AMF All-Time Film Project: List Submissions

[BleuPanda]

It's time for the final vote!

Please rank all of the movies you have watched from the final 256. Submit your excel file to me at christophermames@yahoo.com . Don't post your list on the forum, or at least wait until after the results are revealed.

Please don't change anything in the file outside of the 'your rank' and 'comments' columns. The file can be downloaded here:
https://www.dropbox.com/s/l17aqpk84ljvc ... .xlsx?dl=0

The deadline is March 15th at midnight CST.

[BleuPanda]

After doing all this I completely overlooked the fact that I had recently added Harold and Maude to my all-time top 10...whoops.

[Maschine_Man]

I wasn't fussed when I realized that I missed the deadline to get in a few films that had been missed. Then I had a real good look at the list, bah humbug, why didn't I do it before Christmas! the films in my Top 10 that I'm surprised aren't on the list; All About Eve (#2) Rocky Horror (#4) Pan's Labyrinth (#5) and, if I included shorts, A Trip to the Moon. Next time.

That being said I really like what did make the list! I have seen about 150 of the films and many of them are in my Top 400. The films I haven't seen are ones that I have been procrastinating about watching, Kurosawa is the biggest gap in my film knowledge that I hope to close.

[BleuPanda]

If I counted correctly, I have right around ~40 I still need to see. Hope to take a chunk out of that before the deadline.

[BleuPanda]

Since this is a bigger list, I think some changes need to be made to the formula. Since this is an all-time list, I think there needs to be a less steep scale; the difference between my #1 and #2 is a lot closer, etc. If we keep the current scale, our individual ranks after a certain point become insignificant, and I'd like to minimize that.

Our old formula is 2100/(20+rank). I'm considering changing it to something like 3100/(30+rank) or 4100/(40+rank). Any opinions on what's going too far?

With those numbers, to get a sense of the breakdown:
2100/(20+rank):
#100 is worth 17.5% of #1
#200 is worth 9.55% of #1

3100/(30+ rank):
#100 is worth 23.85% of #1
#200 is worth 13.48% of #1

4100/(40+rank):
#100 is worth 29.29% of #1
#200 is worth 17.08% of #1


On the higher end of the list:
at 20: #8 is worth 75% of #1
at 40: #14 is worth 75% of #1


I feel like using a lower number puts way too much emphasis on your top few spots, when all of us are trying to support a much larger selection. A bigger number in the formula should also result in films doing better if they manage to do reasonably well with most people opposed to some films doing really well based on a single voter's positive opinion going against everyone else's.

[whuntva]

The spreadsheet seemed a bit complicated. I got the hang of it. I am watching a few older films and getting caught up.

55 of my top 100 is present, though. However, I'm having trouble deciding how to rank the rest of those films.

[BleuPanda]

I have already received lists from three other people:
whuntva
Gillingham
bootsy
BleuPanda


I'm of course planning to watch as many movies as I can before the deadline, which is still over a month away. I have 38 to go.

Also, would anyone mind a rule where a movie needs at least a single vote to be counted? Just in case there's some obscure piece that got in through people who aren't here anymore? Hopefully everything here has at least one fan, but you never know. It would be sort of ridiculous if a film ranks higher than others while going completely unwatched, even if it did qualify.

[Nick]

Bleu, I intend to submit a list, but not until very close to the deadline, as there are so many movies that I wish to watch.

I am at over 100 movies watched from the list though!

[Nick]

Also, I've got a question, Bleu. Looking at 2012's all-time film poll (which seems to be the last time the AMF did an all-time film poll), the final presentation had comments from the users posted with the results. Will you be doing this?

[Gillingham]

Also, would anyone mind a rule where a movie needs at least a single vote to be counted? Just in case there's some obscure piece that got in through people who aren't here anymore? Hopefully everything here has at least one fan, but you never know. It would be sort of ridiculous if a film ranks higher than others while going completely unwatched, even if it did qualify.

That seems perfectly reasonable to me. Though I would be curious to know what films didn't get a single vote in the end.

[whuntva]

It would be sort of ridiculous if a film ranks higher than others while going completely unwatched, even if it did qualify.

This does not seem like it should be mathematically possible. I mean, no votes is zero points, and everything else would at least have SOMETHING.

But, I would say yes to the one vote rule and maybe include the unwatched films as a bonus.

[Petri]

Also, would anyone mind a rule where a movie needs at least a single vote to be counted? Just in case there's some obscure piece that got in through people who aren't here anymore? Hopefully everything here has at least one fan, but you never know. It would be sort of ridiculous if a film ranks higher than others while going completely unwatched, even if it did qualify.

That seems perfectly reasonable to me. Though I would be curious to know what films didn't get a single vote in the end.

