Saturday, November 20, 2010

How to pack for a year

Blog post at LostGirlsWorld

The list of gear she relied on heavily is quite interesting:
  1. Bag in a Bag
  2. Zinetic Pocket Slippers
  3. Headlamp
  4. Pack towel
  5. Sleep sack/sleeping bag
  6. Ziploc Big Bags
  7. Fleece Pillowcase
  8. iPod
  9. U-Pillow
  10. Compression Sacks
The list of gear she got rid of is also quite instructive:
  1. Synthetic long sleeve tops (start to smell after prolonged use)
  2. Travel shirts
I'm particularly intrigued by Zinetic Pocket Slippers, as they are cheaper (and perhaps slightly more socially acceptable) than Vibram FiveFingers. Unfortunately, Zinetic appears to be a small manufacturer and it's difficult to find any stores with stock.

Friday, November 12, 2010

You can't see what you don't measure...

Slate
The Big Picture
NPR

If the goal of the Dow Jones Index is to give a single-number summary of the US equity market, the articles listed above argue (convincingly, IMO) that the goal is not being met.

By weighting the prices of a small number of stocks, instead of the market capitalizations of a large number of stocks, the Dow fails to capture the true effects of money flowing in and out of the market as a whole.

Given the amount of time and energy spent following the Dow Jones Index, I would argue that there should be a strong effort to make the index as accurate as possible (or restate the goal). Or, as the Slate article suggests, use a different index.

That's why money managers prefer to use broader market-cap weighted indexes to help create their own portfolios and benchmark their own performances. 

Another alternative: RAFI 1000

EDIT: My title was inspired by a quote I'd read before. Here, it is used in a article describing how airport security in the US could be vastly improved by focusing on the people, rather than the threats.

We have a saying in Hebrew that it's much easier to look for a lost key under the light, than to look for the key where you actually lost it, because it's dark over there.

Wednesday, November 10, 2010

The Paradox of Choice

TED talk by Barry Schwartz

Ever since I saw the TED talk, I find more and more examples that support the idea that limitless possibilities actually cause paralysis. For instance, the rise in popularity of recommendation engines, e.g. Netflix - there are millions of available movies, help me to narrow down my choices to the ones I'm likely to enjoy.

Another example from the New York Times

“The customer walks in the door, and often sees a huge selection of stuff in a multibrand store, and can’t figure out what to buy and ends up buying nothing,” said Paco Underhill, founder and chief executive of Envirosell, a Manhattan-based company that advises stores on shoppers’ behavior.

To maximize the profit per square foot, these retailers have to focus on selling lots of a few items. If you're not limited by physical space, how can you better monetize the Long Tail? Algorithmic recommendation engines (Based on your preferences and/or history, you might like...)? Social recommendation engines (Based on your friends, you might like...)?

Wednesday, November 3, 2010

Non-intuitive Optimization

Article at The New York Times

Dr. Todorov has studied how we use our muscles, and here, too, he finds evidence of optimization at play. He points out that our body movements are “nonrepeatable”: we may make the same motion over and over, but we do it slightly differently every time.

“You might say, well, the human body is sloppy,” he said, “but no, we’re better designed than any robot.”

In making a given motion, the brain focuses on the essential elements of the task, and ignores noise and fluctuations en route to success. If you’re trying to turn on a light switch, who cares if the elbow is down or to the side, or your wrist wobbles — so long as your finger reaches the targeted switch?

Dr. Todorov and his coworkers have modeled different motions and determined that the best approach is the wobbly, ever-varying one. If you try to correct every minor fluctuation, he explained, not only do you expend more energy unnecessarily, and not only do you end up fatiguing your muscles more quickly, you also introduce more noise into the system, amplifying the fluctuations until the entire effort is compromised.

“So we reach the counterintuitive conclusion,” he said, “that the optimal way to control movement allows a certain amount of fluctuation and noise” — a certain lack of control.


Blog article about genetic optimization of Starcraft 2 build orders

A build order refers to the exact opening steps you take early in the game that best supports the strategy you are trying to conduct.

One of the reasons build-order optimization is so important is that you can discover openings that “hard-counter” other openings. If I can get an army of N size into your base when you do opening X, you will always lose.

The most interesting part of this build (the 7-roach rush), however, is how counter-intuitive it is. It violates several well-known (and well-adhered-to) heuristics used by Starcraft players when creating builds.



My take-home lesson: Don't limit yourself by requiring things to "make sense". Allow reality to broaden your mental horizons.