Educated Guesswork

As I said yesterday, the discovery of a missing datum threw some of the conclusions of recent posts into doubt. As of now I have updated those posts with correct information. The affected posts are listed below:

All of these posts have been amended and may now be read with the confidence that you bring to anything you read on the Interwebs.

 

Warning: The following contains moderately technical details which some readers may find boring and others will find overly simplified.

The problem began almost a year-and-a-half ago, when I read the sequel to the historical mystery Roman Blood by Steven Saylor. (This compelling novel is a fictional retelling of the events behind Cicero’s first major trial, which the Roman advocate published as Pro Roscio Amerino. Saylor effectively blends facts from the trial record with speculation well founded in the known ancient history of urbs Roma, in a novel that is obviously a long-planned labor of love.) Back in 2015 I began tracking which books in my library I had actually read — not in the actual book record, but in a related table, “Notes” — by simply recording that a book was ‘read yyyy-mm-dd‘, where yyyy-mm-dd was replaced with the actual date I finished the book. Though this table has a “Read” field (which the UI displays as a checkbox), that binary field is not easily searched, so it was the text field which was later used in the query identifying finished books. This “Notes” text field was also used for sorting the books in the order completed, where the date format made such ordering simple using an alphabetic sort. During the most recent hundred books I finally created a specific field (two, in fact) for enumerating the book read order, and had to go back and update all of the previous records with that information, but I still used the “Notes” table for recording the actual date read. So my mistake was the following: after reading the short story collection Eight Strange Tales after finishing the historical mystery mentioned above, I read the next Steven Saylor novel in that series, and forgot to record its completion in the “Notes” table. From that mistake flowed all of the dire consequences which followed in its wake.

 

As I said, I later began recording the ordinal for each book read, and had to retcon that data for all the earlier books. (In fact, I record two datapoints, one for every book, and a second for those non-comics books completed, since I exclude comics and graphic novels from my books read totals.) But since I forgot to ‘note’ (hee-hee again) that I’d finished Arms Of Nemesis, the 2nd novel in Saylor’s series about Gordianus the Finder, not only were my ordinal counts off by 1, but this novel did not appear in any of my queries regarding finished books. (As an aside, I note that Arms Of Nemesis was the 150th book read since I began tracking this data, and the 117th non-comic book read. Thus the vast majority of the last hundred books read were affected by this problem.) It is only by fortune that I was able to reconstruct this data, however. I began rating my books as I read them, and I rated the second Saylor novel. This rating meant that the “Date Modified” field had the timestamp correlating to the date I finished the book, and thus I was able to reconstruct the data. However, if I had later changed any datum about that book — which I have been doing on a large scale, as I normalize genre information across the database — I would have changed the “Date Modified” field and would have been unable to determine just when I completed that novel, knowing only that I’d read it after the first in the series. So we dodged a bullet there.

 

Fortunately, I realized that this book had been read before making any other changes. Thus order has been restored to this small, petty universe. I am attempting to normalize and correct information in the database — particularly genre information, so that I can have a standard set of categories and also call out the specialty categories which apply to my own personal collection of books. These include such groupings as “Secret Societies” and “Shakespeare Wacko”. The off-the-shelf software I use for cataloguing my books is nice because I can simple enter new books by ISBN, but the information is garnered from online databases which have their own issues and idiosyncrasies. For a while, for example, UK databases were being searched first, so prices in pounds sterling (£) were often entered automatically. (I also strive to enter the price actually shown on the book, not the price I pay, which is often different given my predilection for used books.)

 

With this major correction — and my heartfelt apologies for confusing the issue with my mistake — I will not return to compiling the list of the last hundred books read so that it may be perused by you, my impatient readers waiting with baited breath. I’m trying to learn how to use the <div> tag, so…. Wish me luck!

 

One final note: The pace of reading I maintain has decreased significantly, and I may have to take steps to improve by rate of days per book. At my most recent speed, it will take me approximately 140 years to finish all the books in my collection. Besides raising issues of mortality, I need to reflect upon the size of my library and its composition. (Since my book reading rate excludes comics, and since a significant portion of my books are such, I may need to investigate this and other aspects of the problem further; as an example, my total reading rate becomes 4.58 days per book if comics are included.)

 

1 Book per 4.58 Days (including Comics)

 

Time to Finish Collection: 106.5 Years

 

Estimated Finish Date (EFD): December 7, 2124

 

So … I need to start taking better care of my health if I am to see that 22nd century Pearl Harbor Day.

 

ALERT! ALERT! ALERT!

Previous post(s) about reading 200 books have errors. Please disregard. More details follow, with more to come.

Well, that just shows that data analysis is only as good (at best) as the underlying data…

Whilst scrubbing data in my books database — which is something I do, and you should do it too — I discovered that I had read a book almost a year-and-a-half ago, but had neglected to mark it as read in my database! This means that the book I had previously designated as my 200th book, Double Cross Purposes by Ronald Knox, was actually the 201st book (assuming I find no further errors in the underlying data), and that my earlier blog post about the last hundred books I’ve read was in error. Preliminarily, it now appears that my actual 200th book was The 2nd Scientific American Book Of Mathematical Puzzles and Diversions by (of course) Martin Gardner.

