What does a cloud feel like to the touch? –B.P. Jones, Chicago

Like a funeral director’s handshake, B.P.–cold and clammy. But why are you asking me? Chances are you’ve already felt a cloud. It’s called fog, which is nothing more than a cloud at ground level.

A LOTTO FLAK

The gambling gods are not happy, Cecil–you’ve blown it again in your column about systems for picking lottery numbers [November 15]. First, card counting in blackjack has nothing to do with “predicting the next card” or any such nonsense, but rather detecting when a particular subset of the deck or shoe is a favorable financial proposition in the long run. I’ll forgive you this amateur mistake since the sheer pressure of having to write a 300-word column once a week must be tremendous.

Your lottery comment shows fundamental ignorance of a parimutuel gambling system at work, however. Perhaps it would be easier to demonstrate your error using the second most disadvantageous form of gambling in the U.S., the horse races. Imagine for a moment a 10-horse race, where horse number 1 for some reason always wins. After some months of this the betting public will likely catch on and everyone present will bet on #1. Will the track care? NO! The track calculates payouts by subtracting its take (usually 17-20 percent) and then dispersing the remainder to the winners. In the case the crowd “won” -\$17.00 for every \$100 it bet. Next time you need a quick and dirty analysis of a gambling problem, talk to a pro. –Greg Bart, Los Angeles

Try not to be a cowpie, Greg. A few points: (1) It’s “disbursing,” not “dispersing.” (2) “Predicting the next card,” as any reasonable person would have inferred, was a shorthand way of saying, “predicting the likelihood that the next card will work to your advantage.” In the typical card-counting system you count the 10-point cards (10, jack, queen, king) and bet more heavily when an unusually large number of them remain in a dwindling deck, since they mean trouble for the dealer. (3) In my discussion of lotteries I wrote, “lottery officials . . . know if their drawings do show a bias, the betting public will eventually discover it and start betting heavily on the hot numbers–and soon the lottery will go broke.” The problem here was that I failed to explain a step in my reasoning, which confused things. But the underlying argument was sound.

To clarify, let’s suppose we have a lottery in which each ticket costs one dollar, the number 1 always wins, and everybody knows it. One hundred people buy tickets and all pick 1. The amount in the lottery “pot,” therefore, is \$100. The lottery administrators subtract their cut–let’s say it’s 20 percent–and distribute the remainder, \$80, to the 100 winners. Each “winner” gets 80 cents back for each dollar forked over–not much of a bargain. So nobody bets anymore and the lottery goes broke. Now you get it?

DUH

In your column on height/long names as factors in presidential election victories [January 17], you state: “The only time the longer-named candidate lost was in 1908 . . .” I admire your wide range of research tools, but sometimes there is a danger in sticking too closely to the book. What about Bush-Dukakis in 1988? Pity the theory didn’t hold then. If you print a correction I won’t even make you eat your hat or anything as penance. –Gloria Allaire, Madison, Wisconsin

I’ve got a little sister like you, Gloria. She is lucky to be alive today. I wrote that “of the 22 presidential elections between 1876 and 1960,” the only time the longer-named candidate lost was in 1908. I went on to say that in the seven elections since 1960 the longer-named candidate has triumphed only once. Next time read the column before you complain about it.

Art accompanying story in printed newspaper (not available in this archive): illustration/Slug Signorino.