<< William Empson: Among the Mandarins | Front Page | Wittgenstein Reads Empson >>
Friday, May 20, 2005
Who reads this stuff?
I just for the first time got some detailed visitor stats from our esteemed webmaster, Chris; it's sort of interesting. Turns out we have some wossname, Urchin, that gathers much data. Example. Every time you visit it sends a 'cookie' that crawls through your 'browser', into your 'keyboard', pricks your finger, gathers DNA and analyzes it. On the basis of hit patterns we've already isolated a gene for contemptsmanship. Tomorrow's culture war will be hi-tech, you see, so this is very exciting. Military applications. Genetically engineered culture warrior blog readers grown in vats, like weebles, with tremulous digits only for for clicking, clicking, typing, typing. (As Myles writes somewhere: give a man food, drink and the prospect of scoring points off his enemies and he will ask for no more.) Of course, it would be very 'unethical' to release this data, but rest assured we have our eyes on you, unbeknownst to you.
Anyway, here's the skivvy for the week of 5/8/05 through 5/14/05. (Plus one stat for the whole month of April that I specifically requested, because I was curious.)
A session is defined as: "a series of clicks on your site by an individual visitor during a specific period of time. A Session is initiated when the visitor arrives at your site, and it ends when the browser is closed or there is a period of inactivity." a pageview: "A request to the web server by a visitor's browser for any web page; this excludes images, javascript, and other generally embedded file types."
| Total Sessions | 10,306 |
| Total Pageviews | 56,290 |
| Total Hits | 64,908 |
| Total Bytes Transferred | 1.44 GB |
| Average Sessions Per Day | 1,472.29 |
| Average Pageviews Per Day | 8,041.43 |
| Average Pageviews per Session | 5.46 |
| Average Length of Session | 00:10:14 |
| Number of Unique IP's | 2,375 |
| Number of Unique IP's (for April) | 7,570 |
Oddly, these numbers bear no obvious relation to the simplified hit counter count I see on my control panel. (Perhaps it is just there to amuse.) The numbers also don't look like the ratios we get at J&B, where hits to sessions is more like 3 to 2; not 5 to 1.
At this point I put on my Trilling hat and say lofty stuff, but maybe you know me well enough that we can skip that part. I am rather proud that folks stay for a whole 10 minutes on average. Why, that's long enough to have a cup of coffee. Do you think the number of unique IP's - minus some allowance for spiders, bots and absurd google hits - is a decent heuristic measure of actual readership, approximately? So we have a few thousand actual individual human beings who visit at least occasionally? It's interesting that just a few thousand individuals can generate 50,000+ pageviews.
I have one question for you.
I generally take 'academic' to mean university level faculty or graduate students aiming to become such. But suit yourself, if you don't like that definition.
Please feel free to use the comment box to ponder readership, circulation, blogs, academic life and stuff like that.
Comments
I would like to know what the timeout is for sessions, since, AFAIK, there’s actually no way to tell when someone’s left the site. (Followed an outward-bound link? Nope, could be in a tab or new window.)
That would affect the average visit length, for one thing; also, if the first thing I do when coming here is open all the new posts and links to new comments in new tabs, and then read through them at my leisure, it will seem as if I’ve visited for a shorter period of time than if I read the same posts and comments sequentially (not that I would read them in parallel in the first case, but they would all be opened within a short span of time, and that’s the only way *you* can find out).
Or rather: I don’t really care what the timeout is, but I’d be curious.
Yeah, I thought of that myself, Ben. I dunno.
Hey everyone, Ben has started blogging again. Click his name and go visit. It’s mostly nonsense and mostly worksafe except for that dog sketch from a few days ago. And the link to the linguistics paper.
I don’t think my browsing practices would register well either. I keep about 18 tabs open in Firefox pretty much all the time; when I get stuck on the ol’ dissertation, I pop into Firefox, hit Ctrl-R and it refreshes all the windows. Then I browse through my pre-selected, pre-sorted reading material. Then back to the dissertation. Then back to Firefox. In other words, if it registers each of my refreshes--some of which occur hours apart--then I may well account for 50% of your hits.
I’ve often wondered how idiosyncratic browsing behavior influences hit-counts, but for quasi-philosophical reasons: if the approx. 250 individuals who read my site daily actually read it daily, that’s a significantly larger audience than I can assume for anything I ever publish, be it an article, a book, etc. These actual readers lack the institutional credentials of the hypothetical readers of my yet-to-be-published work, but there’s a way in which that doesn’t matter to me. As a teacher, as someone who actually wants to inform people of seemingly meaningless trivia about, say, Darwin or academic publishing, an audience is an audience is an audience.
I get reports on my own site from two different sources, and they disagree at a 3:2 ratio. There are also enormous discrepancies in the times reported—i.e., an average session of 15 minutes when none of the top 20 pages have an average visit as long as 5 minutes, and when the average session has 1.6 page views.
In short, lots and lots of of GIGO.
I think that it’s best policy to believe the lower numbers yourself, but to report the higher numbers to others. Unless you think that you’ll get caught.
A. Cephalous—I’ve effectively (not entirely voluntarily) chosen internet self-publication as my primary work. One of the odd things about it is that when you do that, while people do end up reading your stuff, your status in the pecking order remains unknown (except that some will insist that you’re not even worth mentioning at all).
