华尔街日报近日的好文，可供正在选择专业与就业的参考


New Hiring Formula Values Math Pros
Region's Employers Seek Statistical Experts Over Computer-Science Generalists

Being a math geek has never been cooler, at least in Silicon Valley.
As Bay Area technology companies ramp up hiring out of the recession, they are in hot pursuit of a particular
kind of employee: those with experience in statistics and other data-manipulation techniques.
Rather than looking for just plain-vanilla computer scientists, who typically don't have as deep a study of math
and statistics, companies from Facebook Inc. to online advertising company AdMob Inc. say they need more
workers with stronger backgrounds in statistics and a related field called machine learning, which involves
writing algorithms that get smarter over time by looking for patterns in large data sets.
That is leading to opportunities for quantitative specialists like Andr s de Lucca. The 30-year-old, who graduated
from Northwestern University's Kellogg School of Management with a concentration in analytical consulting in
2009, says he found himself sought after by some major technology companies. He was hired by Social Gaming
Network Inc., a Palo Alto online videogame maker, late last year.
At SGN, Mr. de Lucca now manages a three-person team that studies game-play data, such as how long people
spend in certain parts of a videogame, to detect areas to improve. For example, by contrasting various missions
of SGN's "Skies of Glory" fighter pilot game, Mr. de Lucca and his group found that people were quitting when
the game was too easy rather than too hard.
"Some patterns aren't as obvious as they seem," says Mr. de Lucca. Using the findings, SGN gradually increased
the game's difficulty, getting people to play it longer and spend more on virtual goods inside of the game, he says.
The focus on math aces like Mr. de Lucca shows how Silicon Valley tech companies, which have long been
data-minded, are now becoming obsessively so. Cloud computing, the process of storing data remotely rather
than on a company's own equipment, is allowing firms to keep more data than ever by lowering the costs of
handling it. Companies can then use the data they gather to refine services, especially online, as soon as they
detect a pattern.
Now as local high-tech companies step up hiring, many are targeting data-expert specialists to gain a leg up over
competitors. Michael Morell, managing partner at Riviera Partners in San Francisco, says the recruiting firm
has seen more companies seeking to build data-analysis and data-warehousing teams. "They're putting a
massive emphasis on it," he says.
Riviera is working to find quantitative experts for a gaming company and an online software company that want
to build "internal centers of excellence" around data, Mr. Morell says. He says those with strong statistics
backgrounds will earn up to 20% more than generalist engineers, who typically start with salaries in the low six
figures.

The most desirable candidates, employers say, can have a variety of experience and educational backgrounds.
Companies say specific degrees are less important than a focus on data-mining techniques.
Mike Schroepfer, Facebook's vice president of engineering, says the social-networking company has a growing
need for employees with statistics backgrounds. "When you build for the Web, you have quicker real-time access
about how people are using the product," he says. "You have to use the data in the proper way."
The company is tackling the problem in two ways. In the summer of 2008, it created a six-week-long boot camp
required for all new engineers and included a module on training in probability and statistics. Employees teach
new recruits to use tools for comparing the performance of one version of a feature with another and how to
determine what sort of difference in response is meaningful.
Secondly, Mr. Schroepfer says the company has been hiring more people for its data science group, its SWAT
team of fewer than 20 Ph.D.s with the deep background to analyze sample sets of terabytes of data, looking for
patterns that could help optimize products and ads.
Local universities are trying to bulk up the supply of graduates with deep quantitative backgrounds. Stanford
University, one of the largest feeders of engineers into Bay Area tech companies, began requiring its computer-
science undergraduates to take a specialized course in probability two years ago. Enrollment in a more-advanced
version of the class also has been growing steadily, drawing 260 students in 2009, up from 215 in 2008, says
Mehran Sahami, a Stanford associate professor of computer science.
"Interest mirrors industry," says Mr. Sahami, who was previously a senior research scientist at Google Inc. He
says that among other things, the courses teach students to build applications that can predict whether a patient
might have a heart abnormality based on X-ray data and to understand how email spam filters work.