The conventional wisdom these days is that Google is becoming less useful, because people are manipulating its rankings. The storyline goes like this: Once upon a time, back in the Golden Age, nobody knew about Google, so its rankings reflected Truth. But now that Google is famous and web authors think about the Google-implications of the links they create, Google is subject to constant manipulation and its rankings are tainted.
It’s a compelling story, but I think it’s wrong, because it ignores the most important fact about how Google works: Google is a voting scheme. Google is not a mysterious Oracle of Truth but a numerical scheme for aggregating the preferences expressed by web authors. It’s a form of democracy – call it Googlocracy. Web authors vote by creating hyperlinks, and Google counts the votes. If we want to understand Google we need to see democracy as Google’s very nature, and not as an aberration.
Consider the practice of “Google-bombing” in which web authors create links designed to associate two phrases in Google’s output, for instance to link a derogatory phrase to the name of a politician they dislike. Some may call this an unfair manipulation, designed to trick Google into getting a biased result. I call it Googlocracy in action. The web authors have a certain number of Google-votes, and they are casting those votes as they think best. Who are we to complain? They may be foolish to spend their votes that way, but they are entitled to do so. And the fact that many people with frequently-referenced sites choose to cast their Google-votes in a particular way is useful information in itself.
Googlocracy has been a spectacular success, as anyone who used pre-Google search engines can attest. It has succeeded precisely because it has faithfully gathered and aggregated the votes of web authors. If those authors cast their votes for the things they think are important, so much the better.
Like democracy, Googlocracy won’t always get the very best answer. Perfection is far too much to ask. Realistically, all we can hope for is that Googlocracy gets a pretty good answer, almost always. By that standard, it succeeds. Googlocracy is the worst form of page ranking, except for all of the others that have been tried.
Google is a fairly all right search engine, but when I’m looking for what I want it sucks. They don’t seem to want the nice sites or commercial sites to show up on the keywords because I guess they feel they must get paid for it. Just a few months ago, all was fine! Now all my favorite sites are gone from search results and I can’t find what I want at all. All I get is informational crap and not what I’m looking for. A little informational linking is fine, but commercial sites has fallen out of the results. It’s annoying. I think they are cocky and are getting corrupt amidst their new wealth and greed. Face it, Yahoo.com’s search results have a nice mix of info and commercial when I’m looking for what I want and I appreciate that. I don’t feel I should have to click on a pay listing on google to go to a commercial site that should have been in the results somewhere before 1,000! I actually almost feel guilty clicking a paid listing! So I just use yahoo only now. Google’s greed makes me sick. Just look at booble.com. Now what is that site doing to hurt Google? Nothing! It’s funny, and it’s clever. I seriously doubt it would ever infringe on Google’s oh-so-high PROFIT MARGIN! See what happens people. Greed even got to Google.com. There are many instances where they have nixed their “democratic” nature and excluded some sites from the search results almost altogether. I’m turning away from that sort of crap and using Yahoo and sometimes Altavista. They still show the true results!
One interesting side effect of Vivisimo’s clustering of search results is that Vivisimo’s less vulnerable to tyranny of the majority than Googlocracy.
Compare Google’s results for “miserable failure” to Vivisimo’s results for same. The Bush Google-bomb, the Michael Moore Google-bomb, and the Gephardt Google-bomb all get clustered, so it’s possible to factor them out. (Obviously this factoring would be more compelling if there were actually some useful search result for that phrase.)
Anyone attempting to “Vivisimo-bomb” must get a wide variety of unrelated sites to link to some source, and even then the bombing sites might get clustered together. Vivisimocracy is like Googlocracy except that correlated votes “fork” the voting community.
(I don’t have any affiliation with Vivisimo, BTW.)
The political science of Google
Ed Felten has a nice post on Google from a few days ago, suggesting that laments for the halycon days before people tried to manipulate Google are misconceived. His rejoinder: Google results don’t represent some Platonic ideal of the truth…
I’ve recently seen Google-votes been refered to as ghits, which I find rather funny. So “Freedom to Tinker” has 29,100 ghits.
