Gas station without pumps

2018 January 1

Blog stats for 2017

Filed under: Uncategorized — gasstationwithoutpumps @ 11:47
Tags: , ,

According to, my blog had 60,609 views and 35,810 visitors in 2017 (down from 68,443 and 42,499 in 2016).  The reduced viewership is not too surprising, as I posted only 114 blog posts in 2017, rather than the 200 of 2016.  (My biggest year was 2014 with 116,359 views and 69,179 visitors.)  I don’t know how much the decline in my readership is due to a general decline in how much people read blogs and how much is specific to my blog—I’ve not been able to find good statistics on the readership of “average” blogs.

Here are my year’s most-viewed pages (almost all of which are from previous years):

2017-01-01 to 2018-01-01

Title Views
Home page / Archives 19,956
Where you get your BS in CS matters 1,820
Making WAV files from C programs 1,601
Tools and parts list for Applied Electronics W2017 and S2017 1,387
Algorithmic vs. Computational thinking 967
Sum of probabilities in log-prob space 800
How many AP courses are too many? 632
Journals for high school researchers 606
Problems rewriting the Class-D amplifier lab 497
Pressure sensor with air pump 442
Getting text from Amazon’s “Look Inside” 438
Conductivity of saline solution 407
Where PhDs get their Bachelors’ degrees 401
Installing gnuplot—a nightmare 371
More on automatic measurement of conductivity of saline solution 365
Circuits course: Table of Contents 361
Why Discrete Math Is Important and The Calculus Trap 354
Lying to my students 341
Pressure and volume lab 334
labhacks — The $25 scrunchable scientific poster 315
Pullup vs. transimpedance amplifier 312

I think that there are several reasons that old blog posts dominate my views:

  • Most of my viewers (other than subscribers, whose loyalty I really appreciate) come via search engines. Of the 60,609 views, 33,202 were search-engine referrals (55%). Search engines will favor long-established pages that many people have clicked on in the past.  My next highest referrer is Facebook (which I don’t use) at 195 viewer—a tiny number in comparison.
  • My recent posts have been more specialized than older ones, so have a narrower audience.
  • I’ve been less active recently in calling attention to my recent posts on mailing lists and blog comments.

One good trend is that the most popular posts are now mostly contentful ones, rather than ones that are just links to other sites.

I don’t get any revenue from my blog (but I don’t pay anything for it either).  The clicks from my blog mostly go to other of my blog posts (1259), Digikey (1229), AliExpress (179), and LeanPub (175).  The average cost per click for advertising is 50¢–$2, so Digikey and AliExpress probably should be paying me, but they aren’t.  (Of course, the click rate would probably drop way down if I was being paid to push products, rather than just providing links to things I have bought and used or am thinking about buying.)

My top commenters (based on the last 1000 comments, so several years’ worth of comments) are

Commenter Comments
gasstationwithoutpumps 399
CCPhysicist 114
gflint 50
xykademiqz 31
mathproblemsolvingskills 26

My comments are mostly pingbacks caused by links to older posts for continuity, though some are replies to other commenters. I’d like for my comments to be about 25% of the total, rather than 40%, but I’ve not had much success in getting my lurking subscribers and followers to say anything. If each of my followers made just one comment a year, the number of comments would quadruple. Questions, corrections, and suggestions for blog posts are particularly welcome.

I admit to being somewhat envious of bloggers who have active discussions among their commenters—my readers don’t seem to have formed that sort of on-line community, perhaps because my posts are not open-ended enough or because I wander over many different topics rather than staying focused on a specific niche, so the readership may not share many interests with each other.

For those who have been commenting—thank you! It really helps me to know that people are reading my blog (and raw numbers don’t really do that—I can’t really tell whether viewers coming in from search engines are reading what I have to say, or just clicking on a link and deciding it was a mistake).


  1. I think you’re right that part of your drop in viewership can be explained by more specialized posts in 2017. Some of the electronics posts go over my head, but I’m starting to play with some simple arduino projects so maybe Ill go back and reread a bit. I enjoy the personal posts on cats, Santa Cruz, and your son a lot but also pay close attention when you discuss different aspects of university administration and education. Another type of posts that have been useful to me (but I don’t see often enough in your blog) are “opinion-pieces” on other peoples blogs. I found xykademiqz a while back from reading your blog and I’d like to see more long-form versions or rehashes of your comments on other blogs. As far as I know there’s no way of subscribing to your comments, which would be an alternative solution to this.

    I still don’t have a facebook account and hope the world moves on to a more appealing platform before I’m forced to conform. Or maybe blogs will have a renaissance.

    I always feel a bit presumptuous pushing my observations on smarter people when commenting on the blogs that I follow, so I’ll take your final paragraph as a reminder to overcome my imposter syndrome and give something back to the bloggers whose writing I enjoy.

    Happy new year!

    Comment by Ras — 2018 January 1 @ 14:24 | Reply

    • Thanks for delurking! I have about 225 “drafts” of posts that I haven’t written yet, many of which were to be responses to other things I saw posted. A lot of them are rather stale now, but I’ll try to go back over them to pick out some that might still be interesting.

      Comment by gasstationwithoutpumps — 2018 January 1 @ 14:55 | Reply

  2. I greatly enjoy your technical articles, and I enjoy as well your views on the ‘education industry’, altough it is somewhat different in the US than in other parts of the world. But that makes it interesting. I don’t think it is about the number of posts (yes, every post might count up to the views), it is more about if they are too much specialized or not.
    I feel with you about the ‘draft’ articles: 225 is a lot (I keep around 30-40 in draft state, and I feel I have to delete some as they might not be relevant any more).
    Happy blogging and Happy New Year!

