31
waybread wrote: With Method #1 (recommended in my first post one p. 1, above,) only people with known birth times would be admitted to the sample. If their birth time in muinutes were a multiple of 5, only those with birth times at 05, 25, 35, and 55 minutes after the hour would be accepted. Otherwise we could introduce a bigger rounding error of the hospital were rounding to 10- or 15-minute intervals. Obviously if someone has a birth time in minutes that is not a multiple of 5, (like 3:21) we would assume that it was not rounded, and could include them.
We wouldn't need to do this though, remember the goal would be to show something statistically significant. We could take their recorded birth time, whatever that is, and then show that someone can get a time to within X minutes - hell even a half hour would be statistically significant, 15 mins even more so. We wouldn't need to dismiss certain times, just take them as they are and show that you can get within a certain time frame of that.

We are thinking about it from the point of view of astrologers needing an exact time, but really if someone can use a method to reliably and consistently get it within 15 minutes that would be amazing. In fact it would be mindblowing in terms of proving astrology's efficacy.
In statistics (in whch I am not expert, but took a course in it once eons ago) you also don't expect that 100% of your sample will show the predicted result. The reasons could be data errors, or possibly some individuals are special unique cases for some reason. If you interview 25 people and 24 of them produce stellar (pun intended) results, then you are onto something.
Right you'd need a particular p value. I don't know what would be a suitable correlate for a p value in a study such as this one you hypothetically propose, I'm not an expert in statistics either. Of course many studies of astrology leverage the fact that astrologers are not expert in statistics and so use inappropriate p-values to weaken the claims of astrologers. Think of the Carlson study which used a p value normally used for laws of physics rather than psychology, a more appropriate correlate.

32
I appreciate (relative to Polaris/rectification) everyone taking their time to consider and carefully think out the ramifications of any potential tests and of the methodology.

A few points based on the previous posts and in general relative to the problem at hand, in no particular order:

A statistical study of the kind that would be suggested by a statistician is not even possible in this case, as each particular rectification is time/effort intensive. Since many (most, if not nearly all) here do not really have a clear view of what Polaris DOES or how it functions, it would likely be intensely informing to do a (public) sample rectification to demonstrate the methods and so that all can more clearly see any potential problems in ANY future testing that is decided upon...

The average error between rectified time and recorded time has not been statistically tabulated. I would have to ask Isaac if he has formally or informally tried to find an "average error" which might be helpful in our correct formulating of the conditions for some future analysis.

A normal 24-hour search can require as many as 30 dated events or more to be conclusive. It would be much more efficacious to be given, say a 6 hour window in which the birthtime is known to be contained and then for Polaris to deduce that time to say, within +/- 10 mins.

Using this model, a single successful finding of the birthtime (within the 10 mins) would relate to odds of 1 in 18, or just over 5%.

As we repeat this process, consecutively, the odds (against) increase rapidly due to the formulation of probabilities in mutually exclusive events.

Specifically,
Image
By the time we get to just 5 successful demonstrations, we have defeated odds on the same order of magnitude as if we have rolled a (6-sided) die and correctly called the result EIGHT times in a row. ie. quite some accomplishment! ;)

Re: Polaris as Black Box

Although the results from Polaris must be checked with Secondaries and PSSR aspects to determine which is the correct time from Polaris; Polaris itself is pretty much a black box. The astrologer has no method to weight any particular events higher than others, nor to influence the outcome in any way. Polaris "simply" looks at all the events, gives a weighted score based on how strong the astrology is relative to the complete event list for each moment in the search range and then spits out which times performed best. [The more events that are available, the more the tendency of the correct time to rise higher up the list of suggested times.]

I think it might be very constructive to do one demo-rectification based on the data of someone on this forum. (this also would allow others who have contrasting methods to try their hand at it with their normal "go to" methods and see how well they hold up).

Take care,

James

33
Atlantean wrote:
A statistical study of the kind that would be suggested by a statistician is not even possible in this case, as each particular rectification is time/effort intensive.
So what?

The test would be to show if astrologers can get a birth time within a specific time band (+/- 15 minutes) using whatever astrological method they like.

Given (say) twenty case histories along with a date and place of birth, how well can the astrologers do? That is fairly basic statistics.

I think that if rectification works as astrologers seem to take it for granted that it does, then a panel of say five astrologers of recognised competence should be able to agree on the methodology for each case history and come up with an agreed rectified time of birth.

Let us face it, rectification is a pretty routine astrological procedure. Most astrologers do it. Rectification is not some arcane, esoteric procedure written out in Latin on dusty parchment. The test as described above ought to work.

