appears to be fluctuating between 2 sets of results,
and one set of results is once again missing our index page...
bout ready to just throw in the towel this time :(
Greg: There's not a new update today. Despite what some might say, we don't really randomize the results with each update. :-)
We're a bit curious about this, though. Did you issue your query against search.msn.com or against live.com?
On search.msn.com I run my favorite query and find my two favorite sites for that topic at #3 and #4 on page one of the results and #3 and #4 on page two of the results.
On page three, I find the #4 site from pages one and two is listed at #5 and two sites which have been missing for at least a month are now at #2 and number 4 on page three.
Everything else in the top fifty results is spam.
My domains was before on top positions for year before June ... after May update decreased down to 50-100 or 9999+ nowhere posisitons, but was always spidered and indexed .... today are again back at top positions.
It is nice ..because i'm not running spammy or content crap sites..
and I don't know why sites was dropped down before...
Now they are back ...huraaah
Results in my market have changed as well. My site finally made it back to page 1 after several months of vacation. But there are still a few obvious spam sites and non-relavent sites hanging around the top.
But overall it's not too bad. It's just that the bad results are VERY bad. If you ignore 4 or 5 sites MSN would be producing a solid first two pages of results. That's more than I can say for Y! right now and about the same as G.
I see much better results also. It looks like the good old search.msn.com with relevant results.
If I try to get past the subdomain spam present in a few of the industry queries I checked, then the SERPS on the first page are fine, but there is a lot more subdomain spam this morning.
Also, maybe I haven't had enough diet coke yet, but it seems like the first page of results is getting duplicated onto the second page too. Hey, I'm seeing double! *shakes head to unstick eyes*
Yep, not an entire duplication, but a lot of the same URLs on page 1 are on page 2. Couple that with many of the page 1 results being subdomain spam and that creates a problem.
Hope this helps msndude.
I am doing much better on MSN now. Before, I was getting next to nothing on a site with plenty of links so that was very weird. MSN has to be commended for fast indexing too; sites I added 4-5 days ago are already getting traffic.
A dial must have been tweaked somewhere. Hats off to msn this set of results is looking good.
You guys are correct; there was an update last night. Somehow I had thought this would happen later. Apologies for giving you wrong information earlier.
(But I'm VERY glad people seem to like it.) :-)
I am seeing something VERY odd. I have not read the previous posts in this thread, so forgive me if it is a repeat...
For a particular search query, I am seeing the EXACT same results on page 2 as on page 1. That is, after I hit the 'search' button, I see a set of results, then click to page 2, and see the same set. Some of them are even on page 3, but not all of them. I have cleared cookies, etc...Still see it. It is almost like seeing two different indexes, but for ALL of the urls on page one to be listed on page 2 is a bit more than coincidental, I think.
What is going on? I have seemingly lost communication with MSN also, so I don't know if my spam reports are getting dealt with. I still see the offending sites, so I would say no.
What garbage. No doubt the subdomain spammers are doing cartwheels again.
Too bad. MSN was making some progress but this is back to the old pathetic days of "oh, look a keyword subdomain with 100k blog comment links, goody."
Back to the drawing board.
Rob: That could be because your mailbox is full. Shoot me a sticky with the exact URL you're using to do the search because I cannot reproduce this here.
Steve: Yeah, I know; I can see them outside my window. :-) However, I don't think this particular release made a difference to them.
A large number of domains overflowing with subdomain spam you had diminshed or removed are back at #1.
Personally I don't see any other effect other than a dramatic return of subdomain spam.
I suppose that could be because the algo has turned up the "we love blog links" knob even higher than it was, but the irrelevant, useless spam is far worse than before, primarily on subdomains but also the expired domain type too.
One day when I was 15 I had thirteen screwdrivers at an open bar. The next day I resolved "don't do that again". Looking at this update you should have a similar thought... :)
When you go to page 2 from the first page of results, it is defaulting to "first=7"
I will sticky 2 url's
quotations: Thanks. I've managed to reproduce it. We're looking into it.
steve: Good idea. I'm going to go home and have a screw driver. :-)
MSNdude, I sent you a sticky with examples. I think this is the same issue that Quotations stickied you about, and that you were able to reproduce.
Msndude, it looks like some of our results have gotten lost from the index. Is this possible?
I am happy with the new results. Serp seems clean. Only a little bit too weight on keywords in URL.
I can say that not only I am happy with my results, but I am now using MSN as a preferred search engine: lately I found the information I was looking for only on MSN.
