![\"Visualization](\"https:\/\/www.lesestunden.de\/wp-content\/uploads\/2016\/05\/visualization.png#\")
<\/a><\/figure>\r\n\r\n\r\n\r\n\r\n\r\n\r\n\r\nData<\/h3>\r\n\r\n\r\n\r\n
As is always the case on the internet: there\u2019s no central directory where you can find all blogs. So I went looking for blogs about books and used the following sources:<\/p>\r\n\r\n\r\n\r\n
- \r\n
dmoz.org: The good old DMOZ is always a solid source. Carefully maintained blogs should be listed there and, lo and behold, you can indeed find many blogs there.<\/s> now offline<\/li>\r\n\r\n\r\n\r\n- bloggerei.de<\/a>: Another directory with many literature blogs. In return, blogs have to embed a button and are then added to the directory.<\/li>\r\n\r\n\r\n\r\n
- Twitter<\/a>: I found many bloggers on Twitter \u2014 through Twitter\u2019s recommendations and through other users\u2019 lists. I had over 1,000 Twitter accounts with a bibliophile background on my list. Thanks to a script, I queried all URLs automatically and didn\u2019t have to click through 1,000 profiles.<\/li>\r\n\r\n\r\n\r\n
- Beyond that, I clicked through numerous blogrolls and saved the URLs linked there.<\/li>\r\n<\/ul>\r\n\r\n\r\n\r\n
The most labor-intensive step was cleaning the resulting link list. The list contains exclusively<\/em> book bloggers \u2014 no authors, no blogs about writing, no literary magazines, no bookstores, no publishers. Truly only blogs that write about books, publish reviews, and share their reading pleasures<\/strong>. I also removed all blogs whose last post was older than three months (roughly: the latest post couldn\u2019t be older than December 2014).<\/p>\r\n\r\n\r\n\r\n
This left 585 book blogs<\/strong>, which I examine in more detail below.<\/p>\r\n\r\n\r\n\r\n
The Toplist<\/h3>\r\n\r\n\r\n\r\n
You can find a toplist of all book blogs, sorted by popularity, at the following link: https:\/\/www.lesestunden.de\/topliste\/<\/a><\/p>\r\n\r\n\r\n\r\n
The first result is a toplist \u2014 that is, a ranking of all book blogs. But based on what ranking? It seemed sensible to me to use a ranking built on two foundations:<\/p>\r\n\r\n\r\n\r\n
- \r\n
- How well connected is the blog within the book-blogging scene? A blog that is read by many other bloggers should appear higher up, since it is popular within the blogosphere.<\/li>\r\n\r\n\r\n\r\n
- How popular is the blog outside the book-blogging scene? How likely is it that Google or another site will refer a user from the wider web to the blog?<\/li>\r\n<\/ul>\r\n\r\n\r\n\r\n
I used two values that are easy to determine, though they are of course not the final word on the matter. I had to make a compromise to keep things feasible.<\/p>\r\n\r\n\r\n\r\n
- \r\n
- As a ranking for popularity within the book-blogging scene, I used the number of links<\/strong> from other blogs. If a blog appears in many other blogs\u2019 blogrolls, it is very popular within the blogosphere. To measure this, I simply searched the home pages of all sites for the blog\u2019s URL and counted on how many blogs it appeared.<\/li>\r\n\r\n\r\n\r\n
- As a ranking for broader popularity on the open web, I used the
MozRank<\/s><\/strong> OpenPageRank<\/strong>. It ranges from 0 to 10. The more popular a link is \u2014 i.e., the more pages that link to it \u2014 the higher the value.<\/li>\r\n<\/ul>\r\n\r\n\r\n\r\nI normalized both values to between 0 and 1 and took the arithmetic mean. That is the ranking displayed in the toplist.<\/p>\r\n\r\n\r\n\r\n
Visualization<\/h3>\r\n\r\n\r\n\r\n
You can see a visualization of all 585 book blogs above. Click the image to open a very large view (note: it takes a moment to render and load).<\/p>\r\n\r\n\r\n\r\n
You can also display where an individual blog appears in the visualization and who links to it and whom it links to. In the toplist, just click the little bar-chart icon. If a blog links to you, it\u2019s shown in green; if only the selected blog links to the other blog, it\u2019s shown in red; if both blogs list each other in their blogrolls, the other blog is shown in blue.<\/p>\r\n\r\n\r\n\r\n
