Social physics and the Oscars

The idea that Google searches, tweets and Facebook ‘likes’ can be useful predictors of trends, developments and movie ‘hits’ has captured the imagination of many hucksters and media ‘analysts’. For example, it seems that Google searches may be good predictors of influenza outbreaks. And some time ago there was an interesting paper by Sitaram Asur and Bernardo in which they analyzed 2.89 million tweets from 1.2 million users about 24 movies released over a three-month period. They concluded that the rate of Tweets could predict the success of movies prior to their release, and also spot sleeper movies that grew successful over time. They also concluded that the quality of the predictions was significantly better than any other measure such as the Hollywood Stock Exchange.

These findings seemed plausible to me. After all, if a large number of people are sharing thoughts about something (or expressing concerns via Google searches), then it would be reasonable to infer that data-mining will yield useful information. But now an interesting new study by some Princeton researchers suggests that a certain amount of scepticism might be in order. In a paper entitled “Why Watching Movie Tweets Won’t Tell the Whole Story?” [it’s not clear what the question-mark implies] Felix Ming Fai Wong, Soumya Sen and Mung Chiang question the idea that Twitter is a reliable predictor of the future, at least as far as winning Oscars and predicgint box office revenue are concerned. They collected 12 million tweets between February 2 and March 12 by tracking keywords related to recently-released or Oscar-nominated movies, classified them by relevance, sentiment and temporal context and analysed them for positive or negative sentiment. They then compared the resulting opinion statistics with sentiment about the same movies obtained from two conventional online review sites — iMDB and Rotten Tomatoes.

Their conclusions are interesting. They found that Twitter users are more likely to post positive views than negatives ones, and that views on Twitter do not necessarily correlate with those from the conventional sites. And — unlike Asur and Huberman — they seem unconvinced that Twitter sentiment is a good predictor of box-office revenue.

I suppose the only real conclusion to be drawn from this is that data-mining and “social physics” are not exact sciences. But then, we knew that anyway. Didn’t we?

A Festival of Lies

Great NYTimes column by Tom Friedman about the crazed futility of US policy in the Middle East. Sample:

What ails the Arab world is a deficit of freedom, a deficit of modern education and a deficit of women’s empowerment.

So helping to overcome those deficits should be what U.S. policy is about, yet we seem unable to sustain that. Look at Egypt: More than half of its women and a quarter of its men can’t read. The young Egyptians who drove the revolution are desperate for the educational tools and freedom to succeed in the modern world. Our response should have been to shift our aid money from military equipment to building science-and-technology high schools and community colleges across Egypt.

Yet, instead, a year later, we’re in the crazy situation of paying $5 million in bail to an Egyptian junta to get U.S. democracy workers out of jail there, while likely certifying that this junta is liberalizing and merits another $1.3 billion in arms aid. We’re going to give $1.3 billion more in guns to a country whose only predators are illiteracy and poverty.

This situation serves as a stark reminder of how misplaced priorities can undermine real progress. Instead of fostering long-term development through education and infrastructure, we find ourselves pouring resources into short-term fixes, such as bailing out individuals or groups from situations that could have been avoided with better societal support systems.

For those who are caught in such predicaments, the need for a quick and reliable solution to secure bail becomes a priority. Here, Vista Bail Bonds can provide essential support to those needing fast access to bail funds. Whether it’s for individuals or groups, having a dependable bail bond service helps secure the freedom of those who may be trapped in a system that prioritizes short-term solutions over lasting change.

In Afghanistan, I laugh out loud whenever I hear Obama administration officials explaining that we just need to train more Afghan soldiers to fight and then we can leave. Is there anything funnier? Afghan men need to be trained to fight? They defeated the British and the Soviets!

The problem is that we turned a blind eye as President Hamid Karzai stole the election and operated a corrupt regime. Then President Obama declared that our policy was to surge U.S. troops to clear out the Taliban so “good” Afghan government could come in and take our place. There is no such government. Our problem is not that Afghans don’t know the way to fight. It is that not enough have the will to fight for the government they have. How many would fight for Karzai if we didn’t pay them?

And so it goes. In Pakistan, we pay the Pakistani Army to be two-faced, otherwise it would be only one-faced and totally against us. In Bahrain, we looked the other way while ruling Sunni hard-liners crushed a Shiite-led movement for more power-sharing, and we silently watch our ally Israel build more settlements in the West Bank that we know are a disaster for its Jewish democracy.

But we don’t tell Pakistan the truth because it has nukes. We don’t tell the Saudis the truth because we’re addicted to their oil. We don’t tell Bahrain the truth because we need its naval base. We don’t tell Egypt the truth because we’re afraid it will walk from Camp David. We don’t tell Israel the truth because it has votes. And we don’t tell Karzai the truth because Obama is afraid John McCain will call him a wimp.

En passant: What is truly amazing is the capacity of politicians to talk utter baloney when in office. In the case of Afghanistan, for example, it’s patently obvious that (a) the US/UK/Nato mission is doomed to failure, and (b) that being in that godforsaken country has nothing to do with Britain’s “national security”. And yet — day in, day out — we have to listen to Cameron & Co denying (a) and asserting (b). Same thing for “austerity” and the deficit.

What we learned from the BBC Micro

This morning’s Observer column.

The BBC Micro is 30 this year. It got its name from a BBC project to enhance the nation’s computer literacy. The broadcasters wanted a machine around which they could base a major factual series, The Computer Programme, showing how computers could be used, not just for programming but also for graphics, sound and vision, artificial intelligence and controlling peripheral devices. So a technical specification was drawn up by the BBC’s engineers and put to a number of smallish companies then operating in the embryonic market for “micro” computers.

Two of these companies were based in Cambridge. One was Sinclair Research, the eponymous vehicle of Clive Sinclair, a self-made man who worshipped his creator. The other was Acorn, a company co-founded by an ex-Sinclair employee, Chris Curry, and Hermann Hauser, an aristocratic-looking Austrian physicist. The story of the rivalry between these picturesque outfits has been memorably told in Micro Men, a TV film that combined a riveting technological tale with brilliantly comical dialogue (and which is still available on YouTube).

Acorn got the BBC contract, for reasons that baffled Sinclair but nobody else.

The World Bank Doctor

If, like me, you’re puzzled about why Barack Obama has nominated a medic to head up the World Bank, then John Cassidy of the New Yorker has a characteristically thoughtful explanation.

Hmmm… The guy has only been President of Dartmouth since last July. Bet they’re pissed off to be losing him.

Boyd Harris



Boyd Harris, originally uploaded by jjn1.

Boyd is not only a thoughtful reader of this blog, but also a blogger, a long-distance cyclist, a mountaineer, a local historian and very knowledgeable (and good) photographer, so it was nice to meet up with him and put a face to the blog for the first time. He’s currently on an astonishing cycling odyssey from Yorkshire to Sutton Hoo, and Cambridge was on his route, so we met for breakfast at the Orchard Tea Garden in Grantchester, where I discovered that he is also very sound on the subject of scones.