    Comment by Erich Styger — 2018 January 1 @ 22:20 | Reply

  3. Happy new year, and thanks for blogging; I’ve enjoyed following your blog.

    I started following your blog as we prepared for home schooling around August 2010. I found you via the TAGFAM mailing list. By my recollection, you started blogging about Applied Electronics for Bioengineers around the time I started picking up electronics as a hobby, and your posts helped me learn.

    Professionally, I write software. And it happens that the software I write hasn’t exercised the calculus I took so long ago that I have forgotten most of what I did learn. This is of course a detriment to my new hobby. Besides the obvious limitations when designing anything new, it limits my ability to follow some of the posts end-to-end and it makes reading Art of Electronics hard work. I might have retained and developed calculus knowledge if I had picked up this hobby earlier! ☺

    Comment by Michael Johnson — 2018 January 2 @ 03:52 | Reply

    • I must admit that I have lost most of the mathematics I learned in college (though with an MS in math, I got a lot more than just calculus). I’ve also lost the foreign languages I studied in college. As with most learning, there is a use-it-or-lose-it phenomenon. Luckily, the amount of calculus needed for doing hobbyist electronics is fairly small: I’ve not needed anything beyond differentiation of exponentials and polynomials (well, maybe an occasional rational function).

      I find that my students don’t have much trouble with the little bit of calculus, though a lot of them have trouble with the supposedly simpler algebra, which you need to use much more extensively.

      Comment by gasstationwithoutpumps — 2018 January 2 @ 08:41 | Reply

  4. I found your blog in the middle of last year via search and have been visiting about once every week to see what new interesting things you are doing. It is relaxing to sit down after a long string of 14 hour work shifts and read what is going on especially since a large part of it is what I do in my off time. Happy New Years.

    Comment by trophosphere — 2018 January 2 @ 10:06 | Reply

    • Thank you for delurking. With 14-hour work shifts you still have the energy to do things in your off time? More power to you! What sorts of things would you like to hear about?

      Comment by gasstationwithoutpumps — 2018 January 2 @ 10:36 | Reply

      • I work on a 7-on/7-off schedule so I play catch up on everything I missed during the past week after I pass out for the majority of the first day off. I would like to hear more about medical electronic devices. I am a physician working in a hospital, and I find electrophysiology really interesting in regards to its implementation down to the circuit level. I have designed and built a couple of ECGs in the past and it is always exciting to see how others approach it in terms of dealing with electrode implementation, protection, noise filtering, waveform identification, and testing/certification.

        Comment by trophosphere — 2018 January 2 @ 13:05 | Reply

        • Ah, a 7-on, 7-off schedule makes 14-hour days more sustainable, though I suspect I would not be very functional after 7 14-hour days in a row. I will try to do some more EKG posts, though I’ve not done anything new with EKG circuitry lately. I’m particularly interested in trying one of the two-wire approaches, getting rid of the reference electrode.

          My electrode implementation has always been the simplest one: I buy commercial disposable EKG electrodes, for about 25¢ each. I like the Vermed ones for my class, but I’ve also used foam ones from BIO Protech. I use electrodes with snaps, so that I can connect to them easily with alligator clips.

          My protection circuitry tends to minimal: telling students to run off laptop batteries rather than power supplies and using series resistors in the signal leads. I’ve considered having the students use a pair of clamping diodes on each signal line to protect the electronics, but I’ve not bothered with that yet.

          Comment by gasstationwithoutpumps — 2018 January 2 @ 14:48 | Reply

          • Sounds great!

            I have been experimenting with designing and building capacitively coupled electrodes recently and it’s great being able to pick up an ECG waveform through nonconductive mediums such as clothing. That being said, offset (both static and dynamic) as well as noise pickup is certainly arduous to deal with. Guard traces and shielding are a must since biasing is certainly finicky given the small amount of contamination that can wreck a good electrode when the input impedance is >10G ohms via bootstrapping.

            Usually if I don’t want to have an external reference electrode I will tie the inverted Vcm signal to the sense electrodes in the device via 10M ohm resistors and that seems to do well enough. The CMRR loss ranges from 20-50dB though in my experiments using this method. It can be less if you decrease the series resistance or put more gain into the inverting amplifier but the magnitude of the waveform or stability suffers respectively. Maxim Integrated has a great single lead ECG front-end chip that uses a non-external reference electrode configuration called the MAX30003 which you may find interesting.

            My usual method of dealing with front-end protection is to have two series resistors on each electrode with a GDT (shunt) in between the two resistors. A clamping diode right at the end of the resistor pair before it goes straight into the instrumentation amplifier is a good thing to put in as well (as the standoff voltage from the GDT will still be around 75-150v). I find that if the first resistor is at least 51K ohms and the second one is around a tenth of that then the device should be able to withstand a 200J biphasic or 360J monophasic defib (99% of the time) without failing the energy reduction test – IEC 60601-2-49. Naturally, the lower the series resistor the less the CMRR degredation but one probably doesn’t want to go below 10K for the first resistor as the energy reduction test tends to start to fail around there.

            Comment by trophosphere — 2018 January 2 @ 15:24 | Reply

RSS feed for comments on this post. TrackBack URI

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.

%d bloggers like this: