In “Russia is short of soldiers…”, the estimate of 400-650 Russian deaths is provided. It may eventually join the pile of  published wrong guesses. But it is an example of how disparate bits of apparently unrelated information can come together to provide an intelligence estimate. Some of these bits are “general knowledge”, while others would … Continue reading Russian Casualties estimate; a technique note →

If you become a predictor, and make a prediction that is justified by fair use of the tricks of the trade, you are bound to encounter a heckler who wil assail the fallibility of each of your datums. Your datums are things you read, things people said, personality assessments, things that have happened,  things that … Continue reading How to handle hecklers →

Fencing the problem is an important part of the predictor’s toolkit. Sometimes the fence is made of facts; other times, pseudo facts, things that have higher probabilities than the swirling cloud of amorphous possibilities. Since the Sunnis and Kurds have abandoned parliament, one part of the fence is that they are out of the picture. … Continue reading No successor to Maliki named; fencing the problem →