Proof of Fake Ratings

  • 1
  • Question
  • Updated 10 months ago
  • Answered
I have what I consider fairly conclusive proof of fake ratings. About 100 consecutive episodes having exactly 10 1.0 ratings and no other ratings. On another series, I stopped at 1 page (about 10 episodes) of 9 1.0 ratings. It's easy to find sites that sell ratings, so that obviously the culprit (or a very bored person with 10 accounts). Is there ANY recourse for this? Does iMDB protect the integrity of its ratings/reviews?
Photo of Adam Feinsilver

Adam Feinsilver

  • 20 Posts
  • 10 Reply Likes

Posted 1 year ago

  • 1
Photo of ACT_1

ACT_1

  • 3637 Posts
  • 3513 Reply Likes

Proof of Fake Ratings
I have what I consider fairly conclusive proof of fake ratings
by Adam Feinsilver
Joined on December 23, 2014
https://getsatisfaction.com/imdb/people/adamesq_3245898
- - -

Could you posts links to these samples ?
- - -


https://getsatisfaction.com/imdb/topics/sonic-mania-adventures?topic-reply-list[settings][filter_by]...
June 22 2018
Col Needham, Official Rep
Thank you for sharing your concern, all we can say about voting is covered in the FAQ at
https://help.imdb.com/article/imdb/track-movies-tv/faq-for-imdb-ratings/G67Y87TFYYP6TWAV


(Edited)
Photo of Adam Feinsilver

Adam Feinsilver

  • 20 Posts
  • 10 Reply Likes
Sure (and thanks for the reply quick reply): https://www.imdb.com/title/tt1915208/episodes?year=2011&ref_=ttep_ep_yr_pv If you flip through the years - almost every episode has 10 1.0 ratings. They seem to have been applied recently. 
Photo of Steve

Steve, Official Rep

  • 1509 Posts
  • 1178 Reply Likes
Hi Adam,

We do have measures in place to detect attempts to skew the ratings and these are weighted such that they will not have an effect on the overall rating. For more information on our ratings, please see:

https://help.imdb.com/article/imdb/track-movies-tv/faq-for-imdb-ratings/G67Y87TFYYP6TWAV


Cheers,

Steve
Photo of Adam Feinsilver

Adam Feinsilver

  • 20 Posts
  • 10 Reply Likes
Hi Steve,

First...Love the Dexter icon! One of my favorite cartoons. Piper (from the shows with the fake ratings above) actually interviewed Genndy Tartakovsky, creator fo the series here: https://www.youtube.com/watch?v=hueSUj2yKwU

As a matter of fact, THIS is one of the titles with FAKE ratings on IMDB - right here: https://www.imdb.com/title/tt6424522/?ref_=fn_al_tt_2 Yes...10 1.0 ratings on that one too.

I've looked at the page you cited. The info mentions safeguards without any explanation of how they work.

The bottom line is that the evidence here is overwhelmingly clear. IMDB is not properly catching false ratings. It's hard to refute far more than 100 episodes that were clearly and methodically rated 1.0 with the exact amount of 1.0 ratings on every episode. It actually appears that this was all done at the end of last year or beginning of this year because the most recent 2 episodes do NOT have the fake ratings...they are the ONLY ones that don't. Properly implemented safeguards should catch this kind of manipulation. Additionally, the overall title for the series was ranked in the 7.X range UNTIL the fake ratings were put on the episodes.

There's no other recourse here? 
(Edited)
Photo of ACT_1

ACT_1

  • 3503 Posts
  • 3367 Reply Likes

As a matter of fact,
THIS is one of the titles with FAKE ratings on IMDB - right here:
https://www.imdb.com/title/tt6424522/
Yes...10 1.0 ratings on that one too.
Posted 4 weeks ago June 25 2018
by Adam Feinsilver
Joined on December 23, 2014
https://getsatisfaction.com/imdb/people/adamesq_3245898
- - -

July 23 2018:

Piper's Picks TV (TV Series)

GENNDY TARTAKOVSKY Talks Hotel Transylvania 2, Popeye, Dexter's Laboratory
+ Producer Michelle Murdocca at Hotel Transylvania DVD Launch  (2013)

https://www.imdb.com/title/tt6424522/ratings

18 IMDb users have given a weighted average vote of 5.0 / 10
Rating .......... Votes
10 ...   44.4%  .... 8
01 ...   55.6%  ... 10
- - -


Solo: A Star Wars Story  (2018) 
https://www.imdb.com/title/tt3778644/ratings
111,062 IMDb users have given a weighted average vote of 7.1 / 10

Rating .......... Votes
10 ... 10.3%  ... 11,400
01 ....  4.0% ..... 4,420


(Edited)
Photo of Adam Feinsilver

Adam Feinsilver

  • 20 Posts
  • 10 Reply Likes
Yes, There's quite a spread on Solo! The 10s are from work trying to get others to offset the 1.0 rating that almost all 272 episodes didn't have until some time in the past year or so. There were no other ratings on individual episodes before that time. Deadtime Stories seemed to have the same problem recently - lots of sudden 1.0 ratings.
Photo of Adam Feinsilver

Adam Feinsilver

  • 20 Posts
  • 10 Reply Likes
Checking back in. Surely, imdb has SOME way to rectify false ratings?
Photo of Adam Feinsilver

Adam Feinsilver

  • 20 Posts
  • 10 Reply Likes
OK - trying to get this addressed on Twitter.
Photo of Bruce Cochran

Bruce Cochran

  • 2 Posts
  • 4 Reply Likes
Being a regular IMDb and Rotten Tomatoes visitor, I am frequently frustrated with the consistent attempts to game the movie reviews and associated ratings on IMDb.  It's clear, if you want consistently "fair & objective reviews/ratings" you can depend more heavily on Rotten Tomatoes to deliver.  On the downside, RT simply does not have the large database, particularly of independent films so IMDb winds up being your only alternative many times.

I've read IMDb's FAQs and understand their philosophy ....to an extent.  That does not excuse the fact that there is routine abuse of the rating system and simply no attempt by IMDb to mitigate those practices.

It would not require a lot of effort by IMDb to start highlighting those reviewers that consistently get "helpful" ratings from other registered members.  People that put in the time and effort to input objective and useful reviews would be rewarded through recognition as "Super Reviewers" and could even carry a more weighted input into an overall rating.  By the same token, users that have only put in 1 or 2 reviews should be flagged as such and maybe not have their rating count as much until they've demonstrated a certain level of consistency.

It also would be helpful if IMDb provided a rating guide popup just before a registered user inputs a rating.  For instance, if you finished watching a film, it probably doesn't deserve a 1 rating.  Similarly, 10 ratings are handed out all too frequently and should probably be reserved for films that you realistically feel will be an Oscar candidate.

I recognize there will always be some level of review & rating abuse, however the problem is rampant as currently structured and in too many cases renders a final weighted rating that is a bold faced lie.
Photo of Ed Jones (XLIX)

Ed Jones (XLIX)

  • 15029 Posts
  • 17154 Reply Likes
I've read IMDb's FAQs and understand their philosophy ....to an extent.  That does not excuse the fact that there is routine abuse of the rating system and simply no attempt by IMDb to mitigate those practices.
That is your quote.


The part that you need to understand is that weighted average Explanation/Non-explanation! They keep secret the formula for the weighted average so that no one knows how to circumvent it. So your no attempt to mitigate those practices is in reality attempted and mitigated all the time.
Photo of Adam Feinsilver

Adam Feinsilver

  • 20 Posts
  • 10 Reply Likes
Thanks for adding in your experience, Bruce. There's still been no resolution to this issue. In my experience, it would just take an easy bot to monitor the time frame all bottom (or top) ratings are entered from multiple accounts in a relatively close time-span! There's no accountability at imdb. I feel that their inaction and failure to take action makes imdb complicit in defamation of projects. Just as an ISP may be liable for knowingly allowing illegal distribution of IP, I think there's a great argument for IMDB being liable fo liable for allowing clearly nefarious and fraudulent ratings on projects.
Photo of Ed Jones (XLIX)

Ed Jones (XLIX)

  • 15029 Posts
  • 17154 Reply Likes
Adam and Bruce
You both choose to ignore my reply.
IMDb does indeed address it.
Your no resolution to the issue is false.
A software program addresses the issue every day, every hour, every minute.
It is not explained so as to not let others know how.
That way they can't know how to circumvent it.
WEIGHTED AVERAGES!!
(Edited)
Photo of Adam Feinsilver

Adam Feinsilver

  • 20 Posts
  • 10 Reply Likes
Hi Ed, I wasn't quite sure what you were saying above in the earlier reply. I posted evidence showing that "weighted averages" is not working earlier in this thread. 1 person posting a bunch of 1s on a project that doesn't have ratings results in 1 (and it shows how many "1" votes were entered. It's dysfunctional. Even if they purport to not be using raw data, it sure seems close after hours looking at many ratings. Basically, what they're saying. I gave a very specific response to this mysterious "weighted averages" assertion above (about 6th months ago).
(Edited)
Photo of Ed Jones (XLIX)

Ed Jones (XLIX)

  • 15029 Posts
  • 17154 Reply Likes
You cant post "EVIDENCE" there is none because of weighted averages.
Titles with less than 500-1000 votes are weighted worse.
Can you explain this rating? Or how it works?