I think that I end up writing / researching a lot of undergrad papers for people, for free. I do a better and cheaper job than the mills, that’s for sure.
As a favor I posted a full-bore 90 page academic paper for a friend. She’s been getting about 100-150 visits a month for a rather difficult paper on an obscure subject. (However, because of her topic, she gets a large number of visitors looking for porn, so you don’t know what the real number is).
In my case, the decision to self-publish felt pretty voluntary. (Except to the extent that it wasn’t voluntary to fall silent and feel suicidal when I concentrated on publication-by-others instead.) That probably made it easier. Turning down jobs always leaves one with a brighter outlook.
The question (as I’ve written and said so many times before) isn’t where you stand in a supposed pecking order. Insofar as what I write has a point, I do not want that point to peck. The question is whether there’s a better way to let what you need to do be accessed by people who’d like to access it. Journals and books can’t be run that way, which is what makes their production so depressing. The economics of web publishing better fit niche markets and midlists.
The comparison point for a posting shouldn’t be its hit-count for the week, but the number of readers the web-published piece has over several years compared to the number of readers an equivalent journal-or-newspaper-or-non-published piece has over several years. (Assuming, of course, that the post is something you’re actually interested in as opposed to something you’re doing to increase hit-count for the week.)
Most of my qualms about involvement with The Valve involve its (unavoidable?) interest in stirring up controversy, getting big view numbers, piling up the comments, moving entries off the front page, and so forth. These seem opposed to other goals of the site which interest me more.
But different people follow different muses.
Nah, Ray, we don’t want to stir up controversy. We just have to pretend to be delighted when it shows up uninvited. Nothing annoys a troll so much as “you got our invitation! come on in!”
One problem with being very low in the pecking order, or an outcast, is that 1.) people who might want to acknowledge your work might be afraid to—to may in the biz, acknowledging a self-published internet writer would be like acknowledging Drudge or Antiwar.com. Second, less well intended sorts might use your ideas unacknowledged, knowing full well that they can get away with it. (Even for those playing the game by the rules, this can happen).
I’m not a careerist nor the biggest egoist in the world, but if I come up with something good, Ido hope to get credit for it sometimes.
Steven Shapin’s “Social History of Truth” talks about how the gentlemanly leaders of research teams in early modern science got all the credit, whereas the names of their plebian assistants remained unknown, regardless of the magnitude of their contribution. Shapin’s book isn’t an expose’, but he does point out that some of the assistants of Boyle or Newton were actually scientists in their own right, and not just lackeys.
You got something to say, Ray?
For I detest controversy. Always have.
It’s pretty hard to avoid controversy on the internet, since it’s not a controlled-access forum. Some say that that’s one of the good things about it.
I apologize for the misrepresentation of individual motives. I felt very odd writing about “The Valve“‘s interest, as though “The Valve” was a living entity with motivations of its own. Yet I felt distinctly dishonest when I tried rewording it to attribute such “interest” to “the Valve contributors”. And yet again I do sense a conflict between something more Crooked Timberish (which perhaps interests more people) and something less Crooked Timberish (which perhaps interests me more).
But growing pains are natural, and most publications try out a number of different models before finding their groove.
i wonder how rss hits factor in. if a lot of your readership is tracking via rss clients that refresh frequently the pageview count might be a little inflated. then again those hits may not be counted as pageviews
re the ratio of unique ip’s to page.views: chris can probably break down for you the top 20, say, users in terms of site percentage. 1000 users might be generating 40,000 page views, with the other 6000 accounting for the remaining 10,000.
but any way you slice it, these are very good numbers for an academic blog--especially in the summer.
arnab, I don’t know if this is really best described as an academic blog, in terms of what is generating the hits. I suspect, based on the history of the site and on the poll (currently running 30 academic, 30 nonacademic, 14 sorta) that many of the people who read here are here because of Holbo, and followed him over from John & Belle Have A Blog. Perhaps some of these people simply like to read witty rambles rather than being intrinsically interested in academic subjects.
I do think the continuing aw-shucks tone of the discussions about hits, controversy, and so on is overdoing it a bit. I mean, the authors here did every tried-and-true method to increase controversy over the recent comments on other blogs: leaving snarky comments in response, writing defensive tracts, metaphorically shouting “Look at me! I’m not taking you seriously!” etc etc.
Rich, I think most of the people are here because of our commenters, who are among the finest in the blogogea.
I’d be interested in seeing the results of a poll about visitors’ genders.
rich,
i am interested in the stats because i am geeky that way. i don’t think there’s a whole lot of point to speculating on who exactly is reading, why they’re here, what paths they’ve followed, what they’re clicking on most or why. (though i am beginning to wonder what percentage of the valve’s hits are made by people who come here largely to carp and snipe.)
academic blog? blog by academics? i don’t think it really matters in the end as long as there are interesting things to read, and usually there are.
but i’ll say again that the comments feature in a blog is not the best way to have engaged discussion on anything. look at what happened to the discussion on this post about site-traffic. i rarely read the discussions following the posts anymore--i’m doubtless missing out on a lot but i don’t have the patience to negotiate the sniping that seems to infest many of them.
arnab