Two blogs are connected whenever one blog lists the other in its blogroll. The blogroll must be on the homepage and not tucked away on a subpage.<\/p>\r\n\r\n\r\n\r\n
What can we conclude from this visualization, besides that it\u2019s pleasant to look at? At a glance, it becomes clear that blogs are linked together in a very diffuse way. There aren\u2019t a few blogs that everyone links to; rather, each blog follows a handful of others, and the selection isn\u2019t limited to a small number of sites. That makes sense: everyone picks the blogs that match their reading tastes, and there doesn\u2019t seem to be a dominant consensus. In my view, that\u2019s a lovely result.<\/p>\r\n\r\n\r\n\r\n
I don\u2019t want to withhold a second visualization either. I asked myself whether there was any cluster formation \u2014 i.e., whether there are \u201ccliques\u201d that are tightly networked and interact heavily among themselves. A graph should shed light on this:<\/p>\r\n\r\n\r\n\r\n
![\"Graph](\"https:\/\/www.lesestunden.de\/wp-content\/uploads\/2015\/03\/graph.png\")
<\/a><\/figure>\r\n\r\n\r\n\r\nTo identify clusters in a rough way, I ran the force-directed algorithm<\/em> on the graph and waited until it stabilized. I repeated the process several times to ensure the result was similar each time. This algorithm minimizes the distances between nodes and tries to render them with as few overlaps as possible. (Note for nerds: since this problem is at least NP-hard, there are only approximation methods; force-directed is one of them.)<\/p>\r\n\r\n\r\n\r\n
Using this approach, subgroups of blogs that follow each other closely didn\u2019t crystallize out. You can see a strong core with numerous blogs that follow each other, around which there are blogs that are less tightly connected (which may be due to activity and quality), and at the very edges, the more weakly connected blogs (unknown or less active blogs). But no cliques, no elite clusters of bloggers who keep to themselves or are strongly bound to one another.<\/p>\r\n\r\n\r\n\r\n
Statistics<\/h3>\r\n\r\n\r\n\r\n
Apparently, we\u2019re dealing with a true network here \u2014 one without significant focal points and with genuinely even connectivity. But is that actually reflected in the numbers? Yes, it is:<\/p>\r\n\r\n\r\n\r\n
0.68%<\/b> of blogs appear in more than 40 blogrolls
1.2%<\/b> of blogs appear in more than 30 blogrolls
4.1%<\/b> of blogs appear in more than 20 blogrolls
14.7%<\/b> of blogs appear in more than 10 blogrolls
85.3%<\/b> of blogs appear in fewer than 10 blogrolls<\/p>\r\n\r\n\r\n\r\nFrom these numbers you can see that the average blog appears in about 10 other blogrolls. Only 14.7%<\/strong> are listed in more than 10 blogrolls. A mere 1.2%<\/strong> appear in more than 30 blogrolls (those could be called popular).<\/p>\r\n\r\n\r\n\r\n
How strongly does a blog\u2019s age correlate with its popularity? If you use the number of blogrolls in which a blog appears as the metric, the result is clear: not at all. The correlation coefficient is only -0.08<\/strong>, practically zero. (A correlation coefficient of -1 or 1 indicates a very strong correlation \u2014 i.e., a significant relationship between age and popularity; a value of 0 indicates no dependency.)<\/p>\r\n\r\n\r\n\r\n
How strongly does a blog\u2019s age correlate with its MozRank (or even Alexa Rank)? Only moderately: just 0.3<\/strong>. Older blogs therefore have no notable advantage over newer blogs. The average age of the 85<\/b> blogs that appear in more than 10 blogrolls is 3 years.<\/b> So anyone who believes that being well-established automatically brings recognition is mistaken.<\/p>\r\n\r\n\r\n\r\n
I determined blog age using the Wayback Machine from archive.org<\/a>. For me, a blog\u2019s age is the point in time when the first snapshot was created. The average age of the 503<\/b> blogs for which I could determine an age is 3 years.<\/b><\/p>\r\n\r\n\r\n\r\n
