Ask HN: feedback on The Great Movie Experiment - an alternative to IMDB star ratings. | Hacker News Reader

Ask HN: feedback on The Great Movie Experiment - an alternative to IMDB star ratings.

greatmovieexperiment.com

24 pointsmatthias18 comments

Hi everyone, I would very much appreciate your feedback on a new site. The gist is that we're compiling a list of movies based on head to head votes, as an alternative to imdb star ratings.

The format is 1000 movies facing off in a round-robin tournament (~500k votes per round). We needed a non-arbitrary sample of movies to vote on, so we've started with a list of the Top 1000 US box-office takings. Each round, the bottom 50 will be culled to make room for the top 50 in the Up & Coming.

We're working on... 1. an 'add movie' page, allowing you to add any movie to the up&coming 2. member pages showing your own votes & reccomendations, 3. actor pages, ranking the movies people have been in by popularity.

Monetising... 1. netflix add-to-queue buttons, 2. amazon links (the site is great at reminding you of things you want to see), and feasibly 3. targetted advertising (ie. only show this ad to people who have voted for romcoms).

Make the images smaller - they tend to load slowly on slower-ish connections meaning the text/title changes but the image is of the previous movie. Cue confusion ;) Otherwise pretty good show.

Also how about doing a mobile version if you can, that would be snazzy (Iphone yeh!) :D

I'd be interested to see the rankings you get from this.

It's running from the server in the attic just now so it won't be as snappy as possible! Our intention would be to put the images on a CDN at some point. In the meantime, would a loading spinner on the graphics avoid confusion with the title text?

Yeh that is a good idea! I think that was the major problem - the the images didnt disappear immediately when I clicked.

Kudos for posting a home-hosted site to HN! thats like the pinnacle of hackerdom :D

Seems like a really weird way to compare ratings, when you have match ups of childrens movies to horror movies or something along those lines I know I for one cannot compare two movies in separate genres well as you can't base them relatively. Just an opinion of mine, if you compare lets say stuart little to the omen, one person may choose the omen because they don't like animated childrens movies, another may choose stuart little because it made them feel fuzzy or w.e, now if you put the omen against halloween or stuart little against madagascar you get a more truthful comparison.

I guess I'm saying I would like it better if you had like different genres, they don't even have to be that specific like say horror action comedy family; and could compare withing those, but I guess that is limited with your selection right now.

Agreed. Too many apple-to-orange comparisons. But categorizing by genres will require some amount of work.

How about having an option "can't compare"? That the user can indicate they saw both but cannot compare. With enough data, maybe you can cluster more frequently compared movies together or even making categories automatically?

(Downside is more complex UI; it's better to keep UI simple, so I'm not sure).

I agree that comparing movies in different categories this way is awkward.

I think the idea of getting an ordering by pairwise comparisons (instead of discrete scores for each item) is sound, but it doesn't work as well for movies as for people.

pickthehottie.com (launched at least before June 22, 2001, according to the Wayback Machine) did the same thing to AmIHotOrNot's photo-rating juggernaut (launched in October 2000, according to Wikipedia). I think the A-B model worked really well for them, and the incentive for sticking around to compare for photos was a little more intrinsic.

Consider:

What satisfaction does the user get from voting? Why vote or visit more than a few times? Why would they recommend to their friends?

I'm missing some visual feedback that a new set of movies is loading. Nothing really happens when you click -- according to Firebug it took 13.6 for the next set of movies to appear.

I'm not sure on the statistics, but maybe you'd be better off showing X movies at a time and asking the user to pick 2 of them; that might generate preferrence data faster. That way, when some of those X movies repeat, you would get more reliable ranking data. you could also ask the user to order the movie as it's common in studies, but that would potentially take too long.

Some kind of "progress" to show to the user would be nice too, e..g "you rated 5123 movies", even when not logged in.

I would prefer a Liked, Disliked, Haven't seen choice pattern where "Liked" and "Disliked" reload both pictures.

It will be less annoying than clicking on "Haven't seen it" ten times on both pictures and will provide more comparison data.

Edit:

If you don't like the "Disliked" choice, you could simply reload both pictures when clicking on "Haven't seen it".

IMDb has a group of ultradedicated movie fans and it has an excellent rankings formula. The only reason their ratings are any good is that they have more movie buffs than any other place, so the reviews and ratings actually indicate a lot of very good movies.

It would be great if an one's own ranked list could be exported straight from your website. Maybe you'd want to keep track of individual users' lists anyway, to prevent people from gaming the results?

Great idea :).

Cool idea - I would change both movies each time, though, if you haven't seen at least one of them (instead of flipping one at a time and keeping the other one up). Also, as people have already mentioned, keep them in similar genres or at least offer an option of "Can't make the comparison," or something similar.

Putting a button to discard both movies at the same time would be better in terms of user experience IMHO.

* I wanted to see more movies in the list (in Up&Coming page) and set it to 100. Now, when I hover over the links towards the bottom of the list, the visuals (such as the movie poster) are no longer in my view.

* Some description/tagline/plot of the movie along with the poster would be good

None of the movies seem to be loading.

A problem on my end or is the site unable to handle all the traffic?

The (wonderful) traffic helped us identify a bottleneck in the code. It's working again now but it's a home server so wish us luck!

Much snappier now. :)

Reply on news.ycombinator.com