Flooding in Thailand, Rioting in Greece
Note: this event analysis is done with only 34 eggs. It will be redone later when the rest of the network is again reporting data after recovery from a server change.
While the public outcry against the government austerity measures in Greece gains voice and grows in numbers and stridency, there is a major natural disaster unfolding in Thailand. Here are some notes on the latter from Southgate Amateur Radio News:
The situation in Greece threatens the Eurozone and with it the world economy, so it garners attention. With the floods now endangering and partially engulfing Bankok, the world media are giving Thailand more coverage, though they tend to focus on the politically contentious. My wife proposed that these very different but deeply meaningful events constitute a powerful presence in the global mind's eye. Both are slow-moving disasters which we can only sample. In Thailand, we can hope that the leaders bring sufficient wisdom and material to bear to avert a greater tragedy. In Greece the demonstrations are huge, and increasingly violent. We can only hope they do not turn deadly.
The conversation about these ongoing events was on the 19th, so that day was selected as representative of the persistent turmoil. The result for the 24 hours (UTC) is Chisquare 85185.209 on 86400 df, for p = 0.998 and Z = -2.935. The graph shows a very steady cumulative deviation in the negative direction for most of the day, which, although it is opposite to the GCP standard prediction, is so strong it looks meaningful.
The continuing turmoil in Greece and the slowly unfolding disaster in Thailand illustrate "events" that our technology cannot adequately address. Not only are they persistent over time, they literally overlap so there is no obvious way to attribute correlations to one or the other situation. Only when we can specify a few hours that "should" represent an event, are the resulting statistics interpretable. To visually represent the difficulty, the graph below provides context for the Thailand/Greece event. It shows October 17-19, and the result suggests that While the final statistic indicates no strong effect, the history of deviations over the three days is remarkable. There are two powerful trends (long periods of consistent deviation) one positive and one negative, that cancel to produce a final statistic that is close to zero deviation.
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.