Guess it isn't a quick fix, since I still see it this morning. This seems like a major concern, to have the exact same results on page 2 as are on page one. Or, is it an isolated problem being caused only on my machine, and Quotation's?
It looks like with page one showing
and page 2 showing
and with only very minor variations between the two pages, that page 1 and page 2 both represent the first page of serps from two different indexes. Page one looks like what I saw BEFORE the update, page two looks like what I saw AFTER. Now, I can see both by just clicking to page 2. Is this accurate MSNdude? Will this be fixed today?
MSNdude, one more really curious thing about the search I stickied you is that on page three there is a result which is #7 or #2 on page 3, depending on how I get there [clicking on (3) from page one gives different results than clicking on (2) then on (3)]
That result is the same as the #4 result on page 1 and page 2 except it has index.html appended to it.
You may recall my sticky asking about duplicate content penalties due to the 10+ different addresses which that one page might have due to where it is hosted.
I have not seen the page listed multiple times with those different urls for several months but it looks like that might be back.
MSNdude >> Sorry about the late answer...I was travelling :) not meant to be rude.
Sooooo.....I was right! cooooool, I like that.
Ok now that fluctuations are gone I checked my usual test queries and I think that it's not bad at all.
Well, the Wikipedia cancer is spreading to MSN though.
Ok, that's probably the way to go, trusted sites should be emphasized (however trust is calculated and seeds deposit on quality sites).
More quality is going on, but still quite a few niche authority sites are missing vs subdomain spammers, my-keywords-domain or bloggers for example IMO.
Is the Strider anti spam stuff turned on already? :)
NOTE: I think that this algo needs a little understanding of singular/plurals in order to definetely leave Yahoo far behind (where it belongs) and move closer to GG.
Wow, there is something that MSN doesn't like about my website - traffic has pretty much gone to zero.
I didn't get a lot of traffic from MSN anyway, so it's not a big deal to me. I'm in a very competitive area that is often targeted by spammers. So I'm trying to cut through some very big players and some very sophisticated ones too.
I'm not looking for top spots, but the same kind of traffic I get from Yahoo - mostly long tail searches for very specific information. I have to admit that I concentrate more on writing content than SERP placement - something I'm pretty sure will pay off in the long term.
I looked back at the example I sent to MSNDude about a month or two ago. While it's not 100% fixed I do see some improvement. Good job.
Single word term brings up exactly 4 of what I would consider authority sites in the top 40 in the niche I am looking at. Only two in the top ten.
#10 is a client's site (good for them!) but it is far, far from worthy of that position (not to mention that they will not be pleased to be so highly ranked on the generic term, but that is another story).
Meanwhile, another site I am associated with that really does deserve to rank highly in this niche is nowhere to be found :(
Change the term to plural and an interior page from the latter site does show up in the top ten. But surrounded by double entries and various other garbage.
What I see is that it is still possible to rank based on listings in phony directories, sitewide links hidden in a small font period, etc.
On the other hand, there is not a bunch of scraper, MFA sites, and that is a definite imporvement.
Thanks, Billy. As I've said before, progress comes in small steps over time -- and includes some missteps -- but it's nice to hear that some results are starting to be visible.
It occurs to me that you can compare building a search engine to solving Rubik's Cube; at the beginning, it's very easy to make progress, but after a while it takes a lot of work to advance without messing up what you've already got. This is why some things that look like they ought to be easy end up taking a while.
I'm seeing spam issues as well. Consider yourself stickied.
As far as I'm concerned, the whole .info domain can go drown itself in Lake Michigan.
hmmm...seeing something entirely different than I saw this morning. Either a couple indexes are bouncing around or this is still cooking..
Looks like they rolled back at least the "let's make love to subdomains again" aspect of the update. Significantly better today, still...
<posting this in two engine forums in one day, gee, is it a theme>
More emphasis on PAGE TITLE. Spam, and especially subdomain spam, has a much harder time competing when exact match in title is weighted heavily. MSN-valued spam is often some expired domain with new random text content that is anchor-text blogged spammed for everything under the sun. You simply can't have the same effect though with a title that can only have 60 characters. Spammers also just throw up random words, while real sites use words like searchers.
More page title emphasis please...
It look's like MSN changed something back now.
My page dropped again to old positions ...to nowhere. traffic decreased 3 times .....
[edited by: xspace at 9:14 pm (utc) on Aug. 25, 2006]
| This 55 message thread spans 2 pages: 55 (  2 ) > > |