You are viewing a single comment's thread from:
RE: Are the dates of 'terror' events predictable by analysing numerology used by secret societies?
Numerology is the art of torturing the numbers until they confess what do you want. Fortunately the numbers do not suffer in the process, but my point is that you can prove whatever you want with that.
The chance of 2 terrorist attacks share the same day of the year is number of total terrorist attacks/365 The chance of 6 terrorist attacks at the same day of the month is number of total terrorist attacks/(30*6), the other calculations are meaningless as you are introducing statistic bias.
i am not attempting to prove anything random using numerology, i am only pointing to the patterns that numerologists hold dear and asking if their interest in them is relevant to the patterns that arise. i am not looking at this as if i am hunting for a winning lottery number, rather i am just looking at the patterns to see if anything is really connected or not in the ways that some claim.
as far as i am aware, the calculation for combining the probabilities of multiple instances of an event which have their own independent probability is the one that i used, which involves multiplying the number of possible outcomes for each instance - as you can see in the 4th graphic down on this page: http://www.wikihow.com/Calculate-Probability
I apologize for the misunderstanding about the author.
The probability of calculations is not correct because the author choosed the terrorist attacks whose dates were according to his theory. In probability the sample choosen for calculations must be random, doing otherwise is cheating.
the probability that has been calculated is to determine specifically the chance that any event which could occur at any time will occur on the days that these ones did occur. in other words, if we allow 6 truly random events to occur in a year, the chance that they will occur on dates with the same day number and in the same 3 months out of twelve, is as far as i am aware nearly 3 trillion to one.
it doesn't really make any difference how many other events occurred that did not fit that pattern and, in fact, the fact that there are other events which don't fit that pattern only affirms the low probability of such a pattern existing.
in short, if the calculation is correct in terms of the scope that it is intended to be relevant to, then the result is correct.
Let´s try with an example. In a daily lottery with numbers from 1 to 100 after 1000 days the number 13 won 7 times. With a 1% chance for every day seems reasonable, isn´t?
The probability for the number 13 for winning 7 times in a row is (1/100)* (1/100)* (1/100)* (1/100)* (1/100)* (1/100)* (1/100)= 1×1014. But I lie and tell you this is the probability for the number 13 to win 7 times along the 1000 days, so the lottery is flawed and you should not play. Would you believe me? Because this is the same the author did.
i am not following your logic here.
you stated the probability of a number winning 7 times in a row and made the claim that to lie and say that the resulting (low) probability is actually the probability of the number winning 7 times over 1000 days (instead of 7 days) is the same kind of mis-representation of the numbers as you are identifying in the calculations i, as author, made in the original post.
i think you might have misunderstood the intention of the calculations in the maths in the original post.
to clarify: the maths demonstrate the probability that any event (of any kind) will occur on specific days and in specific relationship to them IF they ARE random. so in other words, if there were no deliberate and conscious process involved in deciding the pattern of the dates then the odds of the particular pattern being as it has been shown to be with these terror events would be 1 in 3 trillion.
the point is that since the terror events DO fit this unlikely pattern, then there must be some motivating factor that is increasing the odds that these events fit this pattern - they cannot simply be random.
you already stated that they are not random and that is the point. what you appear to be missing though is that the nature of their non randomness is in question. you have stated that they are not random because they are all terror events that happen to fit a hypothesis (note that the hypothesis came AFTER the noticing that the number pattern existed and not before), but my point is that the DATES SHOULD be random if they are not following a predetermined pattern. I am only looking at the dates and nothing else, so why are the dates forming such a pattern that has such low probability?
are you claiming that the probability of such attacks falling on such a day is quite high?
if so, what is the probability of these attacks occurring on the dates that they have occurred on - according to your calculations?
My point is that it can´t be proved that the 6 chosen events were picked up at random, so the probabilities should be calculated with a bigger sample with the formulas I gave in the first comment. For example, all the terrorist attacks in Europe in the last 20 years, or all the terrorist attacks worldwide in the last 5 years.
we are approaching the data with different intentions and understandings about how to make the data useful.
Terrorist attacks are thought to be independent of each other in terms of their date within the year - so i am not sure exactly why it is necessary to include in other terrorist attacks to calculate the probability that a specific set of events are timed randomly or not in relation to each other.
i did consider using all of the available dates of terrorist attacks, but then the issue becomes very cloudy - since many of the attacks are just one guy stabbing one other guy.. so where do we draw the line? maybe some are genuinely terrorist in nature and some are not. from our position, we do not know for sure all of the details of every event.
for that and other reasons i chose the more general approach that i chose.
in any case, you wrote that:
"The chance of 2 terrorist attacks share the same day of the year is number of total terrorist attacks/365"
the formula for probability is generally:
probability = events / number of outcomes
so p = 2 (terror events) / 365
p = 0.0054794520547945
p = 0.54%
I think you are convinced there is a significant pattern between terrorist attacks and there is no statistical explanation that can convince you it may be just a chance not so unlikely. So this discussion is over for me.