Poll

What time should we have the presentation of the Filmspotting Top 30?

Monday,  Sept 1. 5pm PST
8 (50%)
Thursday, Sept 4, 2pm PST
1 (6.3%)
Thursday Sept 4, 5pm PST
4 (25%)
Friday Sept 5, 5pm PST
3 (18.8%)

Total Members Voted: 16

Author Topic: Filmspotter Top 100 2014-- Discussion  (Read 35645 times)

Bondo

  • Objectively Awesome
  • ******
  • Posts: 23082
Re: Filmspotter Top 100 2014-- Discussion
« Reply #20 on: May 21, 2014, 03:13:44 PM »
@Jake, yes, I can't imagine we'd end up on a system that disallows shorter ballots.

As a bit of summary of the history of our group list scoring:

So most group lists on the forum have used the standard system...ranked list where the top selections are weighted more heavily. Tweaking the number of films that can be submitted or the degree of weighting can shift results a bit. Arguably less weighting or longer lists would lead to more uniformity (the most broadly accepted films dominating) while higher weighting or shorter lists would lead to more unpredictability, films getting on with fewer voters. So tweaking one way or the other is possible depending on what we would want to accomplish. I'd be opposed to longer lists (>100) because that's more work for me. Dave's system is just a more flexible version of the standard system that would probably lend itself to higher weighting and shorter lists. If the openness is too complex, we could accomplish similar by having ballots be 50 films and just make the points drop-off more precipitous.

A second system we used was for the Best of the Decade (00s) where there was no weighting. In this case the goal was a bit more toward broad consensus so it was functional enough. However, one facet we built into that one was the ability to cast a negative vote for particular films to catch one's ire. This option was not widely utilized and probably didn't affect too much (I'd have to actually read through the threads). The results of an unweighted system would be easy enough to compute for comparison alongside most other systems. I believe Sight & Sound's list is unweighted in this manner...with small ballots (ten films each).

The third system we've used was the "rate all" system, used for the Documentaries list. This is basically the Ratings Project now and thus wouldn't be usefully different from what we already have.

So what completely novel (but not overly complex) process could we try out that isn't basically one of the above? I'm not sure.

MartinTeller

  • FAB
  • Objectively Awesome
  • ******
  • Posts: 17864
  • martinteller.wordpress.com
    • my movie blog
Re: Filmspotter Top 100 2014-- Discussion
« Reply #21 on: May 21, 2014, 03:27:57 PM »
I'm in favor of a non-weighted system, or a very mildly weighted one.  However, I'm also in favor of doing it the way it's always been done, for consistency's sake.

Ratings Project-style rating seems like it would be a lot of work for everyone.  We should keep it simple to encourage participation.

PeacefulAnarchy

  • Elite Member
  • ****
  • Posts: 2132
    • Criticker reviews
Re: Filmspotter Top 100 2014-- Discussion
« Reply #22 on: May 21, 2014, 03:48:08 PM »
I'd be opposed to longer lists (>100) because that's more work for me.
It doesn't have to be if you ask people to submit imdb lists, or make imdb lists yourself if you think asking people for them is too high a burden. I could help you out with that if you want. It makes tabulation a relative breeze and you can pull out a lot of imdb's metadata from the export to make for a nicer presentation.

I'd love to be able to submit a list longer than 100, even if it's not going to affect much of anything.

As for point systems, if the point system is variable (each user has x points to distribute among up to y films) then you probably want to keep things from being too complex just because otherwise it'll turn people off. If the point system is fixed, (i.e everyone's #1 gets the same points, everyone's #2 gets the same points, etc) then the system can be as complex as you want and it doesn't really affect tabulation because you're just associating ranks with points using whatever system or formula you choose.

I personally think the current system (1st=20 points, 2-5=15 points, 6-10=10 points, 11-20=8 points, 21-30=6 points, 31-40=5 points, 41-60=4 points, 61-80=2 points, 81-100=1 point) is way too top heavy. I do think a weighted system is nice to give meaning to rankings, I don't value my #1 film the same as my #100, but I also don't value it 20 times my #100 and twice my #6. An option for people who want to submit an unranked list can also be given with a weighted system for those who don't like to rank.

