Google's Ongoing Pagination Bugs on Search Results Page

Sep 22, 2009 • 8:54 am | comments (2) by twitter Google+ | Filed Under Google Search Engine

Ever conduct a search and see that you have at least ten pages of search results to browse through but then, as you get to page 4 or 5, Google cuts down those ten pages to only 4 or 5 pages? Here is an example, search for [lockeeeer] and you will see one through nine pages of results at the bottom:

Google Pagination Bug

But when I click to page five, I get stuck there:

Google Pagination Bug

Why is that? In this case, does it have to do with Google possibly first including "omitted results" in the pagination and then removing them when you click forward through the search results? Or is there just this ongoing pagination bug for some search results?

Forum discussion at Google Web Search Help.

Previous story: Google Latitude / Reader Bug Leaks Non-English User's Location?



09/22/2009 02:06 pm

This kind of behavior has been around for a very long time. I don't believe that there's a bug involved in any way. Unless there are a very limited number of results for a query term, the total number of results that we see indicated are usually only an estimate. This has been mentioned in at least one Google patent filing. To save on computational resources, search engines follow a number of strategies that may skew the number of results they indicate as well. One intentional strategy is to calculate, order and return only a limited number of results. Why find every result for a query if most people only look at the first page of results, or the first few pages? This is an effort to limit the computational cost of searches. This can mean using a multi-tiered index, where results might first be captured and available in a cache rather than through an actual lookup in a database. This can save on processing power, especially for popular terms, and can also result in a "number" of results that isn't necessarily any more than an estimate either. If there aren't sufficient results for the first page (or subsequent pages) in the cache, the search engine might look to a main index for results. If there aren't sufficient results in that main index for the first page (or subsequent pages), the search engine may then look to an extended index. There are also other processes that additional results may go through, such as filtering of duplicate results, filtering of adult results, domain collapsing for indented (or tabbed) results, and more. Some of those results might be seen in the "omitted" results, but others won't. If these processes can be limited to the first page of search results, rather than all relevant results, it can keep the results we receive from taking too long to process.

Michael Martinez

09/22/2009 07:06 pm

I have often wondered if there is not also a PageRank aspect to it. They don't seem to index every word and expression on a Supplemental page, and I have often run into shortened keyword queries where I knew the index showed relevant pages in site searches but which were not included in the shortened keyword queries. It's quite annoying, actually, to KNOW there is relevant content in the index(es) that is not being shown to the user.

blog comments powered by Disqus