About
When I was 46 I started writing essays on life, or the state of the human condition as I once called it. Because I was halfway between old enough to vote (21) and planned retirement (72) it was known as the "Halfway Point" series of essays.

Later when I mentioned the essays in one context or another on USENET, I got requests for copies and eventually for future essays. Thus the mailing list was born, and it moved to the Internet when that became widely available. At that time I moved to writing on a schedule, the 1st, 11th, and 21st of the month.

Now the trend is to "blogs," and read on demand. I am therefore making this available as a blog, and we shall see if people read it here, or by mail, or not at all.

My other writing
OddLinks - informal comments

Subscribe
Subscribe to a syndicated feed of my weblog, brought to you by the wonders of RSS.

Links
These are a few of my favourite links.

Anchor chains, NASA, and the NY Yankees (16:24)

What is the connection between anchor chains, NASA, and the New York Yankees? It is that they all risk a failure of the weakest link.

In an anchor chain that's pretty obvious, the failure of any one link breaks the chain and allows the ship to drift. Problems like that are serial, every part has to work perfectly. A multi-stage rocket is a serial problem, if any one stage fails the whole process fails. By going to a five stage design with the Ares series rockets the chances of failure are increased.

Consider a parallel problem, Google's web servers. Google has many of these, if one isn't available (fails), there a many others to do the job. The Yankee batters are a parallel problem, if Jeeter doesn't hit Damon, Teixeira, or Rodrigez can. As long as most of the batters do their job someone will get on base about seven or eight times. Of course only in the first inning are they sure that they will start that sequence with no outs.

Or consider typical home frame construction. There are multiple vertical studs supporting a horizontal top beam. As weight is applied to a point, it first transfers to the stud below the load, and if those have a small "give" more of the weight is taken by the adjoining studs. The floor is the same way, floor joists run one way, the floor runs the other, and load is spread by small elastic deflection of the floor and joists.

But no consider the Yankee pitchers. On any given night there is some small chance that a pitcher will have a bad night. Not just throw a pitch which isn't good, but one of those nights when walks, hits, and hit batters follow one another. Unfortunately this is a serial problem, too. If any one pitcher gives up runs, or puts players on base, there's no way other pitchers can pitch better and take those players or runs off the scoreboard. And because every pitcher has good and bad nights, the more pitchers you use, the higher the chance that you will find the one having a bad night. So cringe with me as I watched the first game of the 2009 World Series, when Joe Girrardi tried pitcher after pitcher until he found the bad one(s).

Part of this is bringing in a pitcher to pitch to a single batter, but over the course of a season the greater impact is caused by preparing relief pitchers to throw a single inning. If the relief pitchers were to be ready to throw two innings, a pitcher having a good night could throw two innings without having to face the same batter twice (if a pitcher faces the same batter twice in two inings, he isn't having a good night).

Baseball lives on statistics, maybe the wrong ones.