Saturday, February 25, 2012

Faster than light, ha!

Here is the latest on the Opera experiment, the one which claimed to have found neutrinos which moved faster than light, and reached Gran Sasso 60 nanoseconds faster than photons would have. `Sources' claim the discrepancy comes from a bad connection between a fibre optics connection with a GPS receiver which was used for the time of flight measurements. There's also a glitch in the oscillator which times the intervals between which the system is synchronised. Although the two glitches are claimed to work in opposite directions in estimating the apparent speed, it is clear that further work and independent measurements are needed to finally close this issue.

Tailpiece: The neutrino and the photon came face to face.
Photon: So, that Bolt was a false start.
Neutrino: It wasn't my fault, it was the cable's!

This blog post is by Neelima Gupte and Sumathi Rao.

Sunday, February 19, 2012

Cricket, anyone?

The cricket season is on with its ups and down, test matches and ODI-s and inflation and deflation of rankings and reputations. Diehard fans, as well as occasional TV watchers wait for Sachin's 100th century, weep for the whitewash in Australia and wonder how things could go so wrong! Let's see what light a bit of quantitative analysis can shed on these events. The ICC rankings, opaque as they are, attempt to quantify the nonquantifiable, the crack of the bat on the ball. A recent paper attempts to do the same, using the techniques of network science, treating the cricket playing nations as a social network. The teams are ranked, depending on their success in test cricket, and one day cricket, for all the years for which the data is available, (1877 onwards, no less), and so are the team captains. The success of a team (or captain) is determined by the ‘quality’ of wins and not on the number of wins alone.

The method consists of forming a weighted network. All competing teams form the nodes of the network. If team A defeats team B, a link is established pointing from B to a A with a thickness (the weight), proportional to the fraction of wins where B wins against A. The importance of the match is assigned via a quantity called the PageRank which uses the normalised weight of the link in a diffusive term which redistributes the credit of a given node to all its neighbours, with maximum credit being transferred to its most successful nemesis. Teams are then ranked by their PageRank, and captains are as successful as their teams.

The results conform to common intuition, which the ICC rankings don't always do. Needless to say, Australia emerges as the strongest test playing nation, followed by South Africa, despite their twenty one year absence from test cricket (1970-1991), England, West Indies, Pakistan, India, Sri Lanka, New Zealand, Zimbabwe and Bangladesh. Since this is an average from 1877, it is unsurprising that this is more or less the order in which the teams started playing test cricket, except for South Africa. For ODI-s, the order is the same, except for Sri Lanka and New Zealand exchanging places. The most successful test captain is Australia's Steve Waugh, followed by South Africa's Graeme Smith, and Australia's Ricky Ponting. From the subcontinent, only M. S. Dhoni and Sourav Ganguly make it to the top 20 list. For ODI-s, Ricky Ponting leads the table, followed by Graeme Smith and Imran Khan of Pakistan. M.S. Dhoni, Kapil Dev, Saurav Ganguly, Mohammed Azharuddin, Rahul Dravid, Javed Miandad, and Wasim Akram also make it to the top 20 list, perhaps indicating that the subcontinent is better at the shorter version of the game. PageRanks computed over successive decades pick up the domainance of the successful teams of those decades, such as the West Indies dominance of both tests and ODI-s from 1981-1990, India's success against strong teams between 1971-1980 (not reflected in the IIC rankings), and the rise and fall of teams like England, Pakistan and Australia in the pecking order. The author (Satyam Mukherjee at Northwestern University, formerly at IIT Madras), modestly says that this scheme cannot replace the ICC ranking, but suggests a novel approach to refine the existing ranking scheme. We hope this scheme will find its way to the cricket commentaries.

This blog post is by Neelima Gupte and Sumathi Rao.

The original paper can be found here .

Thursday, February 2, 2012

Now where were those keys?

This week sees the news of an important breakthrough in the study of degenerative brain diseases such as Alzheimer's and Parkinson's diseases.While it has been known that the diseases spread due to the spread of a distorted protein, (the tau protein for Alzheimer's ), outward from an area where memories are made and stored, the mechanism of spread was not clear. There were two possible mechanisms proposed for the spread. One, that the spread could take place from neuron to neuron, along neuronal pathways, and the other that there were neighbourhoods that were susceptible to the bad protein, and others that could resist it. It is now established that the spread takes place along neuronal pathways.

The experiments that establish this mechanism are ingenious, and involve genetically engineered mice that can create the human tau protein in a localised area called the entorhinal cortex. Cells in the entorhinal cortex of the mice started dying due to the tau protein. In due course, the disease spread to other areas via the neuronal network. Since other cells could not make the tau protein themselves, the only way the tau could show up in other areas was via transmission from nerve cell to nerve cell. It may then be possible to halt the diseases by preventing cell to cell transmission, e.g. by blocking the tau with an antibody. This might provide the key to the prevention of degenerative nerve diseases (and help find those elusive keys!).

This blog post is by Neelima Gupte and Sumathi Rao.