IMDb Top Rated Movie of the Year

@bderoes​ Thanks a great list, thanks! 

@bderoes Nice work on that list, thanks.

It looks like you used a slightly different method than I did. My primary source was the IMDb Top 250 itself - I simply opened the page and used CTRL-F to find the highest ranking movie of every year. I only resorted to the advanced search in cases where none of the films from that year cracked the top 250. For those cases, I filtered the search to the IMDb Top 1000, sorted them by rating and in the case of ties, went with the title with the largest amount of votes. It's not perfect, but it's closer to what the Top 250 weighted average formula would have come up with.

Using the same methodology today, there were a few upsets, but not as many as I expected. Inglourious Basterds surpassed 3 Idiots, Amadeus surpassed Once Upon a Time in America, Judgment at Nuremberg surpassed Yojimbo, and some of the pre-1970 films (including Butch Cassidy and the Sundance Kid) dropped off of the Top 250, changing the calculations. I didn't get around to checking further back than that.

The 25k vote requirement you brought up is an issue I hadn't considered before. IMDb might have to make some slight alterations to the Top 250 formula in order for this idea to work, but it seems like that wouldn't be too hard to accomplish.

@Col_Needham​ Thanks. It took hourzzzzzzzzzz to compile. Partly because of the comments created.

I just added 2 more links in the description to filter by primary language. I wish that were one of the filters available on all Lists. I'd be very interested to see the full cluster of p-lang's on this list and on the Top 250 curated by IMDb.

BTW, it's a shame that you stopped supporting the ! (not) operator in front of the parameter name. One of my links has a list of languages, and to see the list of NOT those, the user needs to put a ! before each language code.

The fact that all 4 films added since the 2019 list, and even more of the replacements, are from India or other countries in that region, clearly indicates IMDb has made terrific headway there. Or, did author @timothy_gray_el34lojg1aih1 deliberately exclude such films in his 2019 list? I am certainly conflicted about the 2023 list as it stands today. Especially given the number of French, Italian, Japanese films on the list that are considered classics, I wouldn't want to create an English-only list, but the inclusion of so many Indian films...

For instance, looking at the entries on the 2023 list that are NOT on the IMDb Top 250: https://www.imdb.com/search/title/?groups=!top_250&lists=ls545877780&sort=release_date,desc

777 Charlie (2022) avg 8.7 (40K) is not there, but Top Gun: Maverick (2022) avg 8.2 (703K) is. (BTW, Top Gun is the SEVENTH highest rated movie with 25K+ ratings that year. All others are from India. So IMDb is doing something more in its Top 250 eligibility that screens out such films. The bottom of the Top 250 pages mentions "value of ratings received from regular users". When Indian users have been on board for more years...?

For now, I'm going to add more footnotes to the 2023 list.

Just for the record, selecting the "Top 1000" group on ATS only creates a list of 1000 films, so it doesn't have anything to do with the Top 1000 voters who seem to have extra influence on rankings (perhaps the "regular users" mentioned on the Top 250 page?).

I added footnotes to each film regarding its Top 250 rank, or the film(s) from that year which appear instead. 

@timothy_gray_el34lojg1aih1 

I'm amazed that you remember your method from 5 years ago. I always try to put that in the description, for myself and for others.

@bderoes​ I know what the Top 1000 filter does, I was just using that as an easy way to filter out films without enough votes to have accurate ratings. What I really wanted was to be able to see the rankings below #250, so I figured using that filter was the next best thing.

Actually, I did the Top 1000 research for me.

That post, and this one, is to document my research.. No criticism should be inferred from any of this.

Hopefully this doesn't interfere with the nature of this thread.

I still _don't_ really know what Top 1000 does.

Is it a coincidence that the lowest vote count is 25K? Or does Top 1000 actually mimic Top 250 in using the Top 250 criteria?

To answer my own question, first I compared the Top 250 list with the ATS with only criteria Top 250, and those lists matched.

I had to sort both by something other than ranking, since that's not an ATS sort. I chose user rating since it should come close to the ranking.

But then using ATS with only criteria Top 1000, again sorted by user rating, the Top 1000 inserts non-Top 250 titles as soon as item 21. (BTW, these are titles on the 2023 list I created.) I kept scanning the Top 1000 ATS through about 60 titles, and non-Top-250 titles continue to be interspersed.

(When I select Top 1000 but _not_ Top 250, I get 750 results. The lowest vote count is 25K.)

So the answer to my question is: no, the Top 1000 does not follow exactly the same criteria as the Top 250. Which means I still don't know what the Top 1000 represents.

I looked in the Help Center, but did not find satisfaction.

I reverted 2 of the titles on the 2023 list to the same member on the 2019 list: 1983 Jedi and 1996 Trainspotting. I did that because in both cases the rating tied with the replacement, but Jedi and Trainspotting ranked higher on Top 250 than their replacements.

Notes to self:

Even when I sorted the Top 250, the replacements came out higher (as they had done in the custom ATS used to compile the 2023 list). But I'm not certain how IMDb orders titles that appear to tie in rating. Do they use decimal places beyond what's displayed for the rating? Do they use the tt constant? (Fargo tt0116282 rank#173 but higher on rating sort, vs Trainspotting tt0117951 rank#170)?

We know that the ranking is not necessarily the user rating we see on the Top 250 page, since it matches the rating on the title's page. (On the prior version of Top 250, they used to display the ranking version of the rating, which often did NOT match the title page's rating.)