28-11-2022, 08:36 AM
(27-11-2022, 03:40 PM)Mark Knowles wrote: You are not allowed to view links. Register or Login to view.(27-11-2022, 12:50 AM)ReneZ wrote: You are not allowed to view links. Register or Login to view.Your own model
All I have done is taken the dataset that Koen lists, unvarnished. I have then calculated the mean and variance on the basis of the data. The only sense in which I have imposed "my model" is in assuming the underlying data is normally distributed; that seems to be a pretty fair assumption and consistent with the distribution of the data. So to say that it is "Your own model" seems to give the impression this is one that I have cooked up to prove a point. I would be very interested if anyone could suggest an alternative model based on the dataset.
You have brought your model forward twice, precisely to 'prove' the point that the 95% probability upper limit is at 1444. You did this also in this thread, where you are trying to argue that a date after 1430 is in fact more likely.
Except that the same model says that this probability is 16% (after 1431). This is not logical and you are not being critical, at least of your own views.
Also, in this post, it is the third time that you challenge readers to argue why your model is not good, or to come up with a better one. So you stand behind it. You should accept the 16% after 1431.
Even though that is a bit on the high side.
The model is not doing very well precisely away from the high-probability area. The 16% probability after 1431 is not matched by Koen's data, which has only 3 (out of 35) points above 1431. It is doing even worse at the other end. Only 2 out of 35 points in the lower 16% of the distribution.