Evolution in the Times

Today's New York Times science section is devoted exclusively to evolution. Cool stuff. 

(Hat tip: MW)

Differentiating the search market

James Fallows writes about a study showing that the first-page results of leading search engines produce widely varying results:

That is, rather than Google, Yahoo, Microsoft Live, Alta Vista, Ask, etc providing overlapping views of the central data repository that is the World Wide Web, each returns a particular sampling of that data, which can differ to a startling degree from the other samples.

This seems to make sense. If Google opens a grocery store in Durango, I could profitably open a store in Alamosa by replicating Google's Durango store because geography obviously matters to customers of bricks-and-matter retailers, like Alamosans who'd prefer not to trek to Durango. A strict me-too strategy on the internet, however, is mostly senseless. Wouldn't Ask, Yahoo and other competitors fail in the search competition if they didn't significantly differentiate their product from Google's?

"bias bias"

Here's a link to an amusing and useful page cited on an Edward Tufte forum: a list of cognitive biases divided into four categories: decision-making and behavioral biases, biases in probability and belief, social biases, and memory errors.
Tufte writes:

There's also the "bias bias," where lists of cognitive biases are used as rhetorical weapons to attack any analysis, regardless of the quality of the analysis. The previous sentence then could be countered by describing it as an example of the "bias bias bias," and so on in an boring infinite regress of tu quoque disputation, or "slashdot."

The way out is to demand evidence for a claim of bias, and not just to rely on an assertion of bias.

Good stuff.
One bias we've seen a lot in recent years but which didn't make the list is truthiness bias. The bias I've probably heard most in everyday conversation--racial bias--isn't on the list. I imagine cognitive scientists might group racial bias into in group bias, and/or outgroup homogeneity bias--not that the wikipedia list is necessarily comprehensive or representative of the discipline's practices.

Free wireless at airports? It depends

Water, food, shelter, wireless... In the last year I've passed through a dozen airports, and my search for free wireless at many of them has been fruitless. This possibly incomplete list of airports worldwide shows that most of the busiest airports aren't on the list. Of the fifteen busiest airports--measured by passenger traffic--travelers only get free access in Hong Kong, Atlanta, and Las Vegas, assuming this list is accurate and up to date. To be sure, many regional airports in the U.S. offer free wireless, but I think it's because they're less likely to serve travelers with large expense accounts. O'Hare, LAX, JFK, Denver, and San Francisco, by contrast, all offer wireless for a fee, and appear to reflect the pricing practices of hotels. Many mid-priced hotels offer free wireless, but if you patronize pricier properties, watch out. (Last summer I stayed at a hotel in New York that charged $7 for fifteen minutes of access in the business center!) I suspect airport executives, like those of pricey hotels, know that a certain percentage of their customers need--or think they need--services like wireless, and are willing to pay steeply for it. I don't know if administrators of busy airports actually see wireless as a "profit center," as high-end hotels do, but the pattern suggests such thinking.

Also: Salon recently published an article on "good" airports, and "bad" airports around the world. The approach isn't particularly scientific, just reader responses, but worth a look if airports and air travel catch your fancy.

mpg of bikes

In my previous post, commenter Jacob added that a human can travel more efficiently by bike than by foot, which shouldn't surprise anyone who's ridden a bike.
He ran the following calculation in Google--(31,548 Calories / gallon / 563 Calories / hr * 12 mi / hr) in mi / gallon. That is, a 155 lb person cycling at 12 mph will get 672 mpg. At 10 mph that person will get roughly 747 mpg.
I'm pretty sure this can't be as good as it gets. Anyone know of any vehicles that are more efficient than bikes? I imagine some engineering students have worked on this problem and outperformed ye olde bicycle.

