Tuesday, March 27, 2012

game theory programmers wanted

I received the following email, which may be of interest to some readers:

The GAMBIT project — Software Tools for Game Theory — has been accepted for the second time as a mentoring organization for the “Google Summer of Code 2012”. This means that Google pays students a stipend (5000 USD if the work is successful) to work on open-source code for this software project over the summer of 2012.

With this email, we would like to ask you to encourage students to contact us if they are interested in contributing to this project. It would involve the coding of algorithms and user interfaces for game-solving software.

The “mentors” for this project are Ted Turocy, Rahul Savani and Bernhard von Stengel, all with a long-standing interest in games and computation. Any enquiry should be directed to the three of us under the mail alias

gambit-mentors “at” nash.lse.ac.uk

The student application period is March 26 - April 6, with final deadline April 6, 2012, 19h UTC.

The webpage


gives a summary of our project, and lists at the bottom under “Application Template” a number of questions an interested student should answer. He or she should definitely contact us before submitting an application to clarify his or her interest. The possible projects are listed under “Ideas” which is linked to


and this list can be extended, for example if someone is interested in computing equilibrium refinements. We have kept this ideas list fairly non-academic but there is no real limit for more advanced projects related to equilibrium computation or game processing.

The timeline to the Google Summer of Code program is at


and other frequently asked questions are at


We expect something like 3-5 student stipends, and of course are interested in serious contributors with knowledge of game theory and coding experience.

Thank you for identifying possible contributors, and please do not hesitate to ask us any questions. We look forward to hearing from you and from your interested students!

With best wishes,

— Bernhard, Ted and Rahul

Sunday, March 04, 2012

fixing maths and CS education

Here are some links to news articles I’ve been reading recently. On the BBC News web site: Poor numeracy 'blights the economy and ruins lives'... maybe it does, but let’s be careful what we wish for; numeracy is to maths the way ICT (as used in the sense below) is to CS, and the wrong attempt to improve “numeracy” could backfire and create a new generation of schoolkids who hate maths even more than the present lot. The article has a link to a maths test, one that is frankly about as mind-numbing as any other maths test I’ve come across.

It’s in contrast to the situation regarding computer programming, where things looks much happier what with the arrival of Raspberry Pi and recent success on reforming computing at schools. In yesterday’s Guardian: The Raspberry Pi can help schools get with the program provides an update on the newly-released Raspberry Pi:
The Raspberry Pi project – a philanthropic effort to create the contemporary equivalent of the BBC Micro of yesteryear – has graduated from idealistic vapourware dreamed up in Cambridge to a finished, deliverable product manufactured in China.

Here’s a quote from the article
Second, we need to persuade Michael Gove and his colleagues that the subject that should be taught to all children is not ICT but something called computer science. The idea that there's a major body of knowledge in this field – complete with a stable and intellectually rigorous conceptual framework that is independent of today's or yesterday's gadgetry – is probably unfamiliar to residents of Whitehall, who think ICT is trivial because it's always becoming obsolete.
There’s another reason to be careful what you wish for: maybe it’s something you already have. Some links: The Guardian’s digital literacy campaign is a good resource for what’s going on with reform of ICT teaching. See Dept. of Education announcement: ‘Harmful’ ICT curriculum set to be dropped this September to make way for rigorous Computer Science; and a consultation; reported here and here; Royal Society report.

The Raspberry Pi is supposed to help to rekindle the excitement for programming that occurred during the ’80s, which I remember well. Mind you, at the same time, it was even then considered questionable to want to become a ‘programmer’; you were supposed to want to be a systems analyst, or some such job title. And of course, this remain an obstacle to programming in that it’s sometimes perceived to be low-level grunt work. Here’s a blog post that argues against that:
Still, competition for the few programmers out there looking for work is very steep. So few Americans know how to program that firms like Google and Facebook are actually buying whole companies just for their code-literate employees, in what are known as "talent acquisitions."

According to Calacanis, each employee who understands how to code is valued at about $500,000 to $1 million toward the total acquisition price. One million dollars just to get someone who learns code.

Firms' other strategy, of course, is to import Chinese and Indian programmers, through a costly and often only temporary visa. (That's because, unlike those countries, we don't teach programming to students in the United States. At best we teach kids how to use programs that are already on the shelves. But that's another article.)

I don’t know whether to be happy or unhappy to hear that the US has the same problem with programming that we have here. It seems like programmers are valuable due to their scarcity; maybe it’s against their interest to highlight the problem.

(added later:) some more links: Computing at School discussion forum; Information pack on teaching Computer Science at schools, distributed to head teachers, prepared by BCS and CAS