|
发表于 2011-12-27 06:42 PM
|
显示全部楼层
The result of the statistical analysis becomes more reliable when you have more samples. You now have 1473+1288 = 2761 days in total. Among these samples, you only get three cases with N=8. Therefore, the conclusion based on this historical data for N=8 and N=9 may not be that useful. It appears useful for N=1 or N=5 because you have got many samples for those small N cases.
The conclusion based on those small N cases appears to be: “next day is almost always with 50-50 chance, under whatever “N” days green.” I expect the same conclusion for those large N cases as well. In other words, when you increases the size of samples for those large N cases, we will see the same trend: 50-50 chance. Currently, for N=6, we have 33-67 chance. To get to a 50-50 chance you need to increase days_up for N=6.
Therefore, we can also arrive at an opposite conclusion: in future, if N = 6, next day has much higher chance (67% = 14/21) to go up to produce a 50-50 chance 20 years later.
|
评分
-
1
查看全部评分
-
|