Sunday 21 February 2010

Its not what you drink, its how you drink



When is drinking good clean fun, and when is it dangerous and destructive? The dividing line is mostly about how you drink rather than how much you drink, according to this recent Malcolm Gladwell article in the New Yorker.

For instance, recent Italian-American immigrants drink a lot. They drink with every meal, they drink with their families, they drink when friends come over, they drink while watching television. Italian-Americans think about drink as if it were food. Alcohol consumption follows the same quotidian rhythms as the consumption of pasta and cheese. So, alcohol-fuelled loutishness or alcoholism are almost unknown, despite the vast amount consumed. Similarly, the Camba of Bolivia drink a lot, within a well-defined social ritual, with no ill-effects.

Contemporary problems with alcohol are more cultural, related to the meaning associated with alcohol, than physiological. Good point.

Unfortunately, Gladwell concedes another myth which goes against the grain of his argument: the belief that alcoholism is genetic. Consider some of his phrases:

- Around the middle of the last century, alcoholism began to be widely considered a disease: it was recognized that some proportion of the population was genetically susceptible to the effects of drinking

- Philomena Sappio (an Italian-American whose alcohol consumption was studied) could have had within her genome a grave susceptibility to alcohol. Because she lived in the protective world of New Haven's Italian community, it would never have become a problem

This excellent paper Bruce Sacerdote of Dartmouth College tries to seperate the impact of nature and nurture on a number of life-outcomes by studying adpoted children. It compares Korean children adopted by American families with their non-adopted siblings. I like this paper because the data is so clean, certainly compared to most natural experiments in the social sciences. These adoptive parents can't choose the children they want. They meet their child for the first time at an airport, unlike, say, in India, where it is not uncommon to adopt from within the clan.

The paper's main finding about drinking is that adopted children behave no differently from biological children. This would not be the case if alcoholism were genetic. Alcoholism does run in families, but this probably has more to do with upbringing than the genome.

Thursday 18 February 2010

Sit down, I think I love you



A Buffalo Springfield classic just shuffled up on my iPod. It goes:

Sit down, I think I love you
Anyway, I’d like to try
I can't stop thinking of you
If you go, I know I'll cry...

...Oh you know what they say about the bird in the hand
And that’s why I ain’t leaving without you...

...So if you want someone to love you
Pretty baby, I’m your guy.

Really?

Experts like Dear Prudence clearly know more about matters of the heart than me. But even I can tell that Dear Prudence would not recommend “I think I love you, anyway I’d like to try” as the ideal declaration of undying love. You had to think about it? And having thought about it, you were so completely convinced that you were willing to “give it a try”? Dude… really...

And what’s the deal with “the bird in the hand”? Is the girl worth trying to love because she is a bird in hand? Or should the girl settle for you because you’re worth two in the bush?

These lyrics are by Stephen Stills, who also wrote "if you can't be with the one you love, love the one you're with". Bit of a theme developing here.

Maybe Stephen Stills knows something that Dear Prudence and I don't. Maybe this ironic, laconic, self-deprecating approach really works. Real-romantik, a la real-politik?

Either way, great song. Click here to hear it on You Tube.

Saturday 13 February 2010

Were we ever #1?



Were we ever #1? This feels like a question worth asking after the whipping at South Africa’s hands in Nagpur. Maybe the ICC ratings don’t actually mean anything.

For several years I have trusted the Rediff ratings more than the ICC ratings. The Rediff ratings suggest that India never were #1. The latest Rediff ratings Google could find, published in December 2009, show India at #2 behind Australia.

The nice thing about the Rediff ratings is that they set more value on wins against better teams, and wins away from home. They were developed back in 2001 by two geeky cricket fans, one of whom was the Director of the Economics Department at Bombay University. The good professor might have felt the need to develop an intelligent ratings scale because the official ICC ratings developed earlier in 2001 were so bad. These ratings were designed by a panel of distinguished cricketers, like Sunil Gavaskar and Ian Chappell, and treated all test wins as equally valuable. This is not a bad attitude for a player, who should play equally hard against any opposition. But from a fan's viewpoint this original ICC scale is asinine. I thought this post was going to be a rant about the stupidity of the ICC ratings.

However, it turns out that over time the ICC have improved their ratings methodology. They have now incorporated the best idea from the Rediff methodology, that wins against stronger teams matter more. With that improvement, the ICC ratings are not meaningless. India topped a meaningful table in 2009.

There still are interesting differences between the Rediff and ICC scales. The ICC scale gives extra weight to test series outcomes, which is nice. It does not weight-up away wins, which is odd. But the biggest difference is that the ICC ratings give double the weight to wins in the last two years, while the Rediff scale treats an entire cycle of home-away tests as one equally important block.

For instance, the Rediff scale gives Australia’s 5-0 whitewash of England in the 2006-07 Ashes as much weight as the 1-2 loss in England in 2009. Rediff’s logic is that these are the two most recent home-away series. In the ICC ratings, the 5-0 hammering in 2006 gets only half the weight as the 1-2 loss in 2009, because the 5-0 hammering happened more than two years ago. Clearly, weighting-up recent matches makes it harder to apply a home-away factor, because very few pairs of teams will have both home and away matches in the most recent two years.

Neither approach is right or wrong, different scales serve different purposes. The ICC ratings will respond more quickly to changes in performance. It will therefore have more predictive power, will generate more rapid rating changes and therefore more news. The Rediff ratings are probably a more fair and comprehensive summing up of a complete block of historical performance. The swapping of ranks indicates that there probably is no real (statistically significant) difference in the performance of the best test teams since Shane Warne and Glenn McGrath retired.

Rediff ratings don’t seem to have been updated and published on schedule. The most current Rediff ratings don’t reflect South Africa’s drawn series against England, or Australia’s annihilation of Pakistan. Unfortunately, this might be for a good reason. As a profit maximizing brand, Rediff might not want to tell the Indian public things they don’t want to hear. Judging by the mean-spirited and jingoistic reader comments that were posted under the last Rediff update, this is a real concern.

Maybe the chest-thumping nationalism of a big chunk of Indian fans is much more worthy of a rant than the ICC’s rating methodology.

Sunday 7 February 2010

Follow your dream, not (2)

Building on the thread from the last post about following one's dreams...

Consider a fairly routine career choice facing a college graduate, say, advertising vs. IT. Neither job quite counts as living-the-dream, like playing cricket for India. But the college grad finds advertising more interesting than IT. The follow-your-dream school of thought would typically recommends taking the advertising job, since that initial interest is a sign that the grad will enjoy advertising, more and therefore be more successful. My hunch is that that initial interest in advertising contains no information about whether the grad will ultimately enjoy her job.

Post Grad, a movie I watched on a flight recently, broadly in the Reality Bites genre, illustrates the point. It is about a smart, spunky college grad who loves literature, and has dreamed her entire life of working in publishing. After many ups and downs, she finally gets her dream job at the top publishing house in town. Her boss is a jerk. The job sucks. She quickly moves on.

The factors that actually predict whether someone enjoys a job are profoundly situational. Things specific to a particular role, in a particular organization, at a particular time. Questions like: is her boss a good people manager? Does she get along with her colleagues? Does she have the right level of independence, the right support, and prospects for advancement? Does she get paid enough to cheerfully suck up the inevitable hiccups? Is the organization as a whole growing, and filling colleagues with a spirit of generousity? Or is it shrinking, and making colleagues mean-spirited?

Asking situational questions like these feels less pure than looking within and asking "Is this the real me?". But they probably matter more.