Subject:
|
Re: Still missing ballots from 5 people
|
Newsgroups:
|
lugnet.build.contests
|
Date:
|
Thu, 28 Dec 2000 13:33:04 GMT
|
Viewed:
|
1130 times
|
| |
|
|
In lugnet.build.contests, Todd Lehman writes:
>
> A completely alternative solution -- one which presents neither of the two
> problems just mentioned -- is simply to _subtract_ 3 points from entries where
> the entrant failed to vote. That way, an entry receiving, say, 8 votes, would
> still end up with 8 - 3 = 5 points if its entrant didn't vote, and thus it
> could still win if no other entry in its category could beat it under the -3
> handicap.
I suppose it depends on which "election" your considering. For
each category, if someone is trying to improve their own chances,
but still votes, then they'll only cast the 1 mandatory vote
in that category (because you can't vote for your own entries).
So by not voting, then perhaps only 1 vote should be removed
for that entry in that category.
If they have more than one entry in category, then 1 could be
removed per entry, but taken from the total for the entry with
the most votes. You have to do this point-by-point. If not
then if you take 10 points away from one single entry (in a category
with 10 entries by that person) then another of their designs may
come out in front. You'd have to level to vote total like a loose
pile of 1x1s, not like stacks of bricks :] The wily competitor
will likely vote-split anyhow, not putting all the mandatory votes
in that category on the same entry. They'll pick the ones they
think won't win and spread votes as evenly as possible across
those. So this method would seem to work.
The problem may come in the overall "election" of the contest-wide
winner. Again the competitor wanting to leverage his chances
will vote-split, scattering them across the categories on entries
that they hope aren't the strongest contenders (and thus won't
affect the outcome). Okay, maybe there is no problem. In this
case I guess you could still strip away the votes on a point-by-
point basis, but across all of that person's entries. Here the
loose pile of 1x1s is all of that competitor's entries. Yeah,
this could work too.
If someone votes, but doesn't use all of them, then you only
whittle away the missing points, again on a point-by-point
consideration. Figuring my bable doesn't make sense, try this:
1) Are there missing votes (left)?
2a) No. No action required. Goto to 6.
2b) Yes, go to 3.
3) Find entry with highest vote tally for this contestent.
4) Subtract 1 point from this entry.
5) Go to 1.
6) End.
Please don't pick on my nested GOTOs, I know they're bad form.
But what the hay, they worked fine for the F-18 simulators
I worked on for the air force ;]
Anyhoo, a problem I see is when their are tied entries in step
3. Do you then take away the point from the entry in the
closest race with a competitor?? What if they're all tied,
then does it matter? What about 3-way and multi-way fights
in a category, etc., etc...
Todd, the 3-from-each-entry idea may be a little lopsided
when all categories/entries are considered. Maybe the
1-from-each done point-by-point would have better results?
But as for these last problems, I dunno. Maybe someone else
can figure these out, I've already wrecked my brain for
today (and it's only 09:30!).
I suppose what we're seeing here is the inherent problem
with "Competitors' Choice" voting: each voter has a vested
interest in the outcome, should they desire to exploit it.
KDJ
P.S., after reviewing this post, I'm not so certain about
my 1-by-1 idea now. It "targets" the popular entries,
so maybe your 3-form-each idea is the best Todd. I gotta
dwell on it some more. Maybe others have another idea?
_______________________________________
LUGNETer #203, Windsor, Ontario, Canada
|
|
Message has 2 Replies:
 | | Re: Still missing ballots from 5 people
|
| (...) No, the wily competitor would mail a small pile of landing plates to me. Then I'd favor his design with 20+ votes! Though it is interesting to take your idea to its conclusion. 2 wily competitors could end up causing the least likely entries (...) (24 years ago, 28-Dec-00, to lugnet.build.contests)
| | | Re: Still missing ballots from 5 people
|
| (...) Ya, you snipped the part about for each vote point you have learned, you either add 1 point to someone else's or subtract 1 point from your own (i.e., by not voting). It wouldn't go -3 per se, but -1 as many times and in as many places as (...) (24 years ago, 28-Dec-00, to lugnet.build.contests)
|
Message is in Reply To:
34 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
|
|
|
|