The Problems of “Google Spam”
Google bombing by some nerds is not invalidating the assumptions of age rank, Google’s result sorting strategy.
But I think the problem Google is facing are deliberate attempts by parties with low ethics and considerable commercial interest to manipu…
I have lots of thoughts on this.. I’ll stick to two comments here, one to-the-point, the other a bit tangential.
1. I think a bigger problem is that we don’t really know Google’s algorithm. We know that there is some of this Googlocracy involved, but we don’t know exactly how. They could tweak it so that some sites get more votes than others.. not necessarily based on popularity. We assume it is based on various network measures and it is to some extent, but we don’t know the details. They claim it’s proprietary information so they won’t share (at least that’s what search engines used to claim, I suspect they would take a similar stance, too). So we’re left in the dark when it comes to specifics.. and thus don’t really know to what extent it’s really a democracy.
2. As I said, this is not directly relevant, but I did want to mention that the reason some people can find answers to all sorts of questions using Google (or frankly, all sorts of other search engines) is because these people know _how_ to use search engines beyond simply typing in one word as a query. However, lots of people don’t know how to do this. And if you’re looking for fairly specific information using one search term, even Google won’t be much help.
The edonkey example was a bad one. But try I’m Feeling Lucky for grünkohl rezept” (curly kale receipt) and see very obvious Google spam. Also check the Links on the Page you will be redirected to, all of them try to install a dialer.
Response to Mike Weiksner:
I didn’t mean to imply that Google does a perfect job or is free from bias. My point was more descriptive. It’s a fact that Google uses a voting scheme to rank pages. Because of that fact one can expect Google to share many of the attributes of democracy, both good and bad.
You can certainly argue that it would be better, or more fair, to count the votes by a different algorithm than Google’s. A voting system with different vote-counting rules would still qualify as Googlocratic in my view. I don’t see any practical way to make a good search engine that isn’t Googlocratic in this sense.
Similarly, you can argue that the rules of American democracy should be tweaked, to increase voter turnout, to encourage or discourage third-party candidacies, to reduce the influence of large donors, or what have you. But the result could still be described as democratic.
The public space issue is an interesting one. Whether you can put comments on Bush’s site is of course up to Bush. On your own site, you can write whatever comments you want about Bush’s site. The hard part is getting people to notice your Bush comments over everybody else’s. Any search engine has to rank the comments in some order. Somebody will get to be at the top of the list, but most people’s comments will be pretty far down.
The fact that most comments don’t get noticed can’t be Google’s fault — that just follows from the fact that each user is willing to read only a limited number of comments. At most, you can criticize Google for making bad choices about which comments are on the top of its list. And it seems to me that this is not a separate “public space” issue but the same old “does Google rank pages well” issue.
It is interesting but naive to think of google as a democratic for several reasons. I somewhat surprised that you-of-all-people would post this, given recent censureship by google, including those publishing tinkering-type info I believe.
First, it does not remotely give one-person, one-vote. As Helen Nissenbaum points out how google rankings favor well-funded commercial sites in her excellent paper published in 2000 “Politics of Search Engines“:
“Our study of search engines suggests that they systematically exclude (in some cases by design and in some accidentally) certain sites, and certain types of sites, in favor of others, systematically give prominence to some at the expense of others. We argue that such biases, which would lead to a narrowing of the Web’s functioning in society, run counter to the basic architecture of the Web as well as the values and ideals that have fueled widespread support for its growth and development.”
(And don’t forget the growing importance of google adwords, either.)
Equally as important, democracy must have a link to governance. This is an important way that the analogy to democracy breaks down.
Finally, as currently constituted, there is little or no public space. You have comments on your blog, but how can others find out about my disagreement with items on Bush’s blog?
Google is great, don’t get me wrong. For research and commercial purposes, especially, it is great. It may be better than previously available alternatives, but I don’t we should give the folks at Google a free-pass just yet.
– Mike
As I pointed out in my own recent article on the Google-bomb “miserable failure”, the phenomena is “an illustration of many people repeating something (popularity) for purposes of having it accepted as meaningful (authority).”
