62
edits
Talk:Mecca: Difference between revisions
Simple statistical analysis added 27 October 2023
Lightyears (talk | contribs) No edit summary |
(Simple statistical analysis added 27 October 2023) |
||
| Line 29: | Line 29: | ||
Of the sample discussed above, the 4 parallel mosques are in north Africa, distant from both Petra and Mecca. So it is understandable that the builders should have adopted some locally approved convention. Nor is a between direction is incompatible with the Petra hypothesis, since in a period of political uncertainty, the builders might have been hedging their bets. So there are no clear counterexamples to Gibson's hypothesis, and what appears to be statistically significant evidence in support of it. | Of the sample discussed above, the 4 parallel mosques are in north Africa, distant from both Petra and Mecca. So it is understandable that the builders should have adopted some locally approved convention. Nor is a between direction is incompatible with the Petra hypothesis, since in a period of political uncertainty, the builders might have been hedging their bets. So there are no clear counterexamples to Gibson's hypothesis, and what appears to be statistically significant evidence in support of it. | ||
Basic statistics | |||
To get an idea of the weight of evidence on Gibson’s side, consider a crude model of the probabilities involved. The orientation of a mosque is a straight line perpendicular to its prayer wall. Allowing for some inaccuracy, it can be accepted that a given mosque is directed to a particular location if this line in within 10 degrees plus or minus of a line to that location. Given the convention that a circular angle is 360 degrees, a circle drawn around the mosque will contain 18 segments of 20 degrees each. The chance of a particular location falling within a given segment is therefore 1/18. | |||
Taking Gibson’s list on his qibla tool up to 707, which is the date of the first building Gibson classifies as directed towards Mecca, there are 35 in total. He classifies 4 as parallel, 3 as between, 1 as Jerusalem, and 1 as Mecca. ‘Parallel’ means parallel to the direction between Petra and Mecca, ‘between’ pointing between Petra and Mecca. The remaining 26 he classifies as being directed towards Petra. Umar ibn 'Abd al-'Aziz is outside the 10 degree error range, at 11.1, the Qasr Humeima is doubtful, and perhaps it is best to ignore the Chinese case. This leaves 22 directed towards Petra. | |||
The probability of a combination of outcomes is obtained by multiplying their individual probabilities. So if the probability of a coin coming down heads is 1/2, the probability of two heads in a row is 1/4, three 1/8 and so on. The probability of an unbroken sequence of heads rapidly decreases with the number of tosses. The probability of Petra falling within the 20 degree error arc of 22 mosques is 1/18 multiplied by itself 22 times, which is a practical impossibility. | |||
To put the point in another way, imagine a roulette wheel with 18 numbers. What is the probability of the same number coming up 22 times? Or, if the same number does come up 22 times, what is the probability that the wheel is biased? | |||
It should be stressed that the above is a crude ways of doing statistics, and Gibson rightly commissioned a professional analysis. But it might be enough to show that the mass of data he has collected should not be ignored. | |||
Fernando | Fernando | ||
:I don't think any of this is compelling content and we already have more than enough text now on what is an obscure theory. If it ever gains traction in wider academic circles perhaps we will revisit. [[User:Lightyears|Lightyears]] ([[User talk:Lightyears|talk]]) 22:27, 23 October 2023 (UTC) | :I don't think any of this is compelling content and we already have more than enough text now on what is an obscure theory. If it ever gains traction in wider academic circles perhaps we will revisit. [[User:Lightyears|Lightyears]] ([[User talk:Lightyears|talk]]) 22:27, 23 October 2023 (UTC) | ||
I've added a brief numerical analysis. You seem to be assuming that statistics can safely be ignored. | |||