Bambi and top 5's

[lupercal]

I notice that, for the first time, Bambi and Bambi II are holding down the top spots in the Features and DTV categories respectively

(and FWIW, Balto and Balto II are holding down the 'most reviewed' spot in those same categories.)

I remember Balto I and III once holding down most popular in both categories as well. Seems that the more votes a popular film gets, the lower its average is inclined to fall (remember The Incredibles being #1, or any other films no longer in the top 5?)

It would be interesting to know what imdb's algorithm is for ranking films - because there are films with smaller number of votes which don't rank in the top 250, even though they technically outrank the #1. Our only condition (that I know of) is a minimum number of votes, but I can't help thinking that a film with 18 votes and an average of 3.6 is possibly as impressive as a film with 5 votes and an average of 3.9

And of course we all know that the fewer votes a film has, the more effect a new rating has - whereas a film with lots of ratings takes a relatively incrimental hit.

None of this is accusatory or critical. I do sort of wish that there were a way of taking all this into account, but I must admit, I don't know one.

Loop

(and in the interests of disclosure, I'd sooner see Bambi at #1 than Toy Story 2, though I gave both of them 4 stars)

[Toonboy]

Short-lived. Our newcomer is reviewing up a storm of Pixar movies, and Toy Story is now the number one movie. Previously it was Toy Story 2.

[GrafSpee]

In the "eep!" thread (in Theatrical Shorts), athena said that the ratings were based on:

a) some minimum number of reviews (usually 5 but could be as low as 3 for some categories)

b) the average of the star ratings for the film. In the case of ties, the film with the most reviews has priority.

Without some type of weighted average, the good films with the smaller amount of reviews will almost always rank higher.

[lupercal]

In the "eep!" thread (in Theatrical Shorts), athena said that the ratings were based on:

a) some minimum number of reviews (usually 5 but could be as low as 3 for some categories)

b) the average of the star ratings for the film. In the case of ties, the film with the most reviews has priority.



Yes, but if I understand that correctly, it would ONLY apply in situations where films were tied. I don't believe there is a system which weighs the relative number of reviews OUTSIDE of a tied score scenario.

I'd be surprised if it were true, because to create
anything like a fair system would be very, very difficult. I suspect imdb's system is very complex; plus they have thousands of votes to play with, and I don't even think that their system is fair. Better than nothing, but nowhere near perfect.

The other side of the coin, re. your argument, is that that a little known film can have a perfect rating, but not enough reviews to get into 'the charts'. What do you do? It might be the greatest thing ever made, but it may have been made in Turkey in 1979, and only ever available on laserdisc.

I guess it's less of a problem that a widely reviewed film tends to eventually drop. And I suppose the overall problem is not enough reviews.

Loop