Photo of Ed Jones (XLIX)

Ed Jones (XLIX)

  • 15029 Posts
  • 17154 Reply Likes
Those 10 one ratings you complain about will not in the big picture screw up the OVERALL ranking because of WEIGHTED or what can be described as FIXING.
IMDB IS AWARE of the practice yo describe and gives 10's and 1's in groupings less creedence. IT IS A SECRET HOW THEY DO IT. How many times must one say it before it sinks in. Those 10 votes in essence are equal to 3 votes. For one to make a dent in the rankings they would have to do it on a MASSIVE SCALE. Not 10 but thousands. Do you get it now?
Photo of Ed Jones (XLIX)

Ed Jones (XLIX)

  • 15029 Posts
  • 17154 Reply Likes
By the way I have seen stuffing on the order of approx 1500 reviews with rankings of 10 on a title. THAT WOULD effect rankings. 10? C-mon!
(Edited)
Photo of Adam Feinsilver

Adam Feinsilver

  • 20 Posts
  • 10 Reply Likes
There's no need to be condescending. It doesn't lend credibility to your argument. On a title that doesn't have many ranking, YES, rankings are easily skewed...I posted clear, incontrovertible data on that. On a show for kids (who often don't know imdb exists), one person stuffing 10 1s does have an effect, particularly when they're being stuffed on individual episodes as well (in a methodical manner on a short time-frame that would be easy for IMDB to rectify with automation). Illegitimate iMDB rankings can be purchased. A corporation with millions invested would easily spend a few hundred or a couple thousand on fake 10 reviews. YouTube nailed a bunch of companies for purchased, fake views a few years ago. Weighted averages are completely insufficient. I agree with you that IMDB is aware of fixing. The failure to take serious corrective action is reprehensible.
Photo of cmonaptite

cmonaptite

  • 1 Post
  • 3 Reply Likes
Hey, Ed CliffJ Jones D. Weaver III (XLIX) calm down. You're supposed to be here to help customers and not get angry and argue with them.
Photo of Jeffrey Strickland

Jeffrey Strickland

  • 16 Posts
  • 16 Reply Likes
That made for quite the read but I'm inclined to agree with Ed on this.

First of all the system works, the ballots cannot be stuffed so to speak I've seen this in action and though I don't fully understand the algorithm I can see that it's effective.

Second based on what your saying there is clearly something fishy going on, however what do you expect IMDB to do exactly? Discount peoples ratings based on it being fishy but with no actual proof? Ratings all appearing at the same time could be from a group viewing of something or just outstanding coincidence.

I agree with what you're saying but ultimately this isn't proof it's just speculation and you can't expect ratings/reviews etc to be taken down on speculation alone.


(Edited)
Photo of Ed Jones (XLIX)

Ed Jones (XLIX)

  • 15029 Posts
  • 17154 Reply Likes
Not Angry
Not Arguing
cmonaptite
An assumption on your part.
And incorrect I might add.
Photo of Ed Jones (XLIX)

Ed Jones (XLIX)

  • 15029 Posts
  • 17154 Reply Likes
Jeffrey Strickland
Thank you. The points you make are a perfect addition to what I said.
If I said all that too, It would have seemed a little over the top.
Thanks again for finishing my thoughts.
Photo of Adam Feinsilver

Adam Feinsilver

  • 20 Posts
  • 10 Reply Likes
Hi Jeffrey,

What I would expect is that, when EXACTLY ten "1" reviews get posted individually on well over one hundred episodes of one title, as well as the title itself, in a very short time span, the ratings are discarded. (I don't have any doubts that the same accounts did the ratings on each episode as well.) IMDB would have the proof that this is what happened. This can easily be tracked automatically - match accounts to times, ratings and posts. I don't think it's unreasonable to suggest that when a title that has only a handful of episode ratings, suddenly has 10 "1" ratings added to over 100 episodes, something is amiss (to put it mildly). I think if precautions were implemented, there would be a LOT of titles the have had similar situations (possibly in higher numbers). It's pervasive in across social media. YouTube wiped out thousand of fake views and ratings across many, many videos (and continues to do so). It happened to Sony and many high profile artists. Instagram wiped out bot accounts. So did Twitter. It's a problem. 
(Edited)
Photo of Jeffrey Strickland

Jeffrey Strickland

  • 16 Posts
  • 16 Reply Likes
You're right Adam but again that's not evidence it's pretty circumstantial. It's overwhelmingly suspect and questionable but it's not 100% proof and there could be explanations behind it.