I second this (although I've seen 246 of the finalists so if everyone who nominated their three wildcard movies will vote in the final I think every movie will have at least one vote. I'm planning to rewatch some movies before voting (at least Element of Crime (seen it in 2003) and Yi Yi (2001 or 2002) and probably some more) so I will submit my list in the end of February.

[BleuPanda]

It would be sort of ridiculous if a film ranks higher than others while going completely unwatched, even if it did qualify.

This does not seem like it should be mathematically possible. I mean, no votes is zero points, and everything else would at least have SOMETHING.

But, I would say yes to the one vote rule and maybe include the unwatched films as a bonus.

Actually, no votes is not zero points; these polls are usually run where an unwatched film nets a person's average rank. It's a method of balancing out less watched films so the top of the list isn't exclusive to films everyone happens to have seen. It's the problem where giving a film no points simply because you haven't seen it biases the results towards popularity, which is what I'm avoiding compared to the previous all-time lists this forum has done. Using this method also allows us to actually hurt the films we have seen and didn't like as much, whereas a straightforward poll can only accommodate positive opinion.

Of course this method breaks down when you go to lower numbers of votes; I'd even propose a single vote is too little to actually mean anything in what is supposed to be a poll representing the entire forum, especially since that single vote will likely rocket it past 50+ films we've all seen but simply put on a slightly lower pedestal than everything else. Every film (except those added by individual users) has already been vetted by this forum, so there's less need to balance out the popular-but-not-as-good versus the obscure-yet-fantastic. Unlike the decade polls, it's very likely a film most of us appreciate will end up in last place, because this formula now greatly favors loving a movie over liking it, and the lower ranks naturally evolve towards films everyone has seen but put on the back end of their list.

[Nick]

It would be sort of ridiculous if a film ranks higher than others while going completely unwatched, even if it did qualify.

This does not seem like it should be mathematically possible. I mean, no votes is zero points, and everything else would at least have SOMETHING.

But, I would say yes to the one vote rule and maybe include the unwatched films as a bonus.

Actually, no votes is not zero points; these polls are usually run where an unwatched film nets a person's average rank. It's a method of balancing out less watched films so the top of the list isn't exclusive to films everyone happens to have seen. It's the problem where giving a film no points simply because you haven't seen it biases the results towards popularity, which is what I'm avoiding compared to the previous all-time lists this forum has done. Using this method also allows us to actually hurt the films we have seen and didn't like as much, whereas a straightforward poll can only accommodate positive opinion.

I was thinking about this poll today, Bleupanda, specifically about how an unwatched film is rated.

Currently we have it so that an unwatched film is rated at the average rank of the user who did not watch it. But is this really fair? I suppose that putting it at the average rank comes from the assumption that there's a 50% chance that if the person were to watch it, it would go higher than the average rank, and a 50% chance that it would go lower. But is this really the proper assumption to make? If anything, I would say that the chance that any given unwatched movie ranks within the upper 50th percentile on someone's list is actually far less than 50%. This is due to the fact that a person is more likely to watch the movies that appear interesting (and therefore, net higher positions in their list) and avoid the ones that do not appear interesting (and would most likely net lower positions on their list).

For example, a fan of westerns who hates romantic comedies would be more inclined to seek out any unwatched westerns, while avoiding any unwatched romantic comedies. Therefore, it's not really fair to say that the unwatched romantic comedies would have a 50% shot at reaching the upper 50th percentile on this hypothetical movie-watchers list, is it? While we may not all hold the same genre prejudices as this hypothetical movie-watcher, we still do tend to "judge the movie by its cover", and gravitate towards the movies we find interesting.

I propose that a more fair way to rank the unwatched movies would be to put them at, say, 75% or 80% of the total movies a user has watched. This is just a pretty rough figure, but I think it's a bit more fair than the 50% figure.

Thoughts?

[BleuPanda]

That's also my general opinion, but this is what other users on this forum do. At the same time, there are plenty of films I intend to watch but simply haven't had the time to, and switching to the 75th percentile would give a whole lot more power to people who have seen less films. So, while there are films I actively avoid watching, there are also plenty that I know deserve more credit than my non-ranking. Even with the score at the 50th percentile, a film will struggle to reach the top ranks if most of us haven't seen it.

[Petri]

There have been several examples in the decade final polls in which a movie would have been probably higher if more people had seen it. For example A Woman Under Influence would have been in top 20 of its decade if I had seen it when we had 70s poll (luckily it was wild card nominated to the superfinal).

I agree that if you haven't seen a vast majority of the movies in the final it's more likely that you have watched and will watch movies you think you going to like before the dead line. (Of course there's a chance that you haven't seen a single french new wave, italian neorealism, japanese jidaigeki or german expressionism movies and would love those genres a lot .)