34
Geoffrey wrote: I think that if rectification works as astrologers seem to take it for granted that it does, then a panel of say five astrologers of recognized competence should be able to agree on the methodology for each case history and come up with an agreed rectified time of birth.
i don't take the idea that rectification works in any for granted way. i don't know that you are going to get 5 astrologers to agree on the methodology to use for rectification either. the divergence in house systems is only one example where it is hard to get consensus within the astro community on basic methodology.

Geoffrey wrote: Let us face it, rectification is a pretty routine astrological procedure. Most astrologers do it. Rectification is not some arcane, esoteric procedure written out in Latin on dusty parchment. The test as described above ought to work.
i don't think it is a routine procedure. some may make like it is routine. using a software program would seem to make it more routine only becuase the software has a built in methodology you either agree with or don't.. bottom line is it would be great to have some study done to help substantiate an approach or software program that claims to be able to do rectification successfully. until that happens it is one persons word over another with no verification.. james a's example with the women who found the birth data was a good one, but not enough to convince anyone skeptical like myself. a real study ought to be done and there hasn't been one so far as i know. rectification has always been viewed as one of the most difficult areas for an astrologer to engage in for good reason.. a person can go wrong very easily. i think it is an ideal area for the use of a computer program, but i am not convinced with the work shared by others using polaris. maybe if i owned the program i would view it differently. as i recall a few years ago it was selling for about 400 US.

Re: Evidence that chart rectification works

35
Hi again,

you guys are in such a "battle" that I think you missed my post. This was the original question for this thread:
Geoffrey wrote:There must be a reasonable body of evidence for charts where the birth time was not known and for which the birth time was rectified - and then the true birth time came to light.

Someone must have done a study of this....?
And as I've replied in page 2 of this thread, we've made in 2010 a study of the syzygy method (which is a rectification method declared in Christian Astrology and Tetrabiblos) and this was the conclusion:
Careful analysis showed us that the syzygy method was crafted in such way that was biased to give a majority of very good results, independently of the data veracity.
So, as far as some rectification methods are concerned, there is strong evidence that they do not work.. The post is in page 2, the study can be read in http://www.skyplux.com/init/static/animodar_effect.pdf

Rectification

36
Paul wrote:
However the reality is that no matter how accurate your technique there are always going to be people who don't want to believe it. More subjects, they'll say. 100 isn't enough. Do 1000. 1000 isn't enough, do 10,000 etc.
Very true. Sceptics can't get convinced. You have to wait until they die out someone once said.
Last edited by Southern Cross on Thu Apr 17, 2014 6:17 pm, edited 1 time in total.

37
I agree, jventura-- the issue is to get some independent verification that a given rectification method produces accurate results.

Astrologers have given rectification methods since the 2nd century AD (and maybe sooner.) But how many had an independent means of checking them against a known birth time?

Your study looked very impressive-- it would be even more interesting to learn whether there is more than one suite of methods that could produce highly accurate results!

1. Importance of a "blind" study. The reason why any researcher, be it Atlantean or anybody else, should verify that he didn't have the birth time in advance of crunching the data is to avoid nasty charges that he "fixed" the research in order to come up with a foregone conclusion.

This is also why I think getting 20 or so ordinary people for whom he would be highly unlikely to know their birth times-- vs. celebrities-- would be the better method. Perhaps the researcher could swear an oath that he had never looked at the celebrities' charts, but it might be suspect, nevertheless, by scoffers.

Re: Paul's post. Frankly, anyone who's looked at rectification knows that you could take a person with a known birth time and apply some common rectification techniques-- just based on the other birth data plus life-events-- that wouldn't work, even so. And that bogus claims would quickly be discovered by other astrologers trying to apply the same method.

I would like to see such a study be as squeaky-clean as possible, however, because I think its implications are huge. If you can predict and minimize any methodological "hooks" or "snags" that might catch a scoffer, it is worth putting in the extra effort. If this stuff works, then astrology as a whole has a much stronger case than most of us ever realized.

2. Rounded birth times. We do not need to use horoscopes in the sample where the time appeared to be rounded to more than 5 minutes. A quick peak at the Astro-DataBank (available via www.astro.com ) convinced me there are plenty of Rodden AA charts there with birth time minutes not divisible by 10 or 15, and even where the minute ends in 1,2,3, or 4. The data bank is huge, and not all of the AAs are rounded more than 5 minutes, by any means.

3. Statistics I take the point that the probability of even 5 horoscopes in a row coming up with verifiable answers is tiny. Five won't seem like enough horoscopes to the scoffers. Probably there are some simple statistical tests around that could be applied, but to me, the odds of getting 19 out of 20 connect could never occur by random chance.

