Perfect potato-chip book

       Isaac Asimov, Martin H. Greenberg and Joseph D. Olander. 100 Malicious Little Mysteries (1981) A wonderful potato chip book. The stories are mostly 3 to 4 pages long, with a few outliers. Common plot twists feature poetic justice, radical shifts in point of view, or sudden recognition of a minor clue’s meaning. Each could be redone as a full-length novel, which makes this book a reference work of sorts. Most are little gems of narrative art, the whole book a master class in how to put much meaning into a very small frame. A keeper. ***

Short stories by Rankin

Ian Rankin. Beggars Banquet (2002) Collection of short stories, several of which are about Rebus. Like Rendell, Rankin enjoys imagining psychopaths. Makes for page-turning, but too many of these tales leave a bad taste, despite slatherings of poetic justice.
  And the title omits th apostrophe. Because including it would spoil the cover design? Probably.
  This collection would be a poor introduction to Rankin. **

Regression to the mean: The real reason Apple shares are falling

The BBC ttoday has a story explaining why the punters are unwilling to pay high prices for Apple shares:
     1. Investors are worried about iPhone sales.
     2. Apple's high prices could leave it exposed if the economy sours.
     3. Investors aren't confident - yet - about Apple's services business.
     4. Apple reflects broader market concerns - like US-China trade tensions.

All very plausible, as is always the case when explaining why an outlier is regressing to the mean. The stock market is a complex, chaotic system. Any given indicator will "perform" better than average some of the time, but it will also perform worse than average. But on average, it will perform -- ta-da! -- on average. What else? So any excursion into well above (or below) average levels will not last. One can always find plausible reasons why the shift towards the average happened now rather than earlier or later, just as one one can find plausible reasons for the non-average performance that preceded it.

Some of those reasons may even the operative ones. But we'll never know for sure which ones triggered the shift, nor can we use those plausible reasons for predicting when or how the next major shift will occur. That's because the indicators are themselves averages. Worse, we don't have causal models of economic behaviour, we have only statistical ones. This is so even though (especially so, actually) when buyers and sellers are algorithms that weigh many different factors to arrive at a price. Human or not, the buyers and sellers include current prices and recent price changes in their calculations. That means that the price calculations are feedback loops. Hence the chaos in the system, and the inevitable regressions to the mean.

Addendum (2018-12-12): I should perhaps emphasise that explicability is not the same as predictability. Events that result from the interplay of many factors cannot be predicted.The best one can do is study past performance and calculate probabilities. Basing a choice on probabilities is always (ta-da!) a gamble. Hence th attraction of horse racing and other objects of wagers. However, once some event has occurred, one can usually point to those factors that played a major or (sometimes) decisive role. Hence "Hindsight is 20-20". And the pointlessness of trying to lay blame on agencies and people tasked with protecting us against calamities.

Clouds over the North Channel

This morning around 10:30am. You're looking at Lake Huron.

Dr Hu, no not the Time Lord, but a cool guy anyhow (links)

Check out Dr Hu on research into animal movement. Interviewed on CBC's Radio One Quirks and Quarks. Will be available as a podcast by the end of the day.

Update 2026-02-14: No longer available. However, his book on animal movement is available from several sources: 

How to Walk on Water and Climb up Walls: Animal Movement and the Robots of the Future

• David Hu

Turn of the Screw: Rybczynski's history of the screwdriver

   Witold Rybczynski. One Good Turn (2000) Rybczynski was asked to write an article about the most important tool of the millennium. Some research showed that pretty well all the tools expected in a good toolbox had been invented much more 1,000 years ago. Eventually, he settled on the screwdriver, and wrote the essay. But the research was incomplete. This book recounts the research.
The screw is quite old: the Greeks and Romans had it. But the screwdriver dates from some time in the 1300s. That discovery prompts Rybczynski to meditate on the nature of invention, and the effects of the machine tool. For the screw enabled the creation of the device that made it possible to mass-produce screws: the machine tool. There were lathes earlier than the screw-cutting lathe, but the quality of the work depended on the skills of their operators: they were basically devices for holding and turning the workpiece, but the operator had to control the cutting tool.
     Modern lathes simplify control of the cutting tool, but still require operator skill. The machine tool does it all. Mount the workpiece, apply power, and the tool shapes the pieces as desired. Above all, the machine tool allows us to make identical parts to almost any degree of precision. Without that, steam engines and other modern machines would be impossible to make.
     A book worth reading and rereading. Rybczynski writes with grace and economy. The book is short, but contains a lot of information and insight. ****

Jim Galloway, jazz musician

Jim Galloway On TVO Take a look and a listen. Also on Youtube.