Today is a good day to code

Google Should Revisit its Search Efficacy

Posted: December 31st, 1969 | Author: | Filed under: Uncategorized | No Comments »

Google Should Revisit its Search Efficacy

Picture of Irv Owens Web DeveloperWe all love Google's new cool features, but has anyone else noticed that Yahoo! and alltheweb seem to be returning better search results? I find myself more and more frequently having to wade through several pages of results to find the one I am looking for on Google, but using the exact same words to search alltheweb and Yahoo! almost always get me what I want on the first page.

Google uses a combination of page rank (importance), backlinks, and ultimately page content to determine how pages are placed when users search. When they first started, and there were fewer pages in their index, things were probably easier. But now that they have 8,058,044,651 and counting pages indexed it gets harder and harder to work the backlinks angle. Assuming they have enough horsepower to find every page that links to every other page, the linking would be so diluted it would be extremely difficult to use this technique to filter results. Blogging has made it even worse since now blogs link to other blogs, of which there are millions. Gaming is pretty easy, apparently people were intentionally linking the text 'miserable failure' to the online biography of George W Bush, and now when you search on the text 'miserable failure' on Google, that site comes up number 1 on Google. The problem comes in, I don't personally have a problem with that activity and I think its pretty funny, when someone releases a book called miserable failure, and someone tries to find the webpage for the book using Google.

The same issue crops up on all the search engines. Many of them have resorted to copying Google's method of searching to different degrees, or it could be that the gamers are just that good. Yahoo! and alltheweb both return the biography of GW Bush for 'miserable failure,' yet when I search for 'XHTML Programming,' on Google I get some tutorialfind site, which gives me a runtime error, while Yahoo! and alltheweb both return the w3schools site, which I find to be excellent, as their first result. And yet, the w3schools site is not in the top 10 results returned by Google. Why is the best search engine around serving up broken, or irrelevant results for a simple search on XHTML programming?

Other engines are still using the older model of actually looking at the page's text to determine relevance. The Google model is better for flash sites where there is no real text to search, but in real world use, I am looking for the site that contains some particular content, not the most popular site that contains that content. Some blogs are particularly informative, but since they are text heavy and link light, or really new, they don't rank so highly on Google, those same sites are up there on Yahoo! Alltheweb is a different beast, they don't have the traffic that Yahoo! and Google have, and they probably don't have the same number of sites indexed.

Maybe it is in Google's strategy to ignore their search results since they know it is impossible to completely thwart the gamers, perhaps they are doing a Wizard of OZ on all their users, 'pay no attention to that man behind the curtain.' Either way, I hope they change their focus back to providing great search results, because that is why they are number one. Without that, the features are pretty irrelevant.