The study jventura linked had over 200 people in their sample.

Re: Evidence that chart rectification works

38
jventura wrote:
Careful analysis showed us that the syzygy method was crafted in such way that was biased to give a majority of very good results, independently of the data veracity.
So, as far as some rectification methods are concerned, there is strong evidence that they do not work...

So they give the right birth-times, but the result was not dependent upon the life events used...so they do not work?

In any case, if a method gives a sufficiently accurate result, I would have no complaints. :)
Interested in Hellenistic astrology? Visit my blog.

The appearance changes, but the essence remains.

Re: Evidence that chart rectification works

39
jventura wrote:
So, as far as some rectification methods are concerned, there is strong evidence that they do not work..
Hello jventura

I did indeed miss your post! Sorry about that.

Your very nice piece of work tested one particular method of rectification and concluded that for good statistical reasons, the animodar method was not viable as detailed by Lilly et at.

What I was proposing, however, was to test if competent astrologers could rectify a birth chart, using whatever methods they chose. This would then be a direct test of the ability of astrologers to do rectification rather than a test of any particular method.

I had assumed that rectification was widespread and commonly done by professional astrologers, which would have indicated that there were methods which were commonly in use (of which the animodar method and it cousins would not appear to be strongly represented) and presumed to be valid. A test of astrologers' ability to do rectification, as described above, would have been interesting in those circumstances.

But I suspect that james_m is correct and that I have overstated the extent to which rectification is done by professional astrologers - or even non-professional astrologers. This kicks rectification into the basket of esoteric and specialised techniques and the fact that a bunch of "competent" astrologers were not able to perform would not be earth shattering news. If there was a statistically meaningful positive result however....

40
Hello Geoffrey,

Re: "So what?"

So what? The point is, each one takes some time and concentration. ie. it is work... Since the average rectification takes a couple of hours, any sizable statistical group would take quite some time. My point relates to how much "free" time I have to work on such a project.

Re: "The test would be to show if astrologers can get a birth time within a specific time band (+/- 15 minutes) using whatever astrological method they like."

There are likely a few methods (PSSR among them) in order to narrow down the search range, but I don't know of any other methods that can reliably find the birth minute and in many cases to just a couple of seconds or so. [This is based on the accuracy and tight maturation of aspects in Topocentric Primary Directions...]

Re: "Given (say) twenty case histories along with a date and place of birth, how well can the astrologers do? That is fairly basic statistics."

Generally, I would expect the answer to be "dismally".

Re: "...then a panel of say five astrologers of recognised competence should be able to agree on the methodology for each case history and come up with an agreed rectified time of birth."

That sounds like a dream. From the (astrological) forums that I have been on, there has been NO general agreement on the correctness of rectifications, regardless of proof or chosen perspective. You can't get 5 astrologers to generally agree on House Systems, let alone a particular, specific rectification.

Re: "Let us face it, rectification is a pretty routine astrological procedure."

This is patently false. Most use Transits to TRY and do it and since, Transits mature as far out as 1? 20' (or more), it doesn't allow THAT tight of fixing of the birthtime. (since 1? 20' on the MC equals over 5 minutes of time, good luck!)

Re: "Most astrologers do it."

Millions of people play guitar. Less than 0.1% do it well.

Re: "Rectification is not some arcane, esoteric procedure..."

CORRECT. It is a very new process, based on a combination of discoveries from several dedicated astrologers working independently. The K?hr refinements to Primary Directions were necessary, as was the discovery (not invention!!!) of Topocentric Houses. The work of Alexander Marr and Isaac Starkman help bring it all together into a usable system.

As I said, if I hadn't seen the software repetitively FIND an unknown birthtime, which was later confirmed by documentation, I wouldn't be so sure that this algorithm works.

For astrologer Ken Haining, it was only necessary to plug in 3 events in order for it to find the correct sequencing of Primary Directions, verifying his birthtime. Now, his is a rare exception, however that doesn't color the results...as the software still was able to find the right minute (1 out of 1440 in which his birth occurred). Think, please, of the odds AGAINST this happening. Think of it analogously. Someone put 1439 black balls and *1* white one in a non-see-through bag. Polaris reached in, in one swoop, and grabbed the single white one. To you, it might not in any way seem special, but to anyone that understands probability, it does.

Anecdotal findings don't prove that it works in the universal case, HOWEVER, the probability does show that Polaris' finding the right birth minute in this case can not be coincidence.

Peace

James

41
Hello james_m,

Re: " i don't know that you are going to get 5 astrologers to agree..."