44%<\/b> of blogs use WordPress; 41%<\/b> use Blogspot. This result doesn\u2019t surprise me at all.<\/p>\r\n\r\n\r\n\r\n
The blog with the best Alexa Rank is http:\/\/www.transatlantik.de<\/a><\/b> with an Alexa Rank of 52,546<\/b>.<\/p>\r\n\r\n\r\n\r\n
The blog with the best MozRank is http:\/\/www.buchhexe.com<\/a><\/b> with a MozRank of 5.7802074443185<\/b>.<\/p>\r\n\r\n\r\n\r\n
Measuring can be misleading<\/h3>\r\n\r\n\r\n\r\n
Like any set of statistics, these should be interpreted with caution. I didn\u2019t falsify them (as you can see from my own blog\u2019s placement in the toplist), but there are a few inaccuracies I want to mention:<\/p>\r\n\r\n\r\n\r\n
- \r\n
- A link \u2014 i.e., a relationship between two blogs \u2014 is counted here whenever Blog A links to Blog B on its homepage<\/span>. I search the homepage rigorously for the link. Spot checks showed that this almost always corresponds to the blogroll, but it\u2019s not guaranteed. Some blogs keep their blogroll on a dedicated subpage, which is not captured here.<\/li>\r\n\r\n\r\n\r\n
- A link in a blogroll is a solid indicator that Blogger A reads Blog B, but of course Blogger A is likely also browsing many other blogs \u2014 perhaps via WordPress, Bloglovin, or a feed reader. That remains unaccounted for here. The results I present may be interpolated to reading behavior, but that isn\u2019t certain.<\/li>\r\n\r\n\r\n\r\n
- These statistics say nothing about actual traffic or, in particular, the quality of visitors. My blog, for example, has many high-quality visits \u2014 i.e., sessions where visitors stay on the site for a long time and actually read the posts. An unknown blog at the end of the toplist could therefore be far more valuable than a blog with a high MozRank that, for instance, gets tens of thousands of clicks via Google Image Search that all bounce immediately.<\/li>\r\n\r\n\r\n\r\n
- Determining age via the Wayback Machine is of course imprecise. Archive.org is usually quick at indexing a site, but it\u2019s certainly not an exact method for determining a site\u2019s age.<\/li>\r\n\r\n\r\n\r\n
- Another source of error is omitting social networks. I often land on book blogs because someone retweeted something on Twitter or pointed out something interesting. Some people may rely exclusively on Twitter for discovery. Connections via Instagram, Facebook, Twitter, Pinterest, etc., are not considered here.<\/li>\r\n<\/ul>\r\n\r\n\r\n\r\n
Despite these inaccuracies, I consider this analysis representative, as it also matches the subjective impression I formed while exploring and getting to know this blogosphere.<\/p>\r\n\r\n\r\n\r\n
Conclusion<\/h3>\r\n\r\n\r\n\r\n
Book bloggers are excellently networked with one another. Each blogger picks out the blogs they find interesting. The weight is evenly distributed. There are no blogs that you could truly call very popular.<\/p>\r\n\r\n\r\n\r\n
There are no subgroups of blogs. Everyone is connected with everyone; there aren\u2019t small factions that are tightly interconnected only among themselves.<\/p>\r\n\r\n\r\n\r\n
The average age of the blogs is 3 years, and older blogs are not significantly better known than younger literary blogs.<\/p>\r\n\r\n\r\n\r\n
From my perspective, this offers a genuinely interesting insight into the world of book bloggers. In a way, the result is rather lovely: this is how the spread of information and news on the web should look \u2014 not centered on a few sites, but diffusely distributed among many people. The fact that information relevant to a large number of bloggers still spreads is shown by the discussion about earning money with one\u2019s blog \u2014 it felt like everyone took part in that discussion.<\/p>\r\n\r\n\r\n\r\n
Note: Did I forget any blogs? Let me know in the comments and I\u2019ll add them!<\/strong><\/p>\r\n\r\n\r\n\r\n
The figures shown here refer to the analysis from March 7, 2015. I will update the toplist regularly. The values that appear on https:\/\/www.lesestunden.de\/toplist\/<\/a> may therefore differ from those in this post.<\/strong><\/a><\/p>\r\n\r\n\r\n\r\n
