The Search Engine Experiment
The Test Results (So Far)
Most Relevant Results For Each Month
Most Relevant Results Over Entire Experiment Period
Sponsored Links
Comments
swoopz
swape!
Unfortunarely
you can do the search in your search engine of choice as well, and see what it gives you.. and then click that one's button and taint the results in that one's favor.. also.. i think i actually voted twice just now.. by accident
Re: Unfortunarely
"i think i actually voted twice just now.. by accident"
It doesn't matter if you vote twice, the system will disregard multiple votes.
Re: Unfortunarely
for me, Google is tops but yahoo comes a close second in terms of relevance. Good test though
- By webtrendz ltd at Aug 25, 2006, 11:54 am
- Reply
Spoiler
So Google is coming out top?!
You say in your spoiler that it isn't - isn't that a little unfair on Google?
Re: Spoiler
I don't see how it was unfair on Google, the test is blind so my predication can't have affected the outcome so far.
(In case you're wondering, preliminary tests indicated the most relevant was Yahoo.)
Re: Re: Spoiler
yes Yahoo does indeed come up higher for me on some queries but not for all.
Identical results?
I got identical results for two search engines, and they were both better than the third option... I voted for the first of the two, and got "MSN" instead of "GOOGLE". Clearly identical results should be judged as votes for both engines.
Results - UK or US
Are these results based on Yahoo.com/MSN.com etc or MSN.co.uk/Yahoo.co.uk?
- By Matt Whittingham at Nov 22, 2005, 4:15 pm
- Reply
Re: Results - UK or US
Based on .com results.
Good!
Good!
Re: Good!
I like this system. For example, give the one with the best results 3 points, second gets two, and third best gets only one point. Not only does this produce better results, you could also allow for ties, give the top ones that are tied two points each, and then if there was a tie for second, give each one point.
Use Borda Count
It's more fair to use this kind of voting system as you propose.
3 point to the best
2 point to the second best
1 point to the poorest
If two are equal (convert 3,3,1 or 3,1,1 to)
2.5
2.5
1
or
3
1.5
1.5
The system is called Borda count (http://en.wikipedia.org/wiki/Borda_count)
Here's an example on how the selected metod could realy make a big difference (applied on the Election held in in Taiwan (Republic of China) in 2000-03).
http://www.pvv.org/~nsaa/symmetrical_democracy.html
Regards,
Nikolai
odd system.
Wouldn't a system of user ranking be a more effective tool than simply "this one is better than the other two"?
For example, there should be a way for the user to state "Y is poor, X is better, Z is best" or something to this extent.
Google
Google is the best for me, and then Msn and finally Yahoo
What about Jeeves?
Wish Ask Jeeves could have been included... I gotta say they are equal to at least Yahoo, and WAY better than MSN search. ugh.
Re: What About ALtavista
Before it was very popular
- By Gift Baskets at Nov 30, 2005, 1:21 pm
- Reply
Re: What About ALtavista
Before it was very popular
- By Gift Baskets at Nov 30, 2005, 1:21 pm
- Reply
luminescence fading
Deliberately obscure topic, but Google produced two really relevant sites, while the other two each produced one plus one marginal one.
wheres altavista
When searching I usually find altavista has more relevant searchs, and beats google for web pages and images, but google beats for news. But why no altavista
Re: wheres altavista
We basically went by market share:
As you can see from the chart, in terms of usage AltaVisa is non-notable.
Also for anyone wondering, AOL wasn't included because it's powered by Google search.
Re: wheres altavista
Altavista is now owned by Yahoo. I believe they use the same search engine for both sites. They actually own a number of other search front ends but I can't recall them off hand.
wheres altavista
When searching I usually find altavista has more relevant searchs, and beats google for web pages and images, but google beats for news. But why no altavista
Thanks for doing this!
Actually, none of the results was as good or relevant as my favorite, Teoma, which I have used for several years. I know, Teoma is not a household name! But I have to do a lot of online research for my graduate degree, I've compared search engines (similar to this test), and Teoma always works best for me.
Re: Thanks for doing this!
Teoma is owned by AskJeeves - looks like they use the same engine.
Re: Thanks for doing this!
Teoma is owned by AskJeeves - looks like they use the same engine.
here we go again
haven't we been having this argument since the "Is Veronica better than WAIS?" days?
None of the above
You should add a choise for none of the above.. searching my words result in notthing really.
different results
did a search for "iran uzbek foreign relations" according to this thing the first two results for google were about tadjikistan - if you go to google's home page and do the same search you get different results- what gives?
Re: different results
Some variation to results is normal. This for two reasons: 1) Google use multiple datacenters each of which is likely to return slightly varying set of results, anyone one of these datacenters can respond to a given query, 2) Google geo-targets, meaning searching from one country/province may return different results than search from another, this is true to some extent even if you use the international google.com.
Personal Site SERP comparison
I tend to find that MSN gives me the most relevant results. Searching for "Charleston web site design" on MSN shows that my site is currently #4. On Yahoo, the same search puts me at #56, and finally Google puts me somewhere after #80. I stopped looking for my site after the 3rd of 4th page of Google's results.
MSN seems to go by semantic markup, headers match page titles, etc, and actual content. Yahoo and Google, seem to go by link count? I've only had my site up for about a month, so it's really too soon to tell. These assumptions are purely a guess. I have a lot of work to do on my personal site's SEO, so a month or so from now will allow for more realsitic testing.
MSN seems to have rewarded me for my clean markup, while others have not...yet at least. It's a work in progress, as always, and I am still studying the top three Search Engine giants to see what works and what doesn't. Here is what I found while researching my current page ranks on MSN, Yahoo, and Google SERPs.