But I guess among people who have seen or tried to watch practically everything (+90% of the finalists). They just haven't had a chance to watch some movies because those movies haven't been available for good quality and reasonable price (and maybe with subtitles). So I would give those unwathced movies the score of median rank like we have done before (they still can't win). Bigger problem in my opinion is that people who have seen less movies have kind of bigger impact to the results (the sum of their points is bigger + the average of the differences in points is bigger). So maybe there should be a coefficient like (new points = old points* sqrt(wathced movies by a person/all the movies)) or something like that. So if you have seen 50 % of the movies every movie (you vote for + unwatched ones) would get 70,7 % of its maximum score and so on. (or maybe just coefficient = (the sum of points if you have seen all the movies in the final)/ (sum of points by voter)

[Petri]

There should be a statisticial pattern that would predict how much you liked the movies you haven't seen (and rank them). It would depent on genre (every movie should have a main genre and maybe subgenre) and the other voters (how their tastes compares to your taste among the movies you both have seen) at least (and country). Maybe some day we we don't need to give every unwatched movie a similar score.

[Nick]

You've made some good points, guys. I certainly wouldn't want my input (as someone who has seen a relatively low number of films) counting more than someone who has seen a high number of films. It just sucks knowing that by watching some movies that I end up really liking, they end up being hurt in the long run, as they place lower than the 50% mark on my list.

[BleuPanda]

I've sometimes countered that by watching the movies I know I won't like...which is admittedly counter to what this type of list should represent.

[Nick]

I've sometimes countered that by watching the movies I know I won't like...which is admittedly counter to what this type of list should represent.

Actually, I've found myself doing the same thing. And then I take a perverse pleasure in finding out that I'm not really all that into the film. Or I take a perverse annoyance in finding out that I AM really that into the film, meaning I'll have to bump down something I like slightly less.

Ultimately, it's your poll, and you choose the way to run it. But I don't really think that we should go with the 50% rule if our biggest justification is "that's how we've always done it".

[BleuPanda]

Well, one big reason I support the 50% rule is because these films are already vetted by the forum. This isn't a random collection of films but ones that all deserve their placement here (the individual favorites might not fit that, but oh well). I have to enter this with the mindset that all of these films are good enough that they could end up on either side of someone's list. If a film really doesn't deserve its placement, I think that should be decided by those who have actually experienced it.

[Petri]

I've sometimes countered that by watching the movies I know I won't like...which is admittedly counter to what this type of list should represent.

Actually, I've found myself doing the same thing.

Yeah same thing with me (I watched Mission Impossible Ghost Protocal and Dawn of the Planet of Apes (and even Rise of the Planet of Apes before that to give a fair chance for the sequal) just before the deadline of 2010s poll. They were entertaining but not in the same league as the most of the movies in the final (imo) and ended into the bottom quarter like I had expected. (But I still think the ones I didn't see could have been in either half of my list (I just hadn't chance to see them (I have seen It Follows after the poll and it would have somewhere in the middle I guess).

Well, one big reason I support the 50% rule is because these films are already vetted by the forum. This isn't a random collection of films but ones that all deserve their placement here (the individual favorites might not fit that, but oh well). I have to enter this with the mindset that all of these films are good enough that they could end up on either side of someone's list. If a film really doesn't deserve its placement, I think that should be decided by those who have actually experienced it.

I agree.

Also, would anyone mind a rule where a movie needs at least a single vote to be counted?

Maybe we could raise this to two or even three votes.

[BleuPanda]

Because I have a self-imposed writing deadline of March 11, I'm not going to actually work on finalizing the results until after that day. As such, the new deadline to submit your list is going to be March 15!

[Maschine_Man]

I started working the comments for my films. Would it be okay to include the formats for the forum (italics and bold, nothing more)?

[Nick]

Because I have a self-imposed writing deadline of March 11, I'm not going to actually work on finalizing the results until after that day. As such, the new deadline to submit your list is going to be March 15!

More time to watch more movies? I am a-okay with this.

[Otisredding]

I sent my list.

For the next issue, I promise to see many more films from the list.

[luney6]

I've sent my list as well.

[Maschine_Man]

Just sent mine off!

[Nick]

With 125 films seen out of the 256 films on the list, I'm coming in at a hair under 50% in terms of films watched from the list. Which might not seem that much to you megafilm nerds out there, but considering the fact that I had only seen 62 of those 125 movies before the start of 2015, I consider this to be a major success. There's still some movies that I just didn't find the time for though. Like "Dekalog" for instance. I'm sure it's a fantastic film, but there's just something about ten hour long movies about the Ten Commandments that's a little off-putting. Not exactly the kind of movie where you invite a couple friends over, grab some beers, order a pizza or two, and sit down for a marathon, you know?