400 mpg, or so

I was in San Diego two weeks ago when I noticed a Prius license plate that read: 54 mpg. That's pretty efficient, especially by contemporary standards, so it's no wonder that the car's owner was proud.
I was walking at the time, however. And surely, I thought, I'm more efficient than a big, heavy non-evolved car, even a Prius, but by how much?
How many miles per gallon of gas does a human get?
One answer is zero--humans obviously can't convert the potential energy contained in gasoline into kinetic energy. That's a boring answer, though.
How many miles per energy contained in a gallon of gas does a human get?
A gallon of gas contains roughly 132 megajoules of energy, 31,548,757 calories, and 31,548 kilocalories, or food calories, or calories, in common parlance (one food/kilocalorie, remember, actually contains 1,000 calories).
A 155 lb human walking at 3 mph will burn 246 kcal/hour, or 82 kcal/mile. Feed that human one gallon of gas in potential energy--31,548 kcal--and he'll have enough energy to walk for 128 hours. At 3 mph, he'll cover 384 miles; in other words, he'll get 384 mpg at that weight and speed.* That's seven times more efficient than the Prius, and 24 times more efficient than a Hummer H2.
It is worth noting that the Prius, say, can carry five people without seriously compromising its efficiency, and it can go well over 60 mph, and carries a big gas tank, consumes zero gas while out of use, and so on (people drive cars for a reason after all!).
The friend who noted this also wondered how many miles per Big Mac a car gets. (Big Macs, he reasoned, are a "generic unit of consumption." Indeed.)
As it happens, caloric as a Big Mac is, a car still couldn't get too far on such a diet. McDonald's says a Big Mac contains 540 calories, which translates to about 57 Big Macs of energy in a gallon of gas. By this measure, a Prius could get almost a mile per Big Mac, while a H2 could only get a quarter mile, and bit further on a quarter pounder.

* A friend ran the numbers from another source that assumed slightly more efficient walking and got 406 mpg, so 400 is a safe approximation, it seems, though we both concede that we might be way off or missing something key in making these calculations. If any readers try to answer the same question, add a comment to tell us what number(s) you got. 

Also, these numbers vary widely according to how much a person weighs and walking speed. Running, for instance, is relatively inefficient, which I suppose explains why people jog to burn calories.

Baude leaving the blogosphere, for now

Will Baude of Crescat Sententia is closing shop to go clerk for a judge in San Francisco. No word on whether any of his co-bloggers will continue posting on the site.

Tiger's consistency

As long as I'm revisiting sports posts: I wrote about how we should appreciate Tiger's consistency more so than his margins of victories because his competition has improved, and in nine tournaments this year he's won three, and finished second in both majors. Not bad.

Cowen's offer

Tyler Cowen is testing a novel marketing (?) conceit: buy his forthcoming book in order to receive access to his secret blog. "[J]ust send an email to IBoughtTylersBook@gmail.com and tell me you bought the book," he writes. "I'll send you the site address as soon as I can. But please, in receiving the site address you are making a pledge not to give it away, publish it, blog it, or otherwise reveal it." 

I can't imagine someone hasn't yet let the cat out of the blog, but I haven't looked. According to one commenter, the book ranked 1,797 on Amazon before the post, and was around 240 by the end of the day.

Revisiting Jack Cust: His first quarter season with the A's

Last month, I wrote about A's designated hitter Jack Cust's long road to the majors, shortly after his debut with the team, so I thought I'd revisit how he's doing. I admit to a little selection bias here--Cust had an oustanding weekend at the plate, giving a nice boost to many of his statistics. Nonetheless, it makes some sense to reconsider his season now that he's played almost 40 games and received roughly 150 plate appearances. In short, Cust continues to prove himself as a premier major league batter. In a variety of statistical categories, he ranks among the league's best. A quick ESPN search reveals that his 1.007 OPS places him 11th in the majors, and his OBP (.423) 9th in the league. His RC/27 is 12th in the league, and he ranks 4th in isolated power. His forte is drawing walks, which he currently does every five plate appearances, a more robust rate than all but three players--superstars Bonds, Thome, and Burrell. (I should add that I sorted all these stats by players with a minimum of 125 ABs because Cust was called up in April).

