Brave doesn’t censor, and Brave doesn’t inject bias into Search results. We go far beyond paying lip-service to this commitment too; we’re developing Goggles to prevent this issue.
What’s Happening?
The query Peter Santilli returns one set of results, while the query Pete Santilli (note the spelling of the first name) returns another. Mr. Santilli’s podcast uses the “Pete” variant, along with numerous other sites. As such, “Peter” may be expected to yield less-relevant results (without the engine trying to be clever, and conflate Pete and Peter).
Both forms returned Mr. Santilli’s podcast URL at, or near, the top of the results.
How can we Improve?
There are a couple of ways the query results could potentially be improved on our end; Peter could be treated as a common variant of Pete, and a couple more sites could be scanned more frequently for content (i.e. rumble.com, and locals.com). But it’s important to note that Brave’s results aren’t being skewed by another other than name-confusion, and breadth of indexing across the Web.
Update: I just took a closer look at locals.com, wondering why we weren’t listing the many videos on that domain. It turns out none of the content is immediately visible when you visit the page (I had to click a “see the content” button). And, once you jump to the content, it is dynamically loaded via scripting, rather than provided in the initial server response. These 2 patterns alone guarantee that most crawlers will not be able to discover the content.
On Neutrality and Search Results
There’s something we need to consider when discussing neutrality, and what to expect in honest search results. Search engines discover the Web via basic means. We’ll find domains and sites when they are linked-to or referenced from elsewhere. This networking effect will almost always benefit larger sites, with larger user bases.
As a Search Engine scours the Web, it crawls over the content of various pages and attempts to learn about the subjects covered on each. To assist search engines in this learning process, authors of sites can use carefully-crafted HTML tags and more to property identify and mark key portions of their pages and publications. This is an engineering task, and requires a trained individual to do properly.
Given these realities, Search Engines are almost always going to display results from larger, well-funded, and more established outlets. This is very likely going to be the case for popular topics and events, since the “competition” for those topics will be more fierce. If ideologically-inclined sites happen to employ authors, engineers, and content-distribution managers to write, code, and cast their content out onto the Web, they will tend to appear higher in results pages.
How you can Help!
If you would like to help us make sure more properties are indexed (such as Rumble, Locals, etc.), please consider opting-in to the Web Discovery Project. It’s our best hope in competing with Google and Bing (the source of DuckDuckGo’s search results).
Anytime you see results that raise an eyebrow, or give you cause for concern, please do let us know. While we are not (and will not be) injecting bias into our results, there may be ways that we can improve our discovery/indexing/crawling process to result in a better, more replete view of the Web.