I have lots of opinions about better fixed point systems, but not sure it's worth typing it all out.
« Last Edit: May 21, 2014, 03:50:18 PM by PeacefulAnarchy »

oldkid

  • Objectively Awesome
  • ******
  • Posts: 19044
  • Hi there! Feed me worlds!
Re: Filmspotter Top 100 2014-- Discussion
« Reply #23 on: May 21, 2014, 08:49:09 PM »
I wonder if it's possible to give some people an option to not weight their lists.  If a person doesn't say anything, it's the standard system.  Otherwise all films get the same weight (5 points?).  Would that be fair and possible?
"It's not art unless it has the potential to be a disaster." Bansky

Bondo

  • Objectively Awesome
  • ******
  • Posts: 23082
Re: Filmspotter Top 100 2014-- Discussion
« Reply #24 on: May 21, 2014, 09:33:20 PM »
The way I enter things, there is no effort difference between standard and individualized scoring. I enter point values, not tiers. So standard, Dave's, flat, it's all the same on my end. I'm pretty content with people choosing their own weighting if single film and total list limits are in effect.

smirnoff

  • Objectively Awesome
  • ******
  • Posts: 26251
    • smirnoff's Top 100
Re: Filmspotter Top 100 2014-- Discussion
« Reply #25 on: May 23, 2014, 05:15:07 PM »
So what completely novel (but not overly complex) process could we try out that isn't basically one of the above? I'm not sure.

If you want fun & surprise nothing can beat a lottery system!



Each film mentioned on someone's top 100 adds one ball representing that film to the tumbler. Fill up the tumbler with all the balls, and draw 100 and see what you get!

A popular film, like The Godfather, will have many balls, and thus a very good chance of making the list. A film with only solo support (one ball), like Grandma's Boy, will be extremely unlikely to make the list. But it could! Anything could happen! Which is why it would be so fun and make the chat room really crazy during the drawing process! Every draw someone wins! And you could argue, that's the best part of the top 100's anyways... the unveiling! :))

Too random? I thought the same thing. A way around this, but still preserving the lottery fun, would be to allow a film to move up the ranks for any additional balls draw in it's name. So the Godfather may start out at 35th spot, but if another one of its balls was selected it could jump 10 spots (or however many spots we determine... maybe it's auto #1, only bumped by subsequent two-ball picks). Whereas a film like Grandma's Boy could never advance, and probably wouldn't make the list anyways.

You could also have it so when balls are selected they get thrown back into the tumbler after, so films preserve their popularity throughout the process, rather than see it dwindle.



Advantages of this system:

-No more laborious tabulating, which involves proofreading ballots for uniformity in formatting & misspelling. All that would need to be done with this new method is cut/paste every ballot into one long A column in excel, and then in column B just put  =rand() next to every title. Excel recalculates the =rand function every time you sort column b, so you could build a macro to have a one-button sort B function. And every time you pressed it, whatever title showed up in Row 1 would essentially be the ball you are drawing from the tumbler.

-Wildly exciting and surprising! :))

Would love to hear the feedback for this crazy idea!

Junior

  • Bert Macklin, FBI
  • Global Moderator
  • Objectively Awesome
  • ******
  • Posts: 28709
  • What's the rumpus?
    • Benefits of a Classical Education
Re: Filmspotter Top 100 2014-- Discussion
« Reply #26 on: May 23, 2014, 05:54:22 PM »
After the Celtics got screwed in the recent NBA draft lottery, I'm all for a system that will show the exciting, good things that might happen in a lottery. If this doesn't gain any support, I'd still like to see some version of it implemented in some way. Maybe a supplemental top 100. Or take out any movies that make the top 100 and make a top 100 (or 50 or whatever) from the remaining films in this way. That's kinda fun!
Check out my blog of many topics

“I’m not a quitter, Kimmy! I watched Interstellar all the way to the end!”

smirnoff

  • Objectively Awesome
  • ******
  • Posts: 26251
    • smirnoff's Top 100
Re: Filmspotter Top 100 2014-- Discussion
« Reply #27 on: May 23, 2014, 06:03:55 PM »
Participation ought to be high because it's easy (no ranking required) and everyone has a chance to celebrate. And with promotion it could maybe draw out some people to vote who've never done it before, because they just might see their favourite films get the kind of exposure they've never gotten before. :)

Jared

  • Elite Member
  • ****
  • Posts: 3492
Re: Filmspotter Top 100 2014-- Discussion
« Reply #28 on: May 23, 2014, 06:20:25 PM »
You're insane but you might also be brilliant.

I don't know about that system for the official top 100, but 100 films selected that way would be fun thing to do. I would pledge into a marathon pact to watch anything I hadn't seen if we made such a list. 

smirnoff

  • Objectively Awesome
  • ******
  • Posts: 26251
    • smirnoff's Top 100
Re: Filmspotter Top 100 2014-- Discussion
« Reply #29 on: May 23, 2014, 06:28:18 PM »
Oh snap, the stakes just got raised! :))