When it comes to big Blockbuster movies such voting bias can be entirely ineffective but for the smaller independent movies/shows it can be devastating. I entirely understand that but my question is what precautions could be set into place?

What system could be implemented that wouldn't screw over genuine voters/reviewers? I've had this conversation a number of times and ultimately there doesn't seem to be a foolproof one.

Regarding Youtube they've done a lot of things like that, they also purged lesser active accounts. Personally I think the trolling, racism/bigotry filled and distinctly disgusting comments section being monitored better would make more sense but that's their prerogative.


Photo of Adam Feinsilver

Adam Feinsilver

  • 20 Posts
  • 10 Reply Likes
Reviews are a tough one. I don't have a good suggestion beyond flagging for human review. (It's unfortunate that the forums had to be removed a while back, but totally agree - there were disgusting comments in there.)

Ratings are much easier to address. A system can fairly easily be implemented using data IMDB has access to. In this case, one title with a few hundred episodes (most have been listed for many years), almost all unrated - some with very few ratings suddenly gains 10 "1" ratings...all within a very short (maybe a day(s)) of time, a red flag should go up. Data passes on to a look at how quickly the rankings were added and in what order (for example, each episode, in consecutive order gets a 1 rating, then repeats from the first episode again)...some easily discernible pattern. Then, see if these ratings were added by the same accounts and in the same manner. There are SO many data points here. On top of that, NO reviews were added. Additionally, exactly 10 "1" ratings were added to the title...no more, no less...and no further activity for months. No new reviews added at all. IMO, the evidence is direct and it can be validated. This is simple screening. YouTube's algorithm is lightyears more complex (and YouTube didn't come along until 15 years after IMDB). It sounds like IMDB is not even tracking to see if consecutive ratings are coming from different accounts from the same IP address. 

Even in a criminal proceeding, the standard of evidence is "only" beyond a reasonable doubt. 100% proof is impracticable, particularly considering the serious implications of false ratings. It's never foolproof.

I agree that this issue can be (and like is) devastating for smaller indies, especially since imdb rankings pop up in Google results. IMDB should take its impact on others in the Industry far more seriously and strive to eliminate serious shortcomings. It's 28 years old - they should have some solutions.

I do appreciate the civil discussion, BTW! Gotta get back to being productive. This issue is just a big one for me...and Bruce kicked it back to the front burner. 
Photo of Bruce Cochran

Bruce Cochran

  • 2 Posts
  • 4 Reply Likes
Hmmm, first return visit to this forum, so I was somewhat surprised to see so much added dialog (guess I need to work on my spam filter). 

I have a lot of value for IMDb and will continue to use the site regularly for those areas where they excel.  Unfortunately, I've just become too conditioned not to depend significantly on IMDb's weighted average ratings to serve as an indication of whether or not I'm likely to enjoy a particular film.  If the overall rating works for some, that's great - it too frequently just doesn't work for me.  I realize to some extent this is down to individual taste, although I've read many IMDb reviews that seem to express the same wonder and frustration about the frequent proliferation of skewed ratings/reviews for some films.

I don't have the answer, but I remain unconvinced that there simply are no opportunities for improvement.  As an early subscriber to Netflix, I liked that they had developed a rating system that attempted to predict subscriber's ratings for films you hadn't watched based on your rating history.  The system was far from perfect, yet I have to say if the predicted rating was 2 out of 5 stars and I chose to watch that film anyhow, that film indeed generally proved to be a disappointment. 

Sadly, while Netflix seems to have shifted to a more simplified and less accurate approach, I have to believe at some point in the not too distant future an "intelligent" rating system will show up that will take individual tastes into account.  IMDb's prominence in visual media makes them the ideal candidate in my mind to make that step change.
Photo of Ed Jones (XLIX)

Ed Jones (XLIX)

  • 15029 Posts
  • 17154 Reply Likes
Bruce the recommended Movies are ??? Weird to say the least.
I agree.
These are the recommendations for One Over The Cuckoo's N.
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
They do leave a lot to be desired.

This conversation is no longer open for comments or replies.