So this post begins a series of posts that will examine Googles algorithm and how it relates to predicting outcomes. To test if one link one vote actually holds true. Local SEO Guide got me started on this with his great examination of the 2008 Iowa Caucus results in relation to each of the Candidates incoming links. In these posts I will be examining the major categories of the 2008 Hollywood Foreign Press Associations Golden Globes and using the logic set forth by the Google Algorithm try and predict the winners. The Golden Globes will televised on NBC on January 13th 2008, so these posts will wrap up by atleast January 12th or sooner if time permits. I will be analyzing the movies official site URL as well as the iMDB link as many people in the industry or in the Hollywood blogosphere will link straight to the iMDB page instead of the official site. So those two URLs I will evaluate each in Yahoo! Site Explorer, Google's link: command and the number of links in Technorati each has and they will be tabulated together for each film to determine who the winner should be.
Best Motion Picture - Drama
American Gangster
Imagine Entertainment/Scott Free Productions; Universal Pictures
- Yahoo! Site Explorer Totals: 57,457
- Google Totals: 1,122
- Technorati: 2,037
Atonement
Working Title Films Limited; Focus Features
- Yahoo! Site Explorer Totals: 31,381
- Google Totals: 621
- Technorati: 1,253
Eastern Promises
Kudos Pictures/Serendipity Point Films; Focus Features
- Yahoo! Site Explorer Totals: 21,795
- Google Totals: 296
- Technorati: 1,492
The Great Debaters
Harpo Films; The Weinstein Company/MGM
- Yahoo! Site Explorer Totals: 15,542
- Google Totals: 0
- Technorati: 451
Michael Clayton
Samuels Media and Castle Rock Entertainment a Mirage Enterprises/Section 8 Production; Warner Bros. Pictures
- Yahoo! Site Explorer Totals: 33,534
- Google Totals: 1,326
- Technorati: 1,017
No Country For Old Men
A Scott Rudin/Mike Zoss Production; Miramax Films/Paramount Vantage
- Yahoo! Site Explorer Totals: 29,665
- Google Totals: 406
- Technorati: 1785
There Will Be Blood
A Joanne Sellar/Ghoulardi Film Company Production; Paramount Vantage/Miramax Films
- Yahoo! Site Explorer Totals: 11,635
- Google Totals: 198
- Technorati: 902
After looking over the results hands down American Gangster is receiving more buzz than all the other films. It has more links in Google, Yahoo! Site Explorer and Technorati, in fact it more than doubles most of the other films. One thing we must look in this though is that in late 2007 the American Gangster DVD Screener was leaked on to the internet and of course there was a buzz about that on new sites and industry blogs. If I was googlebot I would be spidering all these links and categorizing them so I would know better a true estimate of how many links were an editorial comment about the movie itself and not another news item related to the movie. No Country For Old Men did nearly as well as American Gangster on Technorati so it got nearly as much blog buzz but less over all links. Is it safe to say American Gangster walks away with the prize? I don't know but I am going to have to because I am too lazy to shift through all those links and determine what's what. I have watched both of these films and both were phenomenal, American Gangster probably has more mass appeal than No Country For Old Men and sometimes the Golden Globes will showcase the more out there picture as best picture as opposed to the Academy Awards. So this is a tough call; but as I said this is going to be based on the principle of one link = one vote. So I declare the 2008 Golden Globe Best Picture - Drama Winner as American Gangster. Remember to check back to see the analysis of the other categories and on January 14th to see how Google's algorithm actually did.
Andrew Shotland over at Local SEO Guide had a great posted the other day about his analysis of backlinks to presidential candidate’s websites and remarkably the google algorithm correctly predicted the Democratic Winner; but was wrong on the Republican side according it to his analysis Ron Paul should have won hands down with more than 3 times as many links as Mike Huckabee. As we all saw though that wasn't the case. But anyone who has been following Ron Paul or his rise to internet fame knows that Digg users absolutely love Ron Paul they eat up everything that Ron Paul related. So of course he is going to have more links than anyone else.
So does this mean Google's idea of one link equals one vote is broken? Or is the Ron Paul phenomenon more akin to spam and therefore not as relevant? Because as we saw the number two Republican Mike Huckabee did indeed win. So maybe in Andrew's analysis he should have penalized Dr. Paul just like Matt Cutts would penalize a filthy spammer. Lets face it, Digg users basically spam consistently and blindly promote Ron Paul stories all of which of course point back to Dr. Paul's site. And as we all know it's a very limited number of Digg power users that really control the content of the Digg front page. And once it goes hot on Digg of course more and more people link to the story in hopes to get a little residual traffic from talking about the hot item. Voting up and talking about that item becomes the "in" thing to do. So it then becomes unfair to count each one of those links as a vote, because in essence they are manufactured links - in reality nothing more than a link farm.
So am I trying to same Google's algorithm is broken? No I am actually saying the opposite, that Google's algorithm is fairly on target as it goes with deciding what content is popular and relative. But it also shows a fundamental flaw in the Google algorithm, that people can give links and people can get people to get links and powerful people can reach more and more people garnering more and more links. So good news is that Google gaming will continue on as long as links are part of the equation. Even without links I am sure some portion of the Google algorithm will have some gameable component.