In my anecdotal experience, it's every bit as good as google, and I simply did not notice a difference in the results when I switched.
It must be down to what people search for, or maybe how heavily Google has customized their personal results to fit their behaviors. I mostly use web searches for looking up technical documentation, and 5-second tasks like weather forecasts and unit conversions.
This has happened probably a handful of times but enough times that I would have been stuck in the water without Google, unfortunately. Only one of those handful of cases was significant (I needed that information for work, not a simple curious search) but Google saved the day.
Some of it may be due to years of growing accustomed to building searches subconsciously around Google which is unfair to DDG, but it's hard to say definitively.
1. Index size - Google spider visits more places than Bing spider. This is why Google delivers better results for compiler error strings, serial numbers etc.
Query examples where Bing/DDG completely miss the mark:
SA1DP2CF - a serial number
s2 034 253 033 C - audi rs2 exhaust manifold part no
Denninger Str. 96, 81925 München, Germany - what is on this location?
2. Intent/Semantic
Query examples:
tesla 00001 - who owns first tesla? Bing is completely lost
original elite alien name - looking for thargoids, Google tells me that right away, DDG #8
3. Localized
Query example: hiking trails - Google gives me relevant local results based on my location (IP based), DDG doesn't (although it is aware of my IP-based location in the same way)
Getting these right is incredibly hard as you literally need Google size infrastructure to get this right, which even Bing doesn't have.
Second thing that is hard to catch up for any Google competitor is the rest of Google infrastructure users rely on - Browser, Email, Meet, Docs, Youtube, Android, Drive... Even if you get the search right but the user has to go back to Google for the rest of services, Google has a chance to get the user back.
IMO attacking Google starts with developing a product that does not insult user's intelligence, is innovative and does things Google can't, explores different business models (as ads inevitably lead to conflict of interest - optimize for user experience or advertising revenue?) and offers infrastructure that displaces (at least some) Google services.
"kicad convert wxpoint to millimeters"
Note how many "Autocad" entries clutter the DDG results. Google gets kicad in every single entry on the first page.
My biggest frustration with search engines is "I gave you the terms so USE THEM. Don't be helpful and omit terms you think aren't relevant."
I still miss AltaVista's interactive cluster diagrams where you could see the web of links and pick the relevant one.
If DDG simply implemented that single feature, I would never go back to Google.
The problem is that, like Microsoft in the 90's, Google will simply copy your feature if it takes off. So, there is no point in innovating in a visible way in search. That's the problem with a monopoly.
Here you go. #2 on the list.
https://www.runnaroo.com/blog/the-search-engine-hacker-news-...
Thanks for this. I'll flip it to my default search for a while and see how it goes.
I know Swisscows and Carrot2 both do some topic clustering, maybe those would be interesting as well.
If you are searching technical info like pdf manuals, datasheets, etc.. I find DDG works better than Google, because most companies aren’t buying to be first result, so the manuals show up first.
If you are looking for Stack Overflow type of content “how to config X on linux using Y”. The !g flag becomes handy.
This is what I do as well but regarding results on technical info I consider DDG/BING/Whatever way inferior to Google.
Disclaimer: I switched to DDG and only very infrequently need to resort to google. I happier not being tracked.
Either way, I use Searx[1], which is a search engine aggregator that you can self-host or use one of the public instances[2].
I would be unsurprised if it's true. The reality that DDG is going to have is that search is almost inherently personal, as an artifact of the imprecision of language and the lack of surrounding context. It applies to some searches more than others, but something like "what is vanilla?" Has a whole variety of answers. What is the substance vanilla? What is made out of vanilla? There's probably 80 different software frameworks that use vanilla in the name. Who knows how many companies have vanilla in their name. Oh, or it could be that you mistyped Manilla, which has its own set of results.
Personalization let's you cut down on that. Maybe this person searches for Python stuff a lot, so there's a good chance the frameworks are what they're looking for. Or, the inverse, this person looks up recipes all the time, they probably know what vanilla is so don't show a full page of "this is where vanilla comes from" results.
Lacking that context, they have to assume the same context for everyone, so either everybody gets a full page of web frameworks or everyone gets a full page of frameworks; DDG has no way to tell which you're probably more interested in.
Yeah, if you aren't used enough, it's not worth trying to target you for manipulation. All that really demonstrates is (1) they use different algorithms than Google, so optimizing against Google's algorithm doesn't work on them, and (2) they aren't used enough to be worth gaming separately.
