| |
In lugnet.admin.general, Todd Lehman writes:
>
> A not-too-big-yet-not-to-small number of radio buttons with a neat numeric
> spread is 6, which yields the linear progression of values 0 -> 20 -> 40 ->
> 60 -> 80 -> 100. Implicitly, there's a 50 hidden in there by means of
> abstaining from the vote on that article.
>
> Another neat possibility is 9, which yields the linear progression of values
> 0 -> 12.5 -> 25 -> 37.5 -> 50 -> 62.5 -> 75 -> 83.5 -> 100. (That might be
> too many.)
>
> The number 7 yields the linear progression 0 -> 16.667 -> 33.333 -> 50 ->
> 66.667 -> 83.333 -> 100.
>
> I'm not sure there's much difference between 6 and 7.
OK, pruning this down to be as simple as possible -- while still retaining
the option to include an edit box for fine-tuning later down the road -- it
turns out that a six-position set of radio buttons is really slick. :)
Eliminating the "Neutral" word in the middle saves space, plus ought to lessen
any bad feelings if something ends up less than 50, because 50 isn't labeled
"neutral." 50 is just "50% swingin'" -- or something like that. The end
labels are also now more objective -- "Low" and "High". The rationale: even
though there may be such a thing as a "bad" article or a "good" article,
ultimately the scale is a read or no-read recommendation, with the score
relating some sort of urgency or importance. (This is what Steve pointed out
earlier.)
The tail end of the "Brief" viewing mode shows the score/rating like this
(a few simulated examples...)
(7 minutes ago, 13-Mar-99, to lugnet.off-topic.fun, Unrated:50)
(2 hours ago, 13-Mar-99, to lugnet.announce, Rated:75 by 1)
(4 hours ago, 12-Mar-99, to lugnet.trains, Rated:70 by 2)
(3 days ago, 10-Mar-99, to lugnet.castle, Rated:89 by 15)
(5 days ago, 8-Mar-99, to lugnet.off-topic.debate, Rated:27 by 6)
Mathematically, BTW, the 6 positions {0,20,40,60,80} combine splenderifously
with 50...(no funky numbers like 37.5->37 or 66.7->67 on the first vote)...
50 50 50 50 50 50
0 20 40 60 80 100
--- --- --- --- --- ---
25 35 45 55 65 75
And all 36 possibilities of two votes:
50 50 50 50 50 50
0 0 0 0 0 0
0 20 40 60 80 100
--- --- --- --- --- ---
17 23 30 37 43 50
50 50 50 50 50 50
20 20 20 20 20 20
0 20 40 60 80 100
--- --- --- --- --- ---
23 30 37 43 50 57
50 50 50 50 50 50
40 40 40 40 40 40
0 20 40 60 80 100
--- --- --- --- --- ---
30 37 43 50 57 63
50 50 50 50 50 50
60 60 60 60 60 60
0 20 40 60 80 100
--- --- --- --- --- ---
37 43 50 57 63 70
50 50 50 50 50 50
80 80 80 80 80 80
0 20 40 60 80 100
--- --- --- --- --- ---
43 50 57 63 70 77
50 50 50 50 50 50
100 100 100 100 100 100
0 20 40 60 80 100
--- --- --- --- --- ---
50 57 63 70 77 83
Pretty surprising... If one (human) vote has been cast, the rating/score
always ends in a 5. If two (human) votes have been cast, then it always ends
in a 0, 3, or 7. Not that it really matters, but it's kinda fun to note in a
geeky sort of way.
I haven't run any scripts to enumerate the 216 possibilities of three votes,
but since all the sums would be divided by 4, then probably all the scores
would end in 0, 5, 3 (rounded up from 2.5), or 8 (rounded up from 7.5).
The numbers don't get _truly_ interesting until there are 9 votes. :)
--Todd
|
|
Message is in Reply To:
 | | Re: Article scoring
|
| (...) OK, looks like this is going to work great! On the screen, it's looking much cleaner than what I had before with the drop-down list. A not-too-big-yet-not-to-small number of radio buttons with a neat numeric spread is 6, which yields the (...) (25 years ago, 13-Mar-00, to lugnet.admin.general)
|
52 Messages in This Thread:
 
 
- Entire Thread on One Page:
- Nested:
All | Brief | Compact | Dots
Linear:
All | Brief | Compact
|
|
|
|
