Robert_Charlton - 9:44 am on Sep 22, 2013 (gmt 0)
For this search, basically all of the "major" search engines... (here's a politically-correct sample: Google, Bing, Blekko, DuckDuckGo)... whether they integrate videos or put them on a separate search tab, concluded that users want videos, and that, within those videos, for this query, YouTube overwhelmingly predominates.
You might say that dogs+cats+funny has for years virtually defined YouTube. This particular search is skewed way, way, way more to YouTube than many searches you might come up with. Some of the engines cluster results and allow you to expand a cluster. Others expand to various degrees on their own. There's lots to be discussed about the degree to which Google is currently expanding and, IMO, how it's still testing how well a linear ordering works.
IMO motives aside it's partly lazy engineering, better the devil they know.
I don't think it's lazy. They've been testing an incredible number of interface variants, and this is one that will allow them to triangulate all of the other sort orders they've come up with. I think they've simply been crunching numbers for a long time on this one, and it probably was and is the subject of a lot internal debate at Google. Much has been discussed about it here, and I can understand the frustration anybody feels on searches like this one if they're not YouTube or whatever else Google returns that predominates.
A few news sites with stories on funny cat and dog moments.
Blekko (which you can count in to give you a psychological explanation of humor when you include "funny" or "jokes" in a query) returned a Huffington Post article among its web results, but that in turn linked to several YouTube videos. ;)
A quick note about Bing Video Search, not exactly on topic, but IMO related to what Google is returning in its Universal results. A friend, not a search professional, but knowing that I'm interesting in both search and video, very recently mentioned the new Bing Video Search as by far the best video search he'd seen, particularly noting that he felt it gave exposure to many video search sites that previously he hadn't known about or that he'd needed to browse.
Here it is, and I tried it for this search...
It displayed roughly 150 video thumbnails for the query... a lot of scrolling... so I only had the patience to get an overall impression... but from what I can see, maybe all but five of these 150 videos were from YouTube. I don't think that this means that Bing has got a deal with Google. I think it accurately suggests the preference, or likely preference, for the YouTube videos in this particular search.
I should also note that I think that video search is seeded by topic and by simplistic word matching, but it is then driven a lot by popularity, time spent viewing, perhaps some personalization picked up by related click-throughs, etc... so popularity is not an unfair way to rank these. It's not like keyword relevance can dig into the subtleties of dog+cat+funny. Bing's video results suggest to me that, in terms of popularity, Google's results probably aren't all that biased.
That doesn't mean we have to like these results, or that there isn't a better way to sort and display them, but I do feel that Google has been working through something that is genuine, but getting the data has been taking a long time.
<offtopic> With regard to that last video, a search for Google.com, it's proof that sex sells, that popularity is a ranking factor, and perhaps that many people have bad taste. In any event, it's got 7,614,912 views; and the video's poster has 665 videos in her channel. Beyond the obvious crass jokes that come to mind, I think there's something to be learned from this particular result about how video can get you on page one for extremely competitive terms. I wouldn't simply dismiss it.</offtopic>