A little less selection bias would show that Cust's performance has, unsurprisingly, fluctuated a fair amount over the course of the season. When I first wrote about him, his OPS hovered around 1.230, which was unsustainable. His OPS continued to decline through last week, bottoming out at the still-excellent .883 before his most recent streak of hot batting. Indeed, his decline from the initial Ruthian output dropped his fantasy league ownership rate to 3.6% by June 8, and 1.8% by June 17, according to his ESPN player profile.

It's hard to say exactly how Cust will have performed come October, of course, but we'll probably get more of what he's shown us this far this year, and throughout his lengthy AAA career: lots of walks and strikeouts, a respectable average, and lots of home runs, and few steals. 

In his brief stint with the A's, he's made a name for himself as a good story and great hitter, but the Hardball Times writers have also cited his case as a lesson in player evaluation. Jacob Jackson has written an article titled, In search of the next Jack Cust (Part 1), in which he sets out to find potential major leaguers like Cust. That is, productive but inexpensive players who've toiled in the majors into their late twenties, and qualify as six-year minor league free agents, who "therefore aren’t under control of their original drafting team." The A's are paying Cust, for instance, just under $400,000 this year. Check their site on Wednesday when he's supposed to post the first full article evaluating these players.

Seen and heard

* "Prevention is the quintessential non-event because you cannot easily prove why an infection did not occur."

-Sharon LaFraniere, New York Times South Africa correspondent, on NY Times World View podcast, commenting on why financial donors often find treatment programs more attractive than prevention efforts.

* How to make sense of online user product ratings: Farhad Manjoo of Salon reports on efforts to help would-be consumers navigate the problem of "response bias" inherent in product reviews, such as Amazon's five star system. Manjoo writes that, "Purchasing bias thus suggests some helpful advice when you're looking at online ratings: The more expensive a product, Pavlou says, the more you should discount its high reviews."
Does that mean we should discount Porshche's place at the top of the JD Power initial quality rankings? Would the owner of a new Porsche be more or less likely to report problems after spending so much on the sweet new ride?

I remember reading somewhere recently that people are more likely to read advertising for a product after they've purchased it because advertising, obviously, highlights the product's positive qualities. The article/book called it "confirmation bias," which, according to Wikipedia, seems to be what Manjoo is actually talking about--"response bias," by contrast, occurs when "respondents answer questions in the way they think the questioner wants them to answer," but Amazon customers write rosie reviews not to appease survey questioners, but to appease themselves!

Assorted links and stuff

* Yglesias links to a video of Bush's watching getting stolen in Albania. The deed is obscured, but it definitely looks like one of those hands pulled it off.

* New York magazine profiles Edward Tufte. Their site doesn't offer a single-page option, so we're left using their print page instead. Why not just offer both? Ads and all that, I know, but at what cost? (hat tip: aldaily.com)

* Here's a photo of Tim Hawkinson's highly amusing Uberorgan, which I saw last week at the Getty:
Click on the link to see a video about the piece.

SF's One Rincon

Years in the making, San Francisco's new condo tower, One Rincon, is well under way and sticks out in a neighborhood of mid-rises near the Bay Bridge. It's only the first of four planned towers for Rincon Hill, so it will look more inconspicuous within a few years. When my brother told me about the building's progress, I expected it to look like the Citigroup building in Queens, but given the skyline's future growth it does stand out, but also appears to be anchoring the city's southern boundary of high-rises. I imagine it dominates from the street level, but from most distant angles it appears consistent with the city's scale. John King of the SF Chronicle wrote last August, "All this can be traumatic for those of us who carry settled pictures of the city in our heads... [But] Rincon Hill is just a new piece of the puzzle, shaped to the changing nature of urban life..."

Big airport traffic numbers

I was looking at world airport traffic data and came away, uh, impressed. Atlanta's Hartsfield-Jackson airport processed more passengers than any airport in the world last year, roughly 84 million, which is like the entire population of Germany passing through annually. To Atlanta's discredit, unfortunately, most of this traffic is of the hub-and-spoke variety, while Los Angeles claims the most originating or terminating passengers. All of this is, of course, just a pissing contest. For a different kind of airport pissing contest, here's a board of people discussing the most beautiful airports in the world.

