That is precisely what Google’s Backward Links do NOT indicate, which is one of the reasons that makes it an ideal measure of a blog’s importance, and one of the reasons Technorati’s “Authority” number is so useless as a measure of anything but “the acknowledgement of one blogger by another.”

This is all explained in great detail in the S&F Top 50 methodology to which I refer you for more information.

I don’t understand the problem with blogs with active comment sections at all. An active comment section is not an indication of either substance or the absence of substance; it is an indication of interest and the quality of that interest is, ultimately, subjective.
——————————–

There is no problem with an active comment section. I simply gave an example of how distorting it can be to use subscriptions and unique visitors as a rankings measure rather than Backward Links. In any case, the case is moot. As I’ve already pointed out, there exists no publicly accessible Web-wide data set with this data, and so there’s no point even discussing using subscriptions and unique visitors as a measure for rankings.

ACD

I don’t understand the problem with blogs with active comment sections at all. An active comment section is not an indication of either substance or the absence of substance; it is an indication of interest and the quality of that interest is, ultimately, subjective. What it sounds like, A.C.D., is that you’re interested only in the numbers which reflect dialogue between bloggers, or more precisely the acknowledgement of one blogger by another (which is what backwards links indicate) and not in numbers which reflect readers in general, the vast majority of which, I would hope, are not themselves bloggers.

As a matter of fact, if I were to learn that the majority of my readers were bloggers, I’d shut down immediately. I’m not interested in doing exercises on mirrors, I’m interested in making music and anything else that might make music-making better, more interesting, more moving, and sharing this.

That being said, were there to exist a trustworthy, Web-wide, publicly accessible data set from which to draw such subscription and unique visitor counts for each blog in the population under consideration (in this case, classical music blogs), it would be a neat thing to report them alongside the Google Backward Links counts to provide a more complete picture. Unhappily, no such publicly accessible data set exists, and so, as pointed out in the S&F Top 50 methodology, the whole question is really moot.

What could NOT be done by a blogger whose own blog was part of the population under consideration, even if such a publicly accessible data set existed, would be for that blogger to construct some composite formula using both Backward Links counts and subscription and unique visitor counts, and present the single number resulting from such a formula as a ranking, for then the blogger, no matter how honest, would leave himself wide open to the charge of an “obvious conflict” of interests. Assuming both required data sets were publicly available, such an undertaking would most decidedly have to be done by a totally disinterested and qualified third party trusted Web-wide as is Google.

ACD

While I do think that subscription numbers are meaningful, an indication of more serious interest in a blog, they represent only a small portion of blog readership altogether. The number of subscription service users is just too small; in fact, the percentage of internet users who even know what a subscription service is, is probably miniscule.

The more I think about this, the more I’m convinced that the best solution would be a voluntary use of the same counter by all blogs that wish to be considered in combination with other data, among which Google backwards links, subscriptions, and technorati rankings might be added in decreasing weights.

I don’t understand why people don’t use Feedburner for rankings, as it tells you all of the people who subscribe to your blog.

That reasoning would make sense if the basis of the rankings was some homegrown ranking formula by which metric — mirabile dictu! — the blogger’s own blog always placed near the top in the rankings. But as explained in the S&F Top 50 methodology, the S&F Top 50 uses a metric (Google Backward links) that’s an accepted Web-wide standard as a measurement of a website’s relative importance — a (the) primary determinant of Google PageRank, the cornerstone and foundation of the entire Google empire; that which turned Google from a dorm room enterprise into the trusted Web giant it has become. The S&F Top 50 is nothing more than a compilation; a reporting in a single location of Google’s Backward Links counts for a selected population of blogs, in this case, classical music blogs, and the counts are individually verifiable by anyone when the S&F Top 50 is published each quarter. One does not have to take the S&F Top 50’s word for it. And so, no “obvious conflict.”

Even if the Google-based rank is both the qualitatively best (of which I am not convinced; the volatility I noted was greatest in the two or three days immediately after your ranking was posted and has since decreased somewhat….
————————————

As explained in the latest S&F Top 50 (i.e., 3rd quarter) rankings, I didn’t publish the rankings until after the counts for my test sample of 30 blogs, among which was your blog, had remained stable for a full 5 days indicating that the Google update for September had propagated throughout all of Google’s thousands of servers. Your blog is unusual if not unique in the S&F Top 50 list of eligible blogs as it’s published in Germany but written in English (only English language classical music blogs are eligible for listing in the S&F Top 50). If you saw the kind of volatility you note, I can only guess (and it’s just a guess on my part) that your accessing Google’s German rather than U.S. servers was the cause as the German servers might have gotten the final update numbers some few days after all U.S. servers had been updated.

As always, if you ever discover a publicly accessible statistical data set more appropriate than, and as noise-free and statistically “clean” as, Google’s Backward Links for use as the basis for the S&F Top 50 rankings, I would be more than interested in hearing of it.

ACD

I do not include you blog in my list because its focus is contemporary and experimental music. I do include links within articles whenever relevant and I believe that my outgoing link count is, consequently, significantly higher than my incoming count.

