Temple Stampede in Jodhpur

From the Times of India, Sept 30 2008:

Jodhpur: One hundred and forty-seven dead, over 60 injured and hardly a drop of blood. All it took was some unruly people jumping the queue and some furiously whispered rumours of a bomb going off, to leave this deathly trail.

It was not even dawn. Over 10,000 had turned up at the Chamunda Devi temple in Jodhpur's Mehrangarh Fort for a darshan of the goddess. The temple gates had opened earlier than usual on Tuesday morning, at 3.45 am. Things went smoothly for about two hours. Then, suddenly, some people began pushing forward. A barricade broke, opening up more space on the 8-ft-widepath. Seeing the widened approach, people began jumping the queue, and suddenly everything started going wrong.

A few people fell down and were crushed under the feet of rushing devotees. The news, as it trickled down, got twisted. Someone mentioned a bomb, and all hell broke loose. Panic-stricken people started running in the opposite direction, pushing, trampling. A toll that may have been contained to single digits kept shooting up with every push, every desperate shove.

While the death toll from hospital sources and other eyewitnesses totalled 147, authorities put it at 140. Many suffocated to death while standing in the queue. "Around 80% of the people died on the spot. We just We just have three orthopaedic cases of the 45 persons admitted at the hospital here," said Nitin Negi, a doctor on duty at the city's Mahatma Gandhi Hospital.

The GCP event begins shortly before the stampede, at 5:30 local time and runs for 5 hours to include the developing news and concern over the tragedy. The chisquare is 18292.6 on 18000 df, for a p-value of 0.062 and Z-score of 1.537. The graph shows a ragged history of strong positive and negative trends with an overall positive slope matching the standard prediction.

Temple Stampede
in Jodhpur

It is important to keep in mind that we have only a tiny statistical effect, so that it is always hard to distinguish signal from noise. This means that every "success" might be largely driven by chance, and every "null" might include a real signal overwhelmed by noise. In the long run, a real effect can be identified only by patiently accumulating replications of similar analyses.


GCP Home