Apologies in advance for the long post but I wanted to get into the nitty-gritty of this problem. And I have included what I believe will fix the problem.
The problem is that it is impossible to tell if someone has watched it or not so it is not a good idea IMHO to make assumptions.
For instance:
-
Trakt as much as it hurts me to say this is… not perfect. Many times due to TVDB or TMDB having wrong dates or no dates at all and when you do update it it does take time for Trakt to update. (Have had problems in the past with air dates not updating for quite some time but that is a separate topic for now.)
-
Depending on the theater they often let their employees watch the new movies before they come out to avoid being spoiled by the audience.
-
There are theaters that screen blockbuster movies way in advance of the official release date. (There is a local theater that I know of that consistently does this.)
-
Early test screenings by the production companies.
-
The Tenet example, happens quite often and sometimes there is no release date for the first airing on TMDB/TVDB anyway.
-
Also this happens quite often with TV shows because TVDB requires the release date of the country that the production company runs out of, even if it was aired in another country first. Take for instance Victoria Season 3 the show aired it’s third season in the US on PBS months before it aired in the UK on ITV. Victoria (British TV series) - Season 3 - Wikipedia
But as it was created in the UK that is the dates that TVDB went with. (Was a whole discussion over at TVDB, nothing changed. Please lock season 3 of Victoria - TVDB Forums ) -
Not endorsing but movies do also get leaked onto pirate sites before wide release.
While not many people do watch movies before it is released mainstream it does happen more than you might think.
As I have mentioned before on the old forum the amount of people actually rating before something is released does not really affect the average on movies. Usually ends up being less than 0.5% of the total votes and I have never seen more than about 200 votes usually. Most films have less than 10 votes.
The movies on the first 5 Anticipated pages(Data pool of 190 movies) with over a 100 votes there are… only 8, of which 4 have already had theatrical release(same problem as Tenet).
*averages, votes and plays subject to change
- Ava 672 votes* (Released in Hungary on July 2, 2020)
- Tenet 491votes* (Obviously because people have watched it in other countries)
- Blood Machines 128 votes*(It has been at many film festivals.)
- Possessor 108 votes*(It had its world premiere at the Sundance Film Festival on 01/25/2020)
That leaves us with 4 or 2.11% of movies with more than 100 votes.
- Avatar 2 79% 190 votes*
- After We Collided 87% 171 votes*
- The Flash 74% 128 votes*
- Mulan (2020) 69% 125 votes*
We do have movies to compare these to thankfully.
- Avatar 78% 47.9k votes, 356k watchers*
- After (After We Collided is a sequel) 66% 2.0k votes, 56.8k watchers*
- Shazam 74% 16.7k votes, 281k watchers* (Chose because for a recent DC superhero movie this has the second lowest number of votes, second only to Green Lantern.)
- Mulan (1998) 80% 12.0k votes, 114k watchers*
So that would leave…
- Avatar 2 at 0.4% of the number of votes of Avatar*
- After We Collided (sequel to After) at 8.5% of the number of votes of After*
- The Flash at 0.77% of the number of votes of Shazam*
- Mulan (2020) at 1.04% of the number of votes of Mulan (1998)*
To put this into perspective if these movies were released and the early votes where added to the average how would this many votes affect a rating then? (Keep in mind this is based on the idea that these movies could perform about the same. I also rounded to the nearest hundredth.)
- Avatar 2 would be at 78.00% with 48,090 votes, so 0 points change*
- After We Collided would be at 67.65% with 2,171 votes, so +1.65 points change*
- The Flash would be at 74.00% with 16,828 votes, so 0 points change*
- Mulan (2020) would be at 79.89% with 12,125 votes, so -0.11 points change*
In all fairness (But however very unlikely) that in the most extreme possibility everyone rates a movie as opposite as possible to what the movie would normally average out to, I did the math for that as well.(So if it was going to be 100% after release but early voters all voted 0 hearts what would be the result?)
- Avatar 2 would be at 99.60% with 48,090 votes, so -0.4% points change*
- After We Collided would be at 60.80% with 2,171 votes, so -39.2 points change*
- The Flash would be at 73.44% with 16,828 votes, so -26.56 points change*
- Mulan (2020) would be at 79.18% with 12,125 votes, so -20.82 points change*
I do think that this is a very unlikely result though.
A per user option… this does seem like a reasonable approach. Just a checkmark that says “hide ratings on unreleased content”. I think that ratings should still be shown there if there is no release date because lesser known movies a sometimes do not have a date added to the site(perhaps not a large issue though). You as an individual would have the Tenet example to deal with, so maybe it hides only the percentage while still letting you rate something.
In Summary:
Unfortunately due to the way that Trakt gets its info and the fact that early viewings do happen and that the early ratings are usually not that extreme I do not think that blocking early ratings is a good idea.
So here are my suggestions…
-
A modification to @Sharkiller’s idea, just hide the percentage for everyone until the release date(unless there is none) while still allowing for early ratings. No satisfaction of seeing a rating drop/raise. If implemented highly suggest #2 as well.
-
A way to show you are interested/disinterested before release I believe would work well combined with #1 and divert spam ratings while adding a feature. You can still show your support or lack thereof without affecting the rating system poorly.
-
Forcing the item to be marked as watched if you rate it instead of the option that we have now in the settings page would also probably help solve the issue too. Or at least set it to default.
-
I found that the more votes that Trakt has after release the less early voting matters. So maybe find ways to get more users to vote on content.
-
An integration of all 4 I believe will fix the problem relatively well.