I developed an extensive spreadsheet in determining the best sci-films of each year. I spent weeks hammering out a formula that takes all areas of praise in consideration - viewer ratings, critic ratings, accolades from award organizations, and established All Time lists.
These aspects of praise were not necessarily all given equal value in the equation. For example, All Time lists such as AFI and Sight & Sound, hold a certain amount of value higher than certain other values. The All Time lists essentially represent countless hours of research by these organizations who created their own lists and warrant a sizable value factor in the equation.
Also, awards-won was appointed a much higher value than awards-nominated. Naturally, newer films may have a much higher collection of awards simply because a number of accolade organizations didn't exist for the older films, so a cut-off value was set. In addition, only the more recognized, national, international, and regional awards were applied such as the Oscars, Golden Globe, Saturn Awards, and others.
I disregarded awards given by every little town who happens to have a film social group that announces awards. Another award type disregarded is DVD releases, many of these are based on the quality of the DVD produced, not so much the movie itself. The same for Film Trailer awards, which only merits the production of a film's trailer.
Other factors were taken into consideration such as the number of votes. For example, a given film may be IMDB rated 8.0 by 1,200 IMDB voters. Another given film may be IMDB rated 7.5 by 800,000 IMDB voters. I just couldn't ignore the wide difference in voter input between those two films, so certain values (with minimum and maximum cut-offs) were applied. For RottenTomatoes, rather than using their "percentage liked" value, I used their actual rating of a given film, both critic and site visitor input.