Update March 9, 2015<\/h3>\r\n\r\n\r\n\r\n
I\u2019m truly overwhelmed by the response to this post and delighted that so many of you are enthusiastic about these statistics and the list. This was actually meant to be a one-off, but since there\u2019s demand for a collected overview of book blogs, I\u2019m happy to keep maintaining and expanding the toplist. From time to time I\u2019ll also remove inactive and closed blogs \u2014 that\u2019s quite easy, since I also parsed all RSS feed URLs.<\/p>\r\n\r\n\r\n\r\n
Thanks to your abundant feedback, I\u2019ve adjusted and corrected quite a few things since the initial publication. Here\u2019s a list:<\/p>\r\n\r\n\r\n\r\n
- \r\n
- You sent me an astonishing number of blog URLs, so the toplist now includes almost 650 book blogs. Many thanks to all of you for your lively participation.<\/li>\r\n\r\n\r\n\r\n
- Links for some Blogspot blogs were not counted correctly, since those can have multiple URLs. A Blogspot blog is reachable at beispiel.blogspot.de, but also at beispiel.blogspot.com, www.beispiel.blogspot.de, and www.beispiel.blogspot.com. This is now taken into account.<\/li>\r\n\r\n\r\n\r\n
- In the visualization for an individual blog, the color now correctly indicates the relationship between blogs. If a blog is red, the selected blog links to it but there is no link back. If a blog is green, only that blog links to the selected blog. If it\u2019s blue, both blogs list each other in their blogrolls.<\/li>\r\n\r\n\r\n\r\n
- In this post I had listed the wrong blog with the best Alexa ranking. Alexa ranks evaluate domains, which of course makes no sense for Blogspot. All blogs hosted on WordPress and Blogspot therefore can\u2019t be considered for the Alexa ranking. Hence, that figure is more of a sweetener and not particularly meaningful.<\/li>\r\n\r\n\r\n\r\n
- Duplicate blogs were removed where I found them. A few mistakenly included publisher blogs were also removed.<\/li>\r\n\r\n\r\n\r\n
- Due to the many new blogs, the values changed a bit, but not the fundamental conclusions. I\u2019ll therefore leave the figures here as those from the initial analysis on March 7, 2015.<\/li>\r\n\r\n\r\n\r\n
- <\/a><\/li>\r\n<\/ul>\r\n\r\n\r\n\r\n
Update February 4, 2016<\/h3>\r\n\r\n\r\n\r\n
As you know, I keep the toplist up to date and occasionally add improvements. Here\u2019s a brief overview of what changed over the last year:<\/p>\r\n\r\n\r\n\r\n
- \r\n
- New book blogs are continually being added to the toplist \u2014 either because someone submits a URL (via the link at the top of the toplist) or because I stumble upon a blog myself. The list had grown considerably in the meantime but has now shrunk back to just under 900 blogs.<\/li>\r\n\r\n\r\n\r\n
- I\u2019ve found that after about a year, many blogs are abandoned, no longer maintained, or completely offline. It\u2019s now checked automatically whether a blog has published a new post within the last 6 months. If not, it\u2019s not included in the list. As soon as posts appear again, it\u2019s automatically re-included. If a blog is completely offline and unreachable, it\u2019s removed and must be re-submitted manually. This way the toplist stays current and you won\u2019t find abandoned blogs on it.<\/li>\r\n\r\n\r\n\r\n
- I experimented with a service that evaluates backlinks and tried to incorporate those numbers via an API. I ended up discarding this. On the one hand, MozRank already provides a fairly good measure of a blog\u2019s popularity and the number of web links to it; on the other hand, backlinks are only of limited significance. Often they merely indicate that the blog owner is very active in commenting and thus placing links to their own blog. Perhaps it also depends on the backlink provider.<\/li>\r\n\r\n\r\n\r\n
- A small bug distorted the ranking. I fixed it, which caused bigger shifts during one update.<\/li>\r\n\r\n\r\n\r\n