I have plenty of regard for your blog; it’s a good example of a passionate and articulate music lover being passionate and articulate, but, given my limited time and the thirty or so blogs closer to my own professional interests which I subscribe to, I am only able to read your blog when a theme relevant to my own concerns is flagged at one of the two reblogging sites which I do scan daily and list in my sidebar.

As a matter of fact, I recently saw such an item at Blognoggle, went to your site, and consequently on my blog responded to something in which you wrote which I happened to find both remarkable and naive. I naturally included a link. (Heck, for that matter, I once even included your name in a meme forwarding!)

This discussion is silly. I actually defended your criteria for inclusion to Prof. Bourland. In passing, I happened to mention the obvious conflict which is going to to color the reception of your ranking. No one should be in the business of publishing rankings in which their own blog is also under consideration. Even if the Google-based rank is both the qualitatively best (of which I am not convinced; the volatility I noted was greatest in the two or three days immediately after your ranking was posted and has since decreased somewhat; the combined metric which I suggest would be a considerable inmprovement as it would actually include a count of every single site or page visit) and verifiable (is there an easy way to find historical Google page rankings?), this is best done by a disinterested third party.

Personally, the most interesting metric is the number of blog-related emails I receive. These, even more than on-line comments are the most rewarding part of blogging, with the discussion continuing, in depth, on a one-to-one basis. At present, with about 20 serious messages a day, I frankly don’t think I could handle much more traffic!

My above, “a publicly accessible statistical database”, should have read: “a publicly accessible statistical data set.”

ACD

While one may well argue for or against a given method, there is no mistaking the fact that the particular method of the four singled out by A.C. Douglas is the one which places Sounds & Fury in the highest position, and as long as that is the case, the ranking promoted by Douglas will be viewed in those terms.
————————————–

Once again, your contention is preposterous.

Of the four methods used in the post you linked to (which post, perhaps not so BTW, was a petulant entry by the blogger in response to my refusal to use the Backward Links count for his entire website rather than his blog for ranking in the S&F Top 50), the ONLY method almost universally accepted across the Web as a measure of relative importance (as opposed to relative popularity) is incoming links, and Google’s Backward Links count is the most noise-free and statistically “clean” count of incoming links available publicly which is the ONLY reason the S&F Top 50 uses that measure for its rankings. That you hold S&F in rather low esteem (I know that because you don’t list S&F in your blog’s blogroll), and can’t understand why others don’t as reflected in Google’s Backward Links count for S&F, is your personal issue, sir, not a fault of the measure. Or are you suggesting I choose a measure for the S&F Top 50 rankings on the basis of that measure showing S&F to be low in rank in order for the rankings to be above suspicion?

How preposterous an idea is that?

Since reading the Sounds & Fury ranking for the first time, I have periodically checked my own Google backwards link number. It has ranged from 890 to 1610 in the past few weeks. I honestly don’t believe that my blog should rank particularly high on a classical list … but I am nevertheless suspicious about the volatility of those numbers. It is far from clear what kind of sampling window is at work in these numbers.
—————————————–

Which only goes to show that amateurs ought not to mess about with things of which they’ve little understanding.

The Backward Links count of most blogs would have shown variation over the just past couple weeks or so because Google was in the middle of its October update of Backward Links which began 21 October (Backward Links counts are updated by Google approximately monthly), and which update hadn’t propagated throughout all of Google’s thousands of servers and stabilized. THAT’s the reason for “the volatility of those numbers.”

Once more, if you ever discover a publicly accessible statistical database more appropriate than, and as noise-free and statistically “clean” as, Google’s Backward Links for use as the basis for the S&F Top 50 rankings, I would be more than interested in hearing about it.

ACD

http://classicalconvert.com/2008/04/the-top-50-classical-blogs-using-4-different-methods/

While one may well argue for or against a given method, there is no mistaking the fact that the particular method of the four singled out by A.C. Douglas is the one which places Sounds & Fury in the highest position, and as long as that is the case, the ranking promoted by Douglas will be viewed in those terms. A ranking page organized by a more disinterested party would be less likely to raise such a controversy.

Since reading the Sounds & Fury ranking for the first time, I have periodically checked my own Google backwards link number. It has ranged from 890 to 1610 in the past few weeks. I honestly don’t believe that my blog should rank particularly high on a classical list (especially since I don’t do much opera chat and wander, otherwise, far from the ranch), but I am nevertheless suspicious about the volatility of those numbers. It is far from clear what kind of sampling window is at work in these numbers. While I can assume that readership is generally low for most of the blogs in questions, I suspect that this is a lot more noise than signal in these numbers, with only Alex Ross’s blog in any way accurately reflected by the statistics.

What, then, might be a better statistic? I think it would have to be a composite of several elements. If blogs would voluntarily agree to using the same counter, that would be one useful factor in measuring readership. The precise relationship between page and site views would have to be figured out. Links should definitely be included, but I do have the impression that the music blogging community is quite divided in their attitudes and practices towards linking, with many bloggers purposefully not including links in either articles or sidebars, which constrains the utility of this measure somewhat. Some statistic about search results including particular blogs might be useful. I think subscriptions should be a small factor as well.

So, with the caveat that quanitfying readership among all but the most-read blogs will necessarily include considerable noise, it would be interesting if some disinterested party — that is, not a classical music blogger him- or herself — would organize such a page.