Very much indeed, Google is like (shareholder) democracy – and that means it also has the diseases of such systems, such as special-interest groups using bloc action to capture certain debates. I completely agree that Google isn’t an Oracle Of Truth – but a problem is that too many people think it is!
I do a lot of Google searching; the main topics are either political, tech issues, or science questions. On political stuff Google does great as a rule. On tech it does okay. On science it depends; if it’s a question or issue without commercial connections Google is very good to great — but if there’s something possibly commercial involved it’s lousy.
This is a problem. As for telling the difference, no it’s not always easy. But it can be in some cases, and it still doesn’t make Google more accurate in those instances. For instance, if you’re looking for references about the Hunza and their alleged long lives, as I was last year, you have to wade through virtually endless pages of links to recipes for “Hunza bread” and various other commercial elixirs.
It seems to me that if someone had a pile of stuff on that subject, I could tell them what to ignore to get to something likely to be accurate info, but Google isn’t doing that (and I don’t know how you would do it in a search engine — I certainly couldn’t figure out a word and phrase combo that worked, and I’m usually pretty good at that).
I agree that popularity is not the same as truth. Given a choice between pages that are popular and pages that are true, I’ll take the true pages any day.
But there’s no easy algorithm or procedure for determining the truth. It’s just unreasonable to expect Google, or any other search engine, to discern the truth and present it to us. God knows, democracy doesn’t always give perfect answers. But it’s surprisingly reliable at giving non-disastrous answers, both in real life and online.
It’s still amazing to me that Google is so good at finding answers to simple factual questions. Want to know the names of the Teletubbies, or the population of Chad? Google will get you the answers, from reliable sources, in seconds. I wouldn’t have expected that Google’s voting algorithm would be so good at answering simple factual questions, but somehow it is that good.
Now that people are accustomed to the accuracy of Google, people take that accuracy for granted, and they’re disappointed when Google doesn’t find the one true answer to any question under the sun. That’s just too much to ask from Google, or any search engine.
Google is indeed democracy in action; but the problem is, that’s not its purpose. Google uses democracy as a surrogate for accurate judgement. Unfortunately, sometimes the popular opinion is not the correct opinion.
It’s hard to analyze these issues when you’re in the majority. Is there any part of the online conventional wisdom that you disagree with? Think about Google being manipulated on those issues.
Actually, I just did a search to try to find terms where the first Google hit was an opposition-based site. Strangely, it was quite difficult. Scientology leads to the CoS. Creationism leads to a creationist site. Nazi leads to the American Nazi Party. “White power” leads to a white power site!
The only one I found was the one I already knew about, TCPA, which leads to the FAQ published by Ross Anderson, one of the strongest public opponents of the technology. Apparently even the Nazis get more respect online.
But think about how you will feel when a search on evolution brings up creationist sites explaining why evolution is wrong and evil. That’s a widespread view in the U.S., currently under-represented online but that may well change as the net penetrates more deeply into society.
Democracy makes sense for government, because everyone is affected by governmental actions and so everyone should have a voice. But it isn’t necessarily the best guide to information resources. At some point we will need a mechanism to correct search results based on errors and mistaken beliefs that are sufficiently widespread in society.
Google bombing might not be that effective if you use English search words. If you compare the results of “banana recipe” and “banane rezept” (without the quotes), you’ll notice that Google is no longer very useful to find non-English, mainstream content.
The truth is true no matter how many people believe it. (or vote for it) Do you really want researchers on the net to only find the most popular data?
The issue is that people’s votes are limited only by their time, energy, and money. Googlocracy is like a democracy of the “one dollar, one vote” variety, which many people object to.
Google is trying hard to make their rankings come close to the ideal of representing the composite votes of actual net users, and they do a pretty good job, but it’s a hard fight.
My take on the aforementioned conventional wisdom is that it’s people gearing up for a Google IPO. Once Google is public, there will be a neigh-inevitable shift from reputation maximization to short term revenue maximization. Combined with Google’s 18.5 more years of PageRank exclusivity, there is plenty of reason to be concerned about the goodness of Google’s results.