I apologize for the error, for which I am entirely responsible. I hope that you can forgive, though I realize that absolution must come from a higher source. After I have verified the now corrected data, I will make the analysis once more and present the amended results. I am glad that at least my mistake was discovered before I had completed the list of the books read, and promise to deliver that shortly after I’ve published the updated data. I hope that none of my readers have lost money on any wagers based on my earlier blog post. Again, I crave your forgiveness and will strive to not make such a mistake again.

As I just mentioned, I recently finished reading 200 books, counting from when I commenced recording such information in my books database back in June of 2015. Herewith I present some findings of no ultimate interest, derived from an analysis of the underlying data found in that database. We will be considering the last hundred books (i.e., books #101 – #200) read, where books in the “Comics & Graphic Novel” category do not count towards that total.

The first stat is that already given in the original notification of my 200th book read, which is the fact that I took — on average — just under six-and-a-quarter days to read each volume. This compares with 4.83 days per book read for the first one hundred. (Incidentally, the total average for reading the complete 200 books becomes slightly over 5-and-1/2 days per book. Put another way, three years and ten days were required to read the total set.)

1 Book per 6.24 Days

I am surprised to discover that my ‘comfort food’ reading (discussed in my report on the first 100 books) had subsided somewhat. Whereas the last report showed a 3-to-1 ratio of fiction (of all types) to nonfiction works, now the ratio is 4-to-3, with History books dominating the nonfiction ‘category’. (The Categorizing Imperative is its own fraught issue, one perhaps worthy of discussion another time.) Specifically, 57 books were some type of fiction, leaving 43 nonfiction works. Once again, mysteries dominated the fictional works, with the actual breakdown as given below:

Books Read by Genre

Mystery	30
Fiction	13
SF & Fantasy	14
Nonfiction	29
History	14

Or, for those who prefer charts…

Nonfiction, of course, is a term covering up any number of sins. Excepting the fourteen History tomes read, the complete breakdown of nonfiction books in this last hundred is as follows:

Nonfiction Breakdown

Children’s Books	2
Film	2
Indians of North America	1
Militaria	1
Mythology & Folklore	2
Philosophy	4
Poetry, Drama, & Criticism	6
Psychology	1
Religion & Spirituality	5
Science & Math	3
Wacko	2

Apparently I was getting real literary in my reading lately…. I’ll also note here that I read 7 books in the Comics category, not included in the count towards 200.

As I did in my last report, I present a time series chart showing when these books were read, broken down by genre.

Speaking of time, the earlier six-and-a-quarter days per book is, of course, an average. Reviewing the data I found that there were some periods where much more time passed between completed books. Several books were finished more than ten days after the previous volume, and in two instances more than twenty-five days passed between successive books. The greatest time between books occurred when I was unable to maintain anything like a steady reading pace during the last Christmas season, and forty-eight days passed between the completion of Conversion by Erik Routley (finished November 15, 2017) and Fallacies and Pitfalls of Language: The Language Trap by S. Morris Engel (completed on January 2, 2018). As an aside, I’ll say that I cannot recommend the first slim volume by Routley enough (or anything by the hymnology expert).

Of course, the date I finish a book often has very little relation to the time the book was started. Indeed, there are any number of books by my bedside that have been started and restarted, with varying degrees of progress made within. I have thought of tracking the date I first commence reading a book, but have decided that that way lies madness — or, if I have already crossed that threshold, even deeper insanity.

I generally decline to comment on the relative merits of the books I read, though I’m happy to give recommendations if you want them. I don’t like everything I read, and indeed this last hundred had a few stinkers among them. I do rate the books on a 5-star scale, though I have yet to find something so execrable that it deserves a single star. Thus a perfect mean across all books would be 3.5 stars if an equal number of possible ratings were given. This last batch of a hundred books had an average rating of 3.9, which is pretty good, but has dropped from the average rating for the first hundred, which was exactly 4*. (*for books actually rated. Not all books were rated when I first started tracking my reading. And while I’m in this parenthetical note, I will point out that these ratings include the comics I read, though there was little change due to those (The last hundred, for example, has a total average rating of 3.89 for all books, and a average rating of 3.87 if the comics are excluded.).).

Average Rating for Books Read: 3.9 Stars

Another change to my record-keeping has been the addition of page counts for each book entered. This was begun during the last hundred, so the records are as yet incomplete, but over 90% (including comics) of these books do have this datapoint, so I can state that the average book length (using those books with that data for the calculation, with the error that implies) was 226 pages, or 236 pages if comics are excluded. My total page consumption (including comics, again) was north of 21,990 pages. We’ll check back in after the next tranche with (one hopes) more complete data.

Average Book Length: 236 Pages

I will return in the near future with the complete list of books read.