Lately I noticed that there were some Indian movies which appeared in IMDB top 250 list last year which are now removed. So, I read the criterion page for the movie to appear in Top 250 list. They claim that their algorithm is based on the formula which provides a true ‘Bayesian estimate‘, which takes into account the number of votes each title has received, minimum votes required to be on the list, and the mean vote for all titles:
Weighted rating (WR) = (v ÷ (v+m)) × R + (m ÷ (v+m)) × C
R = average for the movie (mean) = (Rating of the movie by people)
v = number of votes for the movie = (votes) (only votes for regular users are counted and who constitutes a regular user is a secret)
m = minimum votes required to be listed in the Top 250 (currently 25,000)
C = the mean vote across the whole report (currently the value is 7 as per Wikipedia Bayesian estimator page)
I tried to plot the above formula in a 3D graph to observe how the rating and votes effects the chances of a movie appearing in this list. I rearranged the formula to make it intuitive as below:
Weighted rating (WR) = R + (7 – R) / (1 + y)
R = average for the movie (mean) = (Rating of the movie by people) (Range 1 to 10)
y = v/m i.e. the multiple of the minimum number of the votes (Range 0 to 10 i.e. from 0 to 250k regular votes)
So, when I plotted the graph of WR against two variables R as x and y, I found the following result
Using tool from www.math.uri.edu
Side view using tool from www.livephysics.com
So, as you can see, the weighted rating (WR) would approach C (mean overall) when the value of y is small while the WR would start approaching the value R (i.e. Z = X) after the value of y crosses 4 i.e. the number of regular votes starts crossing 100k which happens very rarely for any foreign movie.
Why Indian movies are missing
So, the movie having more rating (R) and more votes (v) would most likely be in the top 250 list, if the formula is followed honestly without bias. To filter out the movies of foreign origin, there are two ways as per the formula above. It could be done through changing the value of m as foreign movies naturally have lower value of m and thus such movies would stop appearing in the list as they won’t cross the threshold value. Or it could be done by not considering voters of certain geographies under the value of v and thus reducing the ratio y and thus decreasing the value of WR.
Last year, there were certain movies in the list which are not there this year. Assuming no modification or manipulation in algorithm, to reduce the value of WR there needs to have a consistent voting of low rating which generally doesn’t happen as people suddenly doesn’t start liking or disliking the movie. Instead the number of votes will increase which will reduce the value of y and that would further push up the movie in the list. So, it seems that company has made some changes on their algorithm side which is reflected in filtering of Indian movies.
For example, movie Dil Chahta Hai with Rating (R) of 8.2 and total number of votes being 49.5k and Udaan with 8.3 rating and 31k votes is in the list but Swades with higher rating of 8.3 and 60.5k isn’t in the list, which is clearly against the formula. Similar is the case with Bahubali with rating of 8.4 and 66k votes which is also not in the list. The list of such movies is long but some of those movies could be PK, Vishwaroopam, Lagaan, Gangs of Wasseypur, Barfi! etc. At present, the movies that are in the list are TZP, 3 Idiots, Andaz Apna Apna, A Wednesday, Hera Pheri, Sholay, Munna Bhai MBBS, RDB, Drishyam, Udaan, Chak de! India, Queen and DCH.
The company claim that their voting system is fair but it seems based on the above analysis that the system isn’t trustworthy and credible but is strongly biased against foreign movies. Also, if the status quo is maintained then in future they would further raise the value of these parameters to filter more foreign movies. It is fruitless for the company to think that they would manipulate the system and nobody would know.
In my personal experience, based on some top Indian and American movies that I have watched, I didn’t find that American movies are generally superior or if big budget movie make a movie better. There could be no better example than Indian Mars mission which we achieved in first attempt itself and in one tenth the cost! What matters is the content of the movie but not the budget or publicity. Though I appreciate the fact that certain graphic movies produced in US are better and innovative with the example of movies like Avatar but in general movies are liked or disliked for their content and not for the money spent.
Perhaps, we Indians have gradually developed a kind of inferiority complex where we consider everything foreign as superior and don’t appreciate what we do or have. And due to this we seek to get acknowledgement for everything we do or have from foreigners as the trust we place on them is greater than the trust we’ve on ourselves. Perhaps, this has happened due to India’s low economy for long time and due to psychology where a person feels that by imitating people whom they consider as superior, they would also become superior. We drifted away from our roots and started this new notion where we consider wealthy people as superior without regard to the way that wealth is accumulated. In ancient times, the philosophers and thinkers were held in highest esteem which doesn’t seems the case today in our society.
If the above analysis is correct then it also shows that the people running the website are racist, untrustworthy and unfair. Thia also shows some very interesting characteristic about their psychology which forced them to tweak the list. It shows their fear of losing the top position, soft power and the psychological domination that they had for long. It is only recently that Indians have started using the IMDB site which resulted in Indian movies started listing in top 250 otherwise we’ve a long list of movies that are worth watching by everyone.
Every country makes stupid movies by spending a lot of money and by casting leading actors but those movies are just for entertainment. India also has the same case that mainstream movies are just stupid and are just for entertainment value while the real movies in Parallel Cinema aren’t much watched because of strong hold of rich production houses on multiplexes and censorship board. There are much greater number of Indian movies which have been watched by less people but could easily fetch good score like Apu trilogy etc.
Finally regarding IMDB, again if the above analysis is correct and they want to atone and absolve themselves for their manipulative actions and want people to consider them as fair then they must make the whole thing transparent and reduce the minimum number of votes required for qualification and consider all votes with equality to get rid of bias.