You and me, as consumers, shouldn’t be forced to use a worse product just to try to balance the market, because we can’t compete with Google either. That’s the point of anti-trust regulations.
I'm sure there's some site the obfuscates results and could probably answer the question a little more scientifically.
My process: think of something I want to find, think of all the keywords I can, type them into the search engine.
Google helps me find what I'm looking for faster than Bing or DDG.
On the other hand, there are often various good local search providers who have bangs that you can use directly, with better SERPs than Google for the relevant queries.
Even with those features on I find constantly outdated information these days.
Yup that article from 2012 Techcrunch with the broken links is really helping me Google.
Disclosure: Currently running an alternative search engine. However I legitimately don't like Google search or Google's practices in the search market, so much so it actually prompted us to throw our hat in the ring.
Edit: I just made the search again, "Techcrunch submission form" The first result for me is from 2006, not 2012, https://techcrunch.com/2006/02/01/how-to-submit-your-company...
Yeah, Google has overtly for, IIRC, about a decade said they are focussing search improvements on personal and even predictive search, so, sure, disable that and the quality goes down.
But that just illustrates that, while there is an obvious advertising incentive for data collection, it actually is relevant to delivering useful search results.
[0]Only 22% of consumers say they are okay with having their information collected indiscriminately by websites to provide a better experience.
Which means by that metric google abuses 88% of it's users unless they go through the means to turn the features off.
Informed consent, access and ownership are all pillars a modern site experience should have when it comes to data collection -- Google fails here and so does DDG in the opposite direction by just never asking which means they can't do any personalization.
The winner here is going to be someone who can provide the better ethical and functional experience. I firmly believe in the next 15 years we will see litigation in the US around data privacy just using California as the canary and Europe recently voted to get even stricter than the GDPR.
That isn't a quote from the referenced link. The quote is "Only 22 percent of consumers agreed that they would be willing to hand over personal data to improve experiences." - which is quite a different thing.
I'm not happy to hand over personal data to improve experiences. But I'm very happy for Google to track what searches I do to improve their results.
Your right it's not a quote I didn't quote it, notice the lack of quotes. I'm also not sure how you are splitting hairs over "information collected indiscriminately" which would be a superset including personal data.
I get https://techcrunch.com/2019/11/20/submit-a-guest-post-to-ext... as the first link
Absolutely. I switched from Google to DDG and Qwant a few weeks ago and was surprised that I didn't see much differences.
But after discussing it with colleagues that were switching back to Google it became apparent that it was because I was siloing Google to its own browser, erasing cookies etc that made it as "inefficient" as the other search engines.
App Store search says.
I use Google with site:apple.com to search all of these instead.
If Apple search capabilities are a hidden asset, then they're very well hidden.
Even indexing the web is a different problem now than in 2000, users expect the search results that are SEO optimized for google.
That said, in general, I appreciate the greater average quality of apps in the App Store.
What's an open source ripoff?
A competitor who wasn’t shackled to that business model would have a lot of freedom to innovate.
Most people seem to disagree with you about the raw results improving. They certainly don’t seem to be for me.
First 4-5 links are ads, the next might be relevant. Page 2 onwards is definitely irrelevant.
A decent search engine is not worth the massive amount of 12 billion dollars. The advertising platform Google build is!
Apple 1) can't recreate that 2) it shouldn't as advertising is not in their DNA and opposite of their privacy core values.
Apple can build a search engine (or just switch to Bing) but they wouldn't be able monetize that for $12000000000.
They certainly have been hiring people who know how to build search tech.
>Apple has hired Google’s chief of search and artificial intelligence, John Giannandrea
https://www.nytimes.com/2018/04/03/business/apple-hires-goog...
This is true, but not in a way you think. Conventional wisdom tells us that you're gonna need completely different approaches across the stack (not just technological but organizational as well) when a problem size grows 10x, and the web has grown about 100x from 2000.
Wouldn't it be ironic if the DOJ case fell apart because Apple released it's own search engine?
It's not 2000 anymore. Building a web scale search engine no longer requires novel distributed systems work, and the most viable search algorithms have strong open source implementations. It's still a very hard problem to get right, but you can create an "interesting" poc in a few weeks. Recent work on unspervised content based recommenders can help reduce click stream data requirements.