Initial impressions of the new Safari 3 beta:

It's good. I like the search/find feature--it's quick and useful; the tabs look sleek and work smoothly; and the drag-tab-to-open-in-new-window feature at least looks cool and feels intuitive. They make claims about the faster-than-the-competition loading, but I haven't been able to tell yet. 

Pepez of On the Way posts his initial impressions here, and Jannis Hermann's crashed as soon as he launched it, but he says he figured out the problem and the solution. 

Download the beta here. 

On Display in San Diego: Adams, Callahan, and Lavine

If you're in San Diego, I'd recommend paying a visit to the Museum of Photographic Arts in Balboa Park. They currently have three exhibits--for 20th C. Americans Ansel Adams, Harry Callahan, and Arthur Lavine--and they're all great. The museum is small, but the selections were impressive, and included such works as Lavine's "Working Hands," seen bastardized by jpeg to the right here.

Tiger Woods at the majors through 2011

Pete projects Tiger's chances at the majors through 2011 (the last year for which we now know the PGA, US Open, or British Open courses):

[L]et's say he has 59 more majors before he gets too old to win (46 years old, the age Jack was when he won his last). Can he win 7 out of 59? That’s 11.86% of the majors for the next 15 years. That’s less than one every two years. For some reason that, at the same time, seems totally do-able and incredibly hard... 2009 and 2010 give him the best chance for the Grand Slam. 2010 especially. And look at all the major courses where he's won or has significant experience; that's scary for everyone else."

The real meat of the post is in Pete's commentary on Tiger's chances at each individual course, for instance:

2007 US Open, Oakmont (originally I thought this course wouldn't be good for Tiger, but they've taken out a lot of trees so it'll be a long course with heavy rough, which should favor strength, much like Bethpage Black)
2007 British Open, Carnoustie (played well there in 1999 and had a chance to win)

Be sure to click over and read the whole thing if you're a golf/Tiger fan.
Revisiting Carnoustie should be interesting, as the media will repeatedly remind us of Jean Van de Velde's dramatic triple-bogey collapse on 18, leaving him tied with Justin Leonard and Paul Lawrie, who ultimately won in the playoff.

The curious incident of Pickleshane and Gapingvoid

On April 30, Hugh Macleod, cartoonist, marketer, and proprietor of the blog Gapingvoid, questioned if his unique style of cartooning was being plagiarized by the author of a blog called Pickleshane. Some of Macleod's readers thought he should sue the bastard, so to speak, but Macleod only questioned the author's failure to attribute, and ultimately appeared unbothered about the resemblances, writing, "I'm just guessing some kid came across my work randomly one day, thought it looked like fun, and decided to have a go himself. I did the same with my favorite cartoonists, when I was young. Whatever."
Shortly after Macleod's post, a response to another blogger's initial reaction to the similiarities, several people commented on the apparent plagiarism directly on Pickleshane's blog on his profile/about page. A few days later in early May, Pickleshane updated his blog with a post saying he'd be traveling for the month with limited access to the internet and his site. May passed, and Pickleshane presumably completed his travels, so I decided to check back in and see if he'd responded. He hadn't. Sometime in the last couple weeks, the blog was deleted. All I found at the URL was this white page with the following Wordpress note: "The authors have deleted this blog. The content is no longer available."
If Macleod wasn't exercised about the incident in the first place, I don't know why he would be now, but it is a curious, if not unlikely, end to an odd incident.
Pickleshane continues to maintain a flickr page, where you can still find some of the cartoons that strongly resemble the style and content that Macleod created.
The author of Pickleshane didn't respond to requests for comment regarding the accusations of plagiarism and the deletion of the blog.

A small tree in Bryce

And here's a photo of a small tree I saw in Bryce that seems to be thriving despite the seemingly hostile landscape (or, as an ecologist would probably explain, because of the hostile landscape):

Yucca blogging

Still mostly away from my computer. In the meantime, here's a photo of a yucca I saw in the Grand Canyon: