How to stop bot ratings

@nick_burfle

You are missing the point here, yes every user should be able to rate in a perfect world and yes most likely many of the reviews won't be seen. 

And yes generally speaking a rating is subjective and depends on perception but, come on, you have to see the forest not the tree almost all sorts of movies nowadays have a big rating even the ones with a simple plot small production and noname and mediocre actors!

You could (if you would like to be the devil's advocate ) argue them per case but not as a whole.

Like compare how movies were rated years ago (before lets say 2014 ) the ratings did not use to be like that e.g you saw an 8/10 rating ? The movie would be a classic/master piece now even run of the mill documentaries can land that score ... e.g the movie Bullies 2018 has a 8.8/10 score its a short that I didnt know existed and is not memorable at all, is it equal to the Matrix? Forrest Gump? LotR? Pulp fiction? Harakiri? The good the bad and the ugly?  or even the godfather? Hell it has a higher rating compared to Silence of the lambs and Green mile lol  and I could give you hundreds of examples of bloated fake rates that are 8/10 or higher from movies you probably have to google (and that on its own means they are not worthy to be even close to 8)  e.g The painted warrior, Not to Forget 2021, DNA 2019 and so on and so forth. 

And the issue is not with 8ish rated movies its all along the spectrum above 5 

The point you are missing is that IMDB ratings are just losing their meaning/usefulness since most of the movies are artificially bloated with high rates which are systematically added either by the production it self or by a (or multiple) PR companies. 

There must be a bottleneck to this to at least reduce those inflated/fake rates by introducing a bottleneck which will make things more expensive (if you are a PR company and want to inflate the rates) both in terms of time and manpower needed.

In my example this could be solved by forcing you to write a review (and also monitoring that review with AI to nerf/exclude generic text etc) because clicks are much cheaper to fake than having to write something genuine that makes sense and doesnt look like a bot typed it. 

The rates themselves (as in the mathematical formula to find a mean rating out of the total ratings)  have to get more sophisticated as well simply because of the above reason so that a PR company not only has to spend more time in order to inflate ratings but also not be able use 1 day accounts or bot accounts or accounts of friends and family of the production etc.

Maybe there is a better way to solve this what I mentioned in the OP is just how I would think that rates would become more realistic. 
 

You made a number of suggestions on how to improve and make more solid the rating system, but I think you missed one.

You mentioned a short movie with 8.8/10 but only 17 ratings. We all know that 17 ratings have no meaning. If you make a home movie and you want it to be rated, you don't need to use complex bots, you just ask your mom, sister, niece and so on to rate it. And yes, they will give you 10/10 because they love you.

So I think that maybe IMDB should set a threshold (eg 1000 votes) before the average rating is shown in the main page or the movie is listed in some research result.

But in general, I do not agree with the fact that ratings are losing their meaning or usefulness.

We have all the instruments and information to use the ratings properly: the rating itself, the number of votes, the demographic breakdown, the top1000voters. 

Also, in the advanced search, you can filter both ratings and number of votes, so you can easily get rid of the meaningless 8/10.

Hi uomotora,

I generally agree with your conclusion; we have the information we need.

For niche movies, though, I think 1000 ratings is too high.  Many of them may never get so many, and for those of us who view them, any rating may be more useful than none.  I looked at the last 7 of these I'd rated, and see that they have 34 to 693 ratings.  In all but the least rated one, a good normal distribution shows up, with the usual surfeit of 10s of varying sizes.  So I would hate to see that information disappear.

With 100 or more ratings, the extra 10s and 1s from bots or shills or indiscriminate viewers rarely affects the ratings for the practical purpose of deciding whether to watch it (in my experience).  So, just spitballing, I would prefer to keep ratings visible if there are 100 or more, but perhaps eliminate them from searches and top lists to avoid annoying folks who care about a 0.1 point difference.  Just a thought.

Hi nick_burfle,

Thanks for your reply.

The 1000 ratings threshold I mentioned was just an example, but as you said it might be set much lower.

Just as a note and to be fully clear, for movies below the threshold I also don't want to lose the rating breakdown information. I just suggest to substitute the value of the average rating in the main page with some sign indicating that it is a niche movie with few votes. 

@nick_burfle​ 

And 

@uomotorta 

Well, I admit that the selection of the movies I used as a reference for being totally overrated was somewhat in purpose. 

"Somewhat", because I didn't pick them strategically just because they suit my narrative, I really just went to a movies catalog I use and filtered the thumbnails it shows by specific years (from 2018 and on I think was my selection) and sorted according to the IMDB rating and chose them almost at random. 

I say almost (and here comes the "in purpose" I mentioned above) because I ignored movies that you could elaborate that they are not bad (since rating a movie is always subjective to a degree but defining that degree via forum posts will be a meaningless and endless war of attrition in my humble opinion so to avoid that I ignored those movies) 

Now to address the points,

Well yes I agree that the rating of a movie shouldn't show up unless it reaches an X number of rates (e.g 1000 as was mentioned ) 

Ok I take the point that there are people that watch those movies, no issue with that, but I also think that "with great power comes great responsibility" :P 

by that I mean you should be prepared for it to suck (or to be likable to your particular taste but objectively suck) and you select it by your own volition, you chose to watch it, I disagree that "any rating" would be helpful since all the ratings will be good so all the movies or 99% of them who have a low number of people rating them will be masterpieces in terms of score so it doesnt differentiate them by a lot  

What it does though is that they are coming up to “average” (for lack of a better word) people like me when they try to sort movies by IMDB rating and really bother them since they get mislead and have to click or google the movie to check the number of people rating them = an extra hustle... and I think those "average" people are the vast majority. 

While people like you could do without that since you know what you are getting up against so to speak :P 

Last but not least if not at least an X number of people (e.g 1000) bother to rate a movie (or even watch it) that on its own tells something about said movie. 

Now to address umotorta's accurate point that the movies I used had a small number of people rating them, I explained as to why I used such movies above in this post but I now will unleash the pandoras box (SMALL SPOILERS MAY BE PRESENT HENCEFORTH) and put out other inflated movies that have many people rating them just to showcase that the issue is "global" and not just about the movies wich are rated by a small number of people.

One of them is Matrix 4 (which increased by .1 since I made my last post I think)  I already argued about it and why it is a cheap shot 

The tomorrow war.... The plot has 1000 stupid plot holes and unnecessary scenes like  5 random dudes found the location of the beasts which was in an unkown place aka somewhere in the general vicinity of the entire daman north pole in a matter of a few hours... which is silly on its own...

and while they had the explosives to detonate the entire spacecraft and kill all of the beasts they preferred to try to use a toxin one at a time risking waking them up (so that essentially the mother beast can awake and have that "epic" fight with our heroes) while still successfully detonating the entire thing.. and the best and brightest of the pentagon or Kremlin or the UN didnt even think of looking up the spaceship yet those 5 random dudes not only thought of it but found it in a matter of hours lol 

In the future time while they made the toxin that will kill all of the aliens (also conveniently they produced the toxi in a few hours and also they couldnt find a single female before that when 50% of the current human population was sent to the future lol  but they had to send only like a dozen of troops to successfully retrieve the female)  the female beast was calling the males to its rescue which all swarmed through the sea and invaded the stronghold (if they are so capable and intelligent why didnt they do that long before all this happened? anyway) in a matter of minutes yet they didnt kill the female beast ...

like they could just shoot it under its belly or something while they had it in chains and subdued, lol so it can stop calling the males and they could return to the past more conveniently (since the female like almost killed them and killed Chris Pratt's daughter in the process of them trying to return to the past with the toxin)  

Also ok the female beast was not dying by the toxin but the males would why not just use the toxin they know kills only the males in the mean time while they were trying to find a toxin that kills that 1 female ? 


Not to mention that tanks could be the solution lol  I mean if you can kill them with bullets means that their body doesnt have the force and the rigidity to crack the armor of a tank (other wise it means it is as strong as the thick tank armor which would render them invulnerable to bullets even 50 cal  ) 

Also why to try a find toxin that kills the female or infantry against the beast, found the the female is? NUKE THAT PLACE WITH 3 NUKES one after an other just for good measure problem solved... kill the rest of the males with the toxin you have and know that kills them lol 

Acting was mediocre at best especially in the emotional heavy scenes... most of the cast were b grade actors 

It is literally like a movie inside joke that would come up on Tropic thunder as a trailer :P

Yet that movie has a 6.6/10!!!!!


Ok I spent much time just presenting points (and not even touching the tip of the iceberg with that one) on the tomorrow war so I will be brief on this one... 


Teenage mutant ninja turtles 2014.... like its just cgi with a mediocre plot (and out of canon in some places if not many) why did that get such a high rating? 

I mean how does it differ from e.g Street fighter 1994 which had prominent actors (for the 90s) such as Jean-Claude Van Damme Raul Julia and Wes Studi ? 

Kids loved it for sure (I loved street fighter when I was a child and watched) but it wasnt much to write home about and it got the rating it deserved because back then IMDB ratings were trustworthy 
 

@Col_Needham Thanks for bothering responding to my post, yes the FAQ addresses some of my points but (without wanting to be rude, just honest) just addressing an issue as in list a reply doesnt solve much. 

I mean e.g on the question if the rating system is broken the short answer to that basically says it is not broken and that the rating system does not simply average out the rates but uses an undisclosed method of weighting the values. 

Ok maybe the algorithm is not broken as in does not have bugs but while it does what it is supposed to do, what it does simply doesnt work that well as in what users would expect the ratings to reflect or what IMDB as a company should/would expect (unless there is an "under the table deal" between IMDB and Hollywood production studios to boost sales, which I am not saying there is I am just saying that this is the only reason I can think of that IMDB should be satisfied with the current state of ratings)

Why should IMDB care about the ratings reflecting or being much closer to the reality?

Well e.g I as an consumer used to choose between a catalog of movies since I cant watch them all, on a given timeframe,  I used to check on IMDB to see which movies are more promising based on their rating and this really used to work, ok I may be able (as a self-proclaimed cinephile :P ) to pass accurate judgment on if I will like a movie with fairly great accuracy just by glancing on the cast and checking out the trailer but there are many people that cant be that decisive and need a more direct advice aka ratings. 

Same goes for "Which movie ticket should I spend my money on" and if IMDB ratings disappoint me (in the way that the movie had a high rating but I wouldnt even consider it to be mediocre) and this persist for several ticket buying decisions I soon will stop using IMDB for its ratings. 

Lets take the movie that triggered me into making the OP Matrix 4 ok it has not a very high score(yet I bet it will inflate higher as time goes by to at least a 6) it currently has a score of 5.7 this would suggest that it is a fair if not good movie to watch not a master piece or a classic (this would be a 7+ movie) but a movie doesnt have to be a master piece in order to be a good watch.

But reality tells us otherwise the tickets sold (and various revenue from streaming/renting services) are ridiculously low  but it doesnt stop only there word of mouth (brwosing through various forums) also depicts a greater picture of disappointment for that movie even if you check IMDB reviews closley for that movie you will see that under the reviews that give it a 5 or lower rating around five people found almost each such review useful while the reviews that have a 10/10 (and usually are brief) or a high rating e.g 7/10 have either 0 or a couple of users that found it useful (and when they dont have 0 users that found it useful but a couple its again a minority e.g 2 out of 6 users found that useful which means that 8 people voted on the usefulness of said high rated review and only two agreed that it is useful as in contrast the low rated reviews of said movie are more like "5 out of 5 users found that useful") 

Again " I do not see coincidence, I see providence"

@nick_burfle 

I know that you did not mean anything bad by it but I can assure you that "big numbers" and how they influence statistical results is not a strange idea to me. 

The thing is though that what you started (and I am sure you had no such intention) is a conversational trap. 

I mean for a given fact or event there can be numerous explanations/apologetics towards why it is how it is  and there can be noumerous excuses/other scenarios that can be used to explain it, IF we ignore historical data and the entire premise I mean there are counterarguments even as to if earth is flat with seemingly rational or even (despite being ultimately wrong) sophisticated explanations as to how a flat earth could work (and I am not a flat earther nor do I imply that you are) 


Yes we could just reduce the issue to big numbers and such, or exchange 10 posts back and forth in a span of days elaborating as to how big or small a factor big numbers are and even come up with new reasons and/or excuses and ultimately end up nowhere..

But we're ignoring by doing so is that IMDB ratings historically used to be darn accurate! 

Even when movies that we now consider classics were fresh e.g look at the rates of an archive.org (waybackmachine) from IMDB's movies ranked by rate back from 2005

https://web.archive.org/web/20050828074614/http://www.imdb.com:80/chart/top

Notice that the people that participated in voting for each title are about 100k or less 


Yet the scores are still fairly accurate and still similar or identical to the current rating most if not all of the movies have in that list, despite quite a few of them not being classics but rather new movies in that time period (2005) 

Edit: Same can be said for the bottom movies of 2005

https://web.archive.org/web/20050829005109/http://www.imdb.com:80/chart/bottom

Including new films (back in 2005) such as "Alone in the Dark" or the older 90s Streetfighter film which again had accurate (bad) reviews that still hold up today and that (as franchises/brands/genre) are comparable to nowadays (highly rated by inflating rates) bad movies that use video game franchise to milk  money from children and fans... 

@Dredderick_Tatum​ I totally agree with you since PR teams will tackle all those big sites (as well as social media they wont just use facebook but all of them) 

But I think most budget goes to inflate IMDB since that's the prominent one in the others e.g rottentomatoes you can see a few movies escape that over inflation (especially in the review score) 

The thing is that hollywood needs to understand that this does more harm than good... 

I mean already most of the movies are mediocre and/or re-re-reboots lazy stuff noone will remember just because everything sells more or less the same since it gets more or less the same exposure... 

They even made a space jam 2 like of all movies this one surely needed a reboot lol 

Everything marginalized and decision taken by bean counters instead of directors cast and plot. 

A travesty truly....