xbt-101

The total is now at 1491 as of September 16. That translates to roughly 6 cases per day. It is fairly unlikely that only 8 or fewer cases are reported in the final 14 days. Using the 1491 as a baseline we would expect another ~80 cases for the remainder of September which would put the total around 1570. We assume the 1500-1649 category is most likely in this case. This means anywhere between 1 case per day and 11 cases per day will stay within this category.  We will use 805 likelihood. 

There has not been a lethal confrontation, and the question resolved in 8 days. It appears very unlikely at this point.  

Ceasefire is still in effect, and there have been no significant events since the April/May timeframe. We will assume a low probability to start (5% as a base). More significantly, there is very little time left between now and 30 September, so we will adjust down to 1%.  

Such a move would account for a ~16% reduction in troops in S Korea. Drop offs of this kind are rare - since 1970 the force levels in Korea have been stable, with a drop off of 16% year to year only occurring a few times. We will nominally put the likelihood at 10%.

If there is a reduction of that size, the likelihood of that announcement coming in the remainder of September is low, we will assume a conditional probability of 10%, for an overall probability of 1%. 

Further, per MSN, Adm. Samuel Paparo, the commander of the US Indo-Pacific Command, reportedly told South Korean government ministers that there would be no reduction in the number of US forces stationed in Korea. 

We will use 1% as an adjusted probability. 

Historically, <= 1 is the most likely value - there have been 4 major hurricanes in the last 4 years. 

No CAT 3+ yet during the time period in question. 

Gabrielle is a current hurricane but as of 9/21 is not on a path to make landfall. 

Looking at Palestine on the ACLED explorer, we see fairly consistently over the past week and month an average about about 40 political violence events per day. This computes to ~1200 events per month. We will assume the majority (~80%) of the events are in Gaza, with some variance. We will assume a normal distribution. This leads to a greatest likelihood of between 751 and 1000, with a larger percentage between 1001 and 1250 than between 501 and 750. 

The data breaches digest shows 533 attacks this month as of 8/31. That makes all options below 533 a probability of zero. There is some probability that there is a delay and the total exceeds 550, so we will allow some probability that this occurs (10%).