajmani wrote:
>
> > But maybe, you could use ip addresses to back check spam votes.???
> >
> > -Nimish
> >
>
> The problem with IP addresses is that many people use DHCP which
> dynamically allocates IP addresses. The same person may get different
> IP addresses each time they logon and the IP address used by a spammer
> may later be used by a legitimate voter. Besides, I would like the
> ability to vote from whatever IP address I happen to be using at the time.
>
> David Buck
That's true. I hadn't thought about that.
As for points, I think 10 is better than 5, because the more points allows for a
greater degree of quality decision.
I liked the 20 point system, because, I used it much like the French grading
system, which was based on 20 points, but if I had to choose between 5 and 10,
I'd prefer 10, again, for a greater degree of quality variance.
I hope that helped.
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
David Buck wrote:
> A 5 point system is just as easy to implement as a 10 point system. I'm
> open to this suggestion if people think it's the best way.
I think 5 points is a bit coarse, there will hopefully be many
"great" images submitted (although not by me) and the voter likely
wants to give them "great" votes but still be able to make his or
her favourite stand out a bit ;) 0-10 sounds reasonable to me.
Regarding the categories, it seems a bit redundant
to vote for the "overall" image. Like, I find it's
technically horrible and of no artistic value, but a
great picture anyway? Also, the distinction between
artistic and concept was always a bit blurred.
How about just two categories "artistic" and "technical",
where "artistic" encompasses the concept as well? If this
puts too much emphasis on the technical aspects, weighted
averages such as 2:1 or 3:2 could be used. Using fewer
categories also reduces tediousness of voting when a
successful round attracts 50 to 100 entries.
Apart from that, I'd also support keeping the voting
confined to panel judges as well as present and past
entrants - whatever the time limit. Prizes could
include winning a longer right to vote ;)
Finally, I think partial votes should not count (in fact
should not be possible to submit in an automated system),
as they might be biased by the ordering of the list, and
votes should not be disclosed until voting period ends.
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
>>>>> "David" == David Buck <dav### [at] simberoncom> writes:
David> The problem with IP addresses is that many people use DHCP which
David> dynamically allocates IP addresses. The same person may get different
David> IP addresses each time they logon and the IP address used by a spammer
David> may later be used by a legitimate voter. Besides, I would like the
David> ability to vote from whatever IP address I happen to be using at the
David> time.
Not only that, but if you're voting from inside, say, yahoo.com, you will be
coming in via the corporate proxy, which is shared by thousands of others, and
might actually change from hit to hit, making your own votes invalid.
"IP" is not "user". Never make that mistake.
--
Randal L. Schwartz - Stonehenge Consulting Services, Inc. - +1 503 777 0095
<mer### [at] stonehengecom> <URL:http://www.stonehenge.com/merlyn/>
Perl/Unix/security consulting, Technical writing, Comedy, etc. etc.
See PerlTraining.Stonehenge.com for onsite and open-enrollment Perl training!
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Christian Froeschlin wrote:
> How about just two categories "artistic" and "technical",
> where "artistic" encompasses the concept as well? If this
> puts too much emphasis on the technical aspects, weighted
> averages such as 2:1 or 3:2 could be used. Using fewer
> categories also reduces tediousness of voting when a
> successful round attracts 50 to 100 entries.
How pretty art is, is different from how clever/witty it is -- just as
how effective a technique is in accomplishing an effect is also
completely different from how technically involved it was to produce.
How do you compare a well chosen/applied simple technique against an
overly ambitious (read Rube-Goldberg-inspired) technique which showed
better promise than first-shot results? I think it would be nice to be
able to credit different types of things in their own distinct
categories so people know in what ways their work is good or not good. I
do agree though that too many would make voting tedious.
2c,
Charles
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
On Sun, 09 Mar 2008 17:42:35 -0400, David Buck wrote:
> Chambers wrote:
>> David Buck wrote:
>>> Your thoughts would be appreciated.
>>
>> Might be too late to mention this, but I've always felt that a 5-point
>> score system is easier for users :)
>
> A 5 point system is just as easy to implement as a 10 point system. I'm
> open to this suggestion if people think it's the best way.
From an implementation standpoint, the difference isn't significant - but
from the standpoint of a user, what's the difference between a "7" and an
"8", for example, on a scale of 9 or 10 points?
This is a question I've been trying to answer for course evaluations in
my "real life" job for a couple of years now. From the standpoint of
what type of scale to use, an odd-number scale works best (I find),
because there is actually a middle number in the scale (assuming an
integer scale). I use a 9-point scale (because that's what's been in
place more than anything) - 5 becomes a "neutral" score in that case.
Just something to think about.
Jim
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Charles C wrote:
> How pretty art is, is different from how clever/witty it is -- just as
> how effective a technique is in accomplishing an effect is also
> completely different from how technically involved it was to produce.
> How do you compare a well chosen/applied simple technique against an
> overly ambitious (read Rube-Goldberg-inspired) technique which showed
> better promise than first-shot results? I think it would be nice to be
> able to credit different types of things in their own distinct
> categories so people know in what ways their work is good or not good. I
> do agree though that too many would make voting tedious.
> 2c,
> Charles
Here's a novel thought: do we really need more than one category? Why
not just have one score, and let that be that? After all, is it really
worthwhile to say, "Well, *this* picture was extremely difficult for the
author to make because of the method he used, so the fact that he pulled
it off makes up for it's being a lousy image"?
---
And another thought about ratings scales: what if every user's score
were "normalized", so that the average the middle two quartiles are
scaled from 2.5-7.5 (so the mean of the two quartiles would be a 5), and
the two outlying quartiles are scaled from 0-2.5 and 7.5-10?
This could make up for individual voter biases (like the fact that some
people, when using a 10 point scale, only hand out between 9 and 10
point scores).
--
...Ben Chambers
www.pacificwebguy.com
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
|
|
