| |
In lugnet.admin.general, Thomas Main writes:
> [...] Consider if, instead of a 0-100 ten point scale, there was just an
> option to rate an article as "I think this is noteworthy" By default, the
> articles an individual didn't think were great would sink to the bottom and
> articles that stood out as particularly important would rise to the top.
> The ambiguity of the ratings would be diminished too...I don't know what a
> rating of "30" versus "40" means...but I can understand "look at these
> articles...a majority of LUGNET members thought these were excellent."
Thomas: First, thanks for your comments (I don't know if I've thanked you
directly yet). Second, things have just been "peeled" way back to their bare
essentials. The underlying system is almost the same, but it's got a
completely new skin -- hopefully one which replaces discomfort with appeal.
Since the ratings were mainly intended as a recommendation-to-read scale (they
were many things, but that was the primary thing), the first thing to get rid
of was the negative stigma associated with people "rating" everything and to
objectify the input. Thus, the first change was to change the input-
solicitation question from:
How would you rate this message? Low * * * * * * * * * * * High
to:
Would you recommend this message to others? No * Yes! * * * * Yes!!!!
(the *'s represent radio buttons). The rationale for the change: Why not
simply ask the most important question directly?
This immediately cuts down on the sheer number of choices as well as the
inclination for one to mark a message highly which isn't necessaily a message
one would recommend to others.
However, this still leaves one to wonder what is the difference between
something with a recommendation of !!!! (the highest) and something with a
recommendation of "only" !!! (the next highest). Also, with four positive
choices ("Yes!" through "Yes!!!!") and one neutral choice ("No"), it still
leaves room for one to feel bad when one's message is left unmarked amid a
dozen other messages marked "!" to "!!!".
Still more simplification was needed.
The next step was to ruthlessly chop again the number of choices -- this time
from 5 down to 2 -- a bare minimum. The philosphy here now is even simpler
than a "would you recommend this?" -- it basically asks, "what type of marking
(if any) would you recommend show up next to this message?" After all, why
not just ask the question even _more_ directly so that its entire purpose is
clear in the question itself?
So, you can now ask that a message be included in the LUGNET Spotlight page by
selecting
°° Spotlight
and you can ask that a message simply by highlighted (with a ° symbol) by
selecting
° Highlight
There are no options to give a message any kind of low score, and in fact the
default choice of "- - -" (empty -- no action) causes an "erase my opinion"
action to be sent rather than an input value of 0. In other words, this
follows the old adage, "If you can't say something nice, don't say anything
at all" -- and even goes so far as to _prevent_ people from saying bad things.
The ONLY input that's allowed is positive input. You can recommend something
for Highlighing, or you can recommend something for Spotlighting, or you can
opt out -- those are the only three choices (so, two choices, really).
(Of course, something marked "Spotlight" by 3 people can still be "marked
down" a very small amount by a 4th person marking it "Highlight", but that's
not a pejorative mark in the same sense that a 0 or hard "no" would be.)
> Simplifying the system would also eliminate some of the subjectivity of the
> scores the articles receive. Someone might think "40" is a ok score whereas
> someone else might choose "60" as a low score. There's less confusion about
> a system that just uses "noteworthy" as a "good" score and lets other
> messages default to "no comment."
Thomas, have a look now -- if you're still with us -- this is basically how
it turned out. No more 40's, no more 60's, no more 100's, no more 0's.
Things not marked as noteworthy (i.e., Highlight or Spotlight) sit pristinely
unmarred.
Internally (this is just some geek details), the "Highlight" recommendation
translates into a numeric value of 75, and the "Spotlight" recommendation
translates into a numeric value of 100. A default "softener" value of 0 is
still included by the system for all messages, in order to prevent "pegging"
by a small number of inputs.
Thus, it now takes two people both rating (er, highlighting) something in
order for any mark at all to even show up next to an article (they would both
have to give it a "Spotlight" recommendation). And it takes at least 4 people
all recommending something as "Spotlight" in order for a °° Spotlight mark to
show up.
A couple of one-person-input examples:
Person A ° Highlight => 75
Softener => 0
COMPOSITE => 37.50 => (blank)
Person A °° Spotlight => 100
Softener => 0
COMPOSITE => 50.00 => (blank)
And a couple of two-person-input examples:
Person A ° Highlight => 75
Person B ° Highlight => 75
Softener => 0
COMPOSITE => 50.00 => (blank)
Person A °° Spotlight => 100
Person B ° Highlight => 75
Softener => 0
COMPOSITE => 58.33 => ° (Highlight)
Person A °° Spotlight => 100
Person B °° Spotlight => 100
Softener => 0
COMPOSITE => 66.67 => ° (Highlight)
And a couple of three-person-input examples:
Person A °° Spotlight => 100
Person B ° Highlight => 75
Person C ° Highlight => 75
Softener => 0
COMPOSITE => 62.50 => ° (Highlight)
Person A °° Spotlight => 100
Person B °° Spotlight => 100
Person C ° Highlight => 75
Softener => 0
COMPOSITE => 68.75 => ° (Highlight)
And a four-person-input example:
Person A °° Spotlight => 100
Person B °° Spotlight => 100
Person C °° Spotlight => 100
Person D °° Spotlight => 100
Softener => 0
COMPOSITE => 80.00 => °° (Spotlight)
Thus, this simplified system requires a very solid base of input from several
people before significant recommendation symbols show up. It's designed to
make abuse very, very difficult.
Internally, there is a table which simply associates the internal composite
numbers with what to show as symbols (output):
0-50 => (blank)
51-75 => ° (Highlight)
76-100 => °° (Spotlight)
Thomas (and anyone else), does this seem like a positive change to you?
My thinking here is if people can see _exactly_ what they're doing when they
mark articles, and if they are physically unable to "begrudge" articles, then
only good can come from this. And if the LUGNET Spotlight page can be
automated to display things based directly on these recommendations, then
we'll have a collaboratively generated Spotlight page more representative
of community opinion, instead of one which is hand-picked by a single person
(currently the case).
In this sense, the highlights (down with ratings!!!) are sort of like a
community project -- people working toward a common goal rather than serving
their personal interests.
--Todd
|
|
Message has 7 Replies:
Message is in Reply To:
 | | Re: the latest news
|
| In lugnet.lego.direct, Todd Lehman writes: <snip> (...) <snip> (...) The rating system has seriously made me (and perhaps others?) consider returning my LUGNET membership card. It seems to me that a few people enjoy rating the "newsworthiness" of (...) (25 years ago, 18-Apr-00, to lugnet.admin.general) !
|
75 Messages in This Thread:
 
- Entire Thread on One Page:
- Nested:
All | Brief | Compact | Dots
Linear:
All | Brief | Compact
This Message and its Replies on One Page:
- Nested:
All | Brief | Compact | Dots
Linear:
All | Brief | Compact
|
|
|
|
