IMDb

What it is

IMDb — short for Internet Movie Database — began in 1990 as a set of fan-maintained film lists on Usenet and has operated as a subsidiary of Amazon since 1998. It is now the most widely referenced public database for film, television, and streaming entertainment, cataloguing titles, cast and crew credits, episode guides, trivia, and release data. Three participation surfaces matter for voting campaigns: the 1-10 star rating that any registered user can submit on any title, the IMDbPro STARmeter ranking that orders actors and industry professionals by weekly popularity derived from page-view activity, and IMDb Polls, a community feature where users create and answer multiple-choice polls on film and TV topics.

The score shown on a title page is not a simple average. IMDb publishes a weighted average rating, meaning individual votes are filtered and re-weighted by undisclosed criteria before the displayed figure is calculated — a design decision aimed squarely at vote manipulation.^[1]

In the context of online contests

Fan communities treat IMDb scores as a public scoreboard, which produces several recurring campaign formats:

Rating drives: Organized fan bases coordinate to submit high ratings during a release window, hoping to lift a film or series up user-facing charts and recommendation surfaces.
Counter-campaigns (“review bombing”): The mirror image — coordinated low ratings directed at a title as protest. IMDb’s weighting system was hardened largely in response to these events.
Top 250 placement pushes: A spot on IMDb’s Top 250 chart is treated by fans as a permanent trophy, prompting long-running campaigns to keep a favorite title above the cutoff.^[2]
IMDb Poll campaigns: Community polls comparing actors, films, or franchises function as informal popularity votes, and fandoms share poll links to swing results.
STARmeter pushes: Because STARmeter reflects page views rather than ballots, fan campaigns for a performer take the form of coordinated profile visiting rather than rating submission.

Unlike a typical online contest, most IMDb campaigns have no closing date and no organizer — the prize is standing in a permanently visible public metric.

Voting mechanics

Rating submission follows strict per-account rules:

One rating per account per title: Each registered IMDb account can rate a given title once, on an integer scale from 1 to 10. The rating can be changed or deleted at any time, and the latest value replaces the previous one.
Weighted display score: The headline figure is a weighted average. The ratings breakdown page additionally shows the unweighted arithmetic mean, the median, and distribution histograms split by demographic group — which makes manipulation patterns publicly visible to anyone who looks.
Chart eligibility rules: The Top 250 chart counts only votes from regular voters — accounts with consistent rating histories — and requires a minimum ratings threshold, so a surge of fresh accounts has little chart effect.
Polls: Each account selects one answer per poll, and running totals become visible after voting.
STARmeter: No vote is cast at all. Rankings are computed from page-view data, so traffic volume is the de facto voting mechanism.

Anti-fraud signals

IMDb is unusually explicit about the fact that it defends its ratings, and unusually secretive about how. The Ratings FAQ states that alternate weighting is applied when unusual voting activity is detected, and that the calculation method is deliberately undisclosed so it cannot be reverse-engineered.^[1] Signals that are publicly observable or widely documented include:

Account history weighting: Ratings from new or sparsely used accounts carry less influence than ratings from long-standing accounts with broad voting histories — the same principle described under account aging.
Distribution transparency: Because the histogram of 1-10 votes is public, a wall of identical maximum scores arriving in a narrow window is immediately visible on the breakdown page.
Velocity checks: Bursts of ratings arriving faster than organic viewing patterns support are a classic trigger for re-weighting, comparable to rate limiting on social platforms.
Network-level duplicate detection: Clusters of registrations or votes sharing one IP address range reduce the weight of the associated ratings.

For marketers

For studios, talent teams, and promoters, IMDb metrics behave differently from contest votes elsewhere:

The weighted mean discounts low-quality input: Raw rating volume means little if the submitting accounts are new, empty, or clustered. Effective contribution depends almost entirely on account quality.
Scores are permanent assets: There is no contest deadline. A rating profile compounds over years and continues shaping perception long after a release window closes.
Pacing beats spikes: Gradual rating accumulation that mirrors an organic viewing curve survives re-weighting far better than a single-day surge.
STARmeter is a traffic game: Campaigns for performers should target sustained page-view growth rather than rating submissions, since the metric is computed from visits. Services such as IMDb vote delivery are evaluated on exactly these dimensions — account age, pacing, and distribution realism.