Do No Harm: Internet Ratings Explained (Sort Of)
By Andrew Robinson
We’ve all been there, logged on to some website or service where inevitably we’re asked to “rate” something, be it a film, product etc. Most of us chose our rating and simply move on and don’t think twice about it. However, the act of rating something, especially a film or product, becomes trickier if you have a vested interest in the success of said film or performance. For instance, what if someone you knew was in the film you’ve been asked to rate? What if you were in the film? Obviously, you want the world to know how good the film is, but more importantly how good your friend is in it. Or maybe you simply want your own name to be associated with a high ranking or positively rated project. Is that wrong? No. It’s perfectly natural.
You’re going to vote favorably and you’re going to vote often.
The problem with this approach isn’t your enthusiasm it’s the fact that services, especially those found on the Internet, have gotten wise and have put algorithms in place to combat ballot stuffing. Take IMDb.com for example, IMDb is pretty much the source for entertainment news, information etc. One of the many features IMDB has on its site is the user generated rating system, which spans from one to 10 stars with one being the lowest and 10 being the highest. Since IMDb is often the first place people look when trying to find information on a film, especially an independent, a lot of people put a great deal of stock in the IMDb rating and try to pump it up, sometimes artificially. In theory this would mean that all you would need is 100 of your closest friends to go to IMDb vote 10/10 and all of a sudden your rating is perfect; or at the very least high enough to thwart the net nerd who trashes everything. No so fast.
You see most rating systems are not based on averages of the total number of votes. They use complex and often proprietary algorithms to generate a “true” average or weighted average that better reflects the “intent” of the voters or votes. These algorithms also take into consideration where the vote is coming from and how new the vote is and will often discount suspicious or repeat votes.
For example: Bob really liked the film “Killer Zombies from Antarctica” and Bob logs onto IMDb and gives it a 10 star rating. Bob then tells 100 of his zombie loving friends to join IMDb and do the same. Overnight “Killer Zombies form Antarctica” has 100, 10 star ratings. The next morning Bob checks his favorite film’s IMDb rating to find that it’s ranked 3.5 out of 10. Bob can’t believe it; he begins to vote 10 out of 10 over and over and over again in hopes of raising the rating. The next day he returns to find that the rating is lower still, even though no one has rated the film below five stars. What gives?
What gives is that Bob is experiencing the weighted average system IMDb uses. While Bob may be a long time IMDb user his friends most likely were not therefore the system views their votes differently, especially if they’ve signed on and voted for only ONE film and gave it a 10 out of 10. The IMDb system can see this and views it as ballot stuffing and will often count these votes for less or throw them away entirely, meaning Bob was the only one out of 100 who gave the film 10 stars. Then when Bob began voting over and over and over again for his favorite film the IMDb system saw this as well and either threw all the votes away (even the legitimate ones from Bob) or began to count them for less, which drives the total overall rating down.
How does IMDb know that Bob is ballot stuffing? Simple, every computer that has an Internet connection has a “fingerprint” of sorts and that fingerprint can be read by IMDb so that when multiple votes come in from the same source they only count the initial vote and not the 100 or so that came after it.
So what Bob has done is essentially “hurt” his favorite film by being the best fan in the world.
I tell you the story about Bob for two reasons. One, I know many of you like to support what it is that we do and in doing so try to rate our work 10 for 10 whenever and where ever possible. First and foremost, those of you who do this are AMAZING and we are forever grateful however we want you to know that simply voting once is enough and whether you vote once on IMDb or once on every site you can find we love you all the same. Second, there are some of you out there that like to combat the net nerds and in doing so will try to counter their negative votes with positive ones, this doesn’t work and I can tell you from personal experience-that you can’t beat the net nerds. It’s best to not engage them or feed them.
I’m writing this because I was on IMDb the other day and noticed that while In The Darkness had jumped into the top 10,000 films on IMDb (a 240% percent increase for the week) the overall “rating” was something like three and a half stars out of 10. When I clicked on the ratings explanation I saw that we had a pretty good spread of votes and when I took the average of what IMDb was showing I came to the same arithmetic mean that they did; six and a half stars out of ten. But we have three and a half stars out of ten because I think many of our fans have been trying to “pump us up” and in doing so have lowered the rating.
I’m not upset, not at all, because I don’t give a shit about what people say or think of me on the Internet. Okay that’s not true; I care what my real fans say but anonymous folks on IMDb or Hulu or where ever can bite me.
That being said, I know many of you do care and want people to know of the good things we do and the people we employ because you truly do support us. Again, we appreciate it so much, but know that we’re okay with what ever comes our way via the Internet because we have to be…we signed up for it.
We appreciate your votes and encourage you to participate on sites like IMDb and Hulu, but understand that once you vote or post your comment you have to be okay with moving on. If you want positive interaction, discussion and banter you’ll have to keep it tuned to our sites, Mattoid sites, for you know we’ll always keep it sexy. As for the criticisms, flames and/or personal attacks swirling around out there in the ether…let ‘em be.
Originally posted on Mattoid Entertainment’s website www.mattoidentertainment.com
Leave a Reply
You must be logged in to post a comment.
April 9th, 2010 at 12:43 pm
[...] This post was mentioned on Twitter by In The Darkness. In The Darkness said: Can you help to much? Internet votes explained…sort of. http://tinyurl.com/y69vysb [...]
July 15th, 2010 at 2:46 pm
Buy:Synthroid.Arimidex.Human Growth Hormone.Accutane.Actos.Prednisolone.Retin-A.Valtrex.Zyban.Nexium.Prevacid.Mega Hoodia.100% Pure Okinawan Coral Calcium.Zovirax.Petcam (Metacam) Oral Suspension.Lumigan….
September 5th, 2010 at 7:16 am
Buy:Super Active ED Pack.Cialis Super Active+.Cialis.Tramadol.Cialis Soft Tabs.Viagra Super Active+.VPXL.Propecia.Viagra.Soma.Viagra Professional.Viagra Soft Tabs.Zithromax.Cialis Professional.Levitra.Maxaman.Viagra Super Force….