If history has shown us anything, 5 astrologers aren't going to agree on anything, except in the most general sense. [As I said, I used a Saturn-MC aspect as corroboration for the event "Death of Father" and one prominent astrologer refused to see how Saturn and the MC/IC axis COULD relate to death of Father. At this point, not much of anything would surprise me.]

Re: "I don't think it is a routine procedure."

Correct. I have yet to see another procedure which can lead me (using strictly the event list) to a particular birth-time in various well-timed births. If it can't work in the specific case, it can't be a universal model working in the general case.

Re: "...James A's example with the women who found the birth data was a good one,..."

It was responsible for another level of added confidence in these methods. [And it was Isaac, who instructed me to always go a bit beyond the given search range, in case of time zone issues or times at the very border of the search range...]

Re: "...but not enough to convince anyone skeptical like myself..."

Some believe we've never landed on the Moon. What's a guy to do? ;)

Re: "...a real study ought to be done..."

We (mostly) all agree with that idea. The devil is in the details...

Re: "...rectification has always been viewed as one of the most difficult areas for an astrologer to engage in for good reason..."

Correct. It requires all the normal astrological "understandings" as well as exposure to several non-mainstream techniques and good math skills.

Re: "...a person can go wrong very easily."

True. Not only a person, software can suck at it too! ;) Look at Jigsaw...it doesn't take into account the different TYPES of events. To Jigsaw, your Father dying in a fiery auto-crash is equated to your winning the lottery. Your marriage and your divorce are INDISTINGUISHABLE (to Jigsaw). With Polaris, there are 50 categories of events and EACH has its own astrological expectations (in Primary Directions). To Polaris, the birth of a male child and the birth of a female child don't even look the same. To Jigsaw, your child's birth and their death are astrologically synonymous.

Re: "I think it is an ideal area for the use of a computer program..."

I agree, but perhaps for different reasons. (?) The computer is NOT subject to astrologer bias. An astrologer (generally) will take an event, form some idea about what astrology SHOULD be there and then fold, spindle, and mutilate ;) the birthtime to make that condtion happen on schedule. Polaris "simply" looks at all moments in the search range and tells you which times perform best (Topocentric Primary Directions) taking every single event into consideration. From all of the rectification methods that I have surveyed (approaching two dozen), it appears to be the most objective...

Re: "...maybe if i owned the program i would view it differently."

Yes, the first time you plugged someone's events in (for whom you have a well-timed birth) and saw the birthtime in the first or second slot from Polaris, I could only expect that you'd be impressed and relatively confident of the methods. As I said before, I am firmly of the opinion that this is the best way to prove astrology scientifically to those who are skeptical. If a computer software can take the events of a life and derive a birthtime, it could ONLY happen if astrology works (and in predictable, reliable ways)...

Re: "...as i recall a few years ago it was selling for about 400 US."

To the best of my knowledge, it has always been $350 U.S. I had honed my birthtime (over years) down to 6:17 am (Birth Certificate: 6:18 am) in terms of minutes, but very close to 6:16:45 am in terms of seconds. Polaris indicates that the correct time is 6:16:44 am. When using this time, all of my events are shown to have correct astrology in many systems (Topocentric Primary Directions, Secondary Progressions, Progressed Sidereal Solar Returns, Age Harmonics, and Transits) and the correct House Cusps activated to just minutes of arc in TPD's, Secondary Progressions, and PSSR aspects), for the events in my event list.

To state that more clearly, once I found the birthtime to just seconds, suddenly all these other systems lined up like a lock tumbler, so that all agreed on the events across the entire list of aspects. Since these are mutually exclusive systems, the probabilites of them all agreeing is arrived at by MULTIPLYING the individual probabilities together. The chances of one system agreeing across all events is a relatively unprobable scenario, however other (incorrect) times CAN also somehow appear to "work". As we are forcing it to work for these other systems as well, we are dispelling coincidence by the probabilites involved.

In other words, ONE SYSTEM working can still be a fluke, since there are also wrong times that are likely to also give halfway decent results in that one system. If we are forcing it to work across 5 systems, then THAT SAME BIRTH TIME has to "magically" function effectively in EACH system for EACH event. That is quite some condition!!! This is why I say that these methods are the closest that we have to a scientifically "proven" rectification.

Take care,

James

42
Hello jventura,

Re: "So, as far as some rectification methods are concerned, there is strong evidence that they do not work..."

I agree. If they can't work in the specific case with a well-timed birth and hearty list of dated events, then it is impossible to have confidence in that method in the general case.

James