- Something that has been available for a while \u2014 and that I want to highlight again \u2014 is the OPML export. You can export all entries with their respective RSS feeds in OPML format via this URL: https:\/\/www.lesestunden.de\/toplist\/index.php?nav=opml<\/a>. This allows you to further process the list, import all blogs into your feed reader (which is truly crazy and impossible to keep up with), or process them automatically in your academic work.<\/li>\r\n<\/ul>\r\n\r\n\r\n\r\n
If you discover a blog that\u2019s still missing from the list, feel free to send it to me \u2014 here in the comments, via the submission function, or simply by email.<\/p>\r\n\r\n\r\n\r\n
Update May 2, 2016<\/h3>\r\n\r\n\r\n\r\n
The toplist is now better integrated into lesestunden.de and is no longer a separate site. I\u2019ve removed the option to generate a visualization for individual blogs \u2014 given the number of blogs, it\u2019s very cluttered anyway and, as far as I could tell, rarely used. The visualization of the blog network and the graph are now only linked in this post.<\/a><\/p>\r\n\r\n\r\n\r\n
Update May 13, 2016<\/h3>\r\n\r\n\r\n\r\n
Since May 9, 2016, Moz has no longer been rating Blogspot and WordPress blogs and always outputs a MozRank of 0. That makes MozRank useless for this toplist, because of course there are well-known blogs hosted on Blogspot or WordPress. Therefore, MozRank no longer factors into the calculation, which is a pity, as it provided a good measure of a site\u2019s general popularity. The ranking is now based solely on the interlinking of the sites with each other \u2014 which is only half the story and reflects a blog\u2019s popularity within<\/em> the book-blogging scene.<\/p>\r\n\r\n\r\n\r\n
I haven\u2019t found a good alternative. Alexa Rank or Semrush, for example, also don\u2019t support subdomains. If you know a provider \u2014 or where I can (programmatically) obtain a meaningful ranking \u2014 please let me know. I also submitted a support request to Moz, but since I\u2019m not a big-paying customer, I don\u2019t expect much to come of it. The next Moz index update is reportedly in one to two months.
We\u2019ll see if that changes.<\/del> Support responded and confirmed it\u2019s a bug. In the next index update on June 21, Blogspot and WordPress subdomains will again receive a rank. I\u2019ll include it in the toplist from then on.<\/a><\/p>\r\n\r\n\r\n\r\nUpdate June 28, 2016<\/h3>\r\n\r\n\r\n\r\n
There\u2019s another major update. Moz now correctly rates Blogspot and WordPress blogs again. MozRank is a good metric for how well-known a site is outside the book-blogging scene, and I\u2019m glad I can use it again for the list. In my view, this yields a more meaningful ranking. Naturally, this update significantly changes the ranking of some sites.<\/p>\r\n\r\n\r\n\r\n
Until now, there was also the limitation that I only searched for blogrolls on the homepage. I\u2019ve changed that too: subpages are now searched as well. Some blogs keep dedicated pages for their blogroll. Of course, this search still has some fuzziness, since it\u2019s automated.<\/p>\r\n\r\n","protected":false},"excerpt":{"rendered":"
When I first started blogging, I only had a vague idea of the book-blogosphere. I already had a few literary bloggers in my RSS reader, but I\u2019d never had a proper overview. Which book blogs are out there at all, what do people write about, how well connected is this blogging community, and what do … <\/p>\n
- As a ranking for broader popularity on the open web, I used the
- As a ranking for popularity within the book-blogging scene, I used the number of links<\/strong> from other blogs. If a blog appears in many other blogs\u2019 blogrolls, it is very popular within the blogosphere. To measure this, I simply searched the home pages of all sites for the blog\u2019s URL and counted on how many blogs it appeared.<\/li>\r\n\r\n\r\n\r\n
- bloggerei.de<\/a>: Another directory with many literature blogs. In return, blogs have to embed a button and are then added to the directory.<\/li>\r\n\r\n\r\n\r\n