Search Terms and Rank as of 11/26/05:
Search Term(s):
"charleston web site design"
Rank:
msn #4
yahoo #56
google #80+ (stopped looking on SERP 3)
--
Search Term(s):
"charleston search engine optimization"
Rank:
msn #3
yahoo #48
google #140+ (stopped looking on SERP 3)
--
Search Term(s):
"charleston custom illustration"
Rank:
msn #1
yahoo #2
google #1
--
Search Term(s):
"charleston logo design"
Rank:
msn #4
yahoo #21
google #70+ (stopped looking on SERP 1)
--
Search Term(s):
"charleston flash animation"
Rank:
msn #1
yahoo #12
google #300+ (stopped looking on SERP 5)
--
Search Term(s):
"charleston ad banners"
Rank:
msn #5
yahoo #9
google #210+ (stopped looking on SERP 3)
--
Search Term(s):
"charleston photo restoration"
Rank:
msn #1
yahoo #27
google #240+ (stopped looking on SERP 4v)
--
Conclusion:
Someone looking for Web site design related services in the Charleston, SC area would be better served by MSN, then Yahoo, then Google.
In the process, I also noted the placement of the local competition for the services searched for above, and the results were very similar. Google returned the most non-relevant results out of the three SEs tested
Results for "charleston flash animation" were perhaps the most shocking.
As I optimize, and retest, I am writing a detailed summary, that will be posted on the (soon to be released) "Odds n' Ends" section of my Web site.
I though you folks might find these early results interesting.
Re: Personal Site SERP comparison
I would assume that your site would do better on the Yahoo and Google Local Searches:
http://local.yahoo.com/
http://local.google.com/
There you can focus more on the location of the business.
MSN results
Its true that MSN is most responseive to on page optimization. But that makes it easy to fool the search engine as well. Just by adding some dummy content, any ameatuer site can rank above a serious site. The only reason why this has not been abused much is that MSN doesnt provide a large number of visitors anyways.
You may be ranked low in Google, but I am sure google will still be bringing in more visitors than MSN.
Google & MSN
I did a few searches, and got good results with Google and MSN. About 50/50. MSN is getting alot better and seems to crawl and update more than Google :-)
- By Tim Linden at Nov 26, 2005, 9:21 pm
- Reply
Mix in some more obscure engines
You write that, in selecting which search engines to include, you "basically went by market share." I would think that this experiment is a great way to strip out the effects of market share and show how search engines really stack up, regardless of popularity. So why the focus on only the biggest?
Right now this experiment is only confirming what I already expected: the big three are fairly similar, with Google taking a slight edge.
- By Bob Malamphey at Nov 27, 2005, 9:59 pm
- Reply
Test
Fantastic Idea this, something really the search engines should keep on top off.
I am not supprised google is still top, allthough its good to see the others have finaly catched up a bit.
I have to agree with what was said about though.
Placeing the results in order of best to worst, and giving identical points for identical results would improve the stats, imo.
- By Thomas Wrobel at Dec 1, 2005, 12:26 am
- Reply
Test
Fantastic Idea this, something really the search engines should keep on top off.
I am not supprised google is still top, allthough its good to see the others have finaly catched up a bit.
I have to agree with what was said about though.
Placeing the results in order of best to worst, and giving identical points for identical results would improve the stats, imo.
- By Thomas Wrobel at Dec 1, 2005, 12:29 am
- Reply
flaw in the test design
In my opinion this test is flawed in one major way. When the search results come up, there's no way for the test taker to click on a link to see how relevant that site is and then to back out and vote on the results. To encourage users to vote on results without looking at the actual sites just precipitates information illiteracy. That's just my two cents.
MSN title scraper not working
Great site. Looks like there's a bug in your MSN title scraper, though -- perhaps they changed their HTML recently? When I tried several queries, I got truncated or nonsense titles for all the MSN results.
Search Engines
GOOGLE appears to do the job competently, while my experience with MSN is downright poor. Yahoo, for my taste was too confusing and I decline to use it.
Search Engines
GOOGLE appears to do the job competently, while my experience with MSN is downright poor. Yahoo, for my taste was too confusing and I decline to use it.
Search Engines
GOOGLE appears to do the job competently, while my experience with MSN is downright poor. Yahoo, for my taste was too confusing and I decline to use it.
Search Engines
GOOGLE appears to do the job competently, while my experience with MSN is downright poor. Yahoo, for my taste was too confusing and I decline to use it.
AllTheWeb
I know that this is the engine behind others and is probably not known as well as most.
Over the years I have found this to be better if searching for things in Europe - it seems to have less of a US bias.
Re: AllTheWeb
i agree.... google comes tops for me
- By webtrendz ltd at Aug 25, 2006, 11:55 am
- Reply
Google
Google wins
- By San Sebastian at Jun 29, 2006, 2:43 pm
- Reply
I thing Altavista is good, but I yse google eh
Well I thing, altavista us a good search engines, because, I use Google, but in many times U could not to access a different pages, and wiht altavisya I find doferent opcions fot a same site
- By Ivan Torres at Aug 23, 2006, 6:56 pm
- Reply
I thing Altavista is good, but I yse google eh
Well I thing, altavista us a good search engines, because, I use Google, but in many times U could not to access a different pages, and wiht altavisya I find doferent opcions fot a same site
- By Ivan Torres at Aug 23, 2006, 6:57 pm
- Reply
Re: I thing Altavista is good, but I yse google eh
hmm, i just tried it and well altavista did not fare well for me.
Chating
Hi all;
Just need to make friends
Chating
Its the fiest time for me to set online but i really hope to make friends
english
chating
- By shuai zheng cheng at Oct 15, 2006, 8:16 am
- Reply
Submit Comment