Question: How do the online yijing readings actually work?

[eowyndernhelme]

Hello, All

I may be in the wrong place, or maybe this question has been answered somewhere on site (if so, please direct me there).

I use several methods interchangeably to ask questions; Russell Cottrell's virtual sticks, LiSe's site, Hilary's 3-coin on the Resonance Journal, and the manual 2-coin method that is supposed to approximate the yarrow stick reading. I use the one that kinda matches my time constraints or the importance of the cast, for instance, I use the 2-coin as a hands-on when it's more important or for a daily reading, the others for quick readings, or just using intuition for which would be better at the time.

What I'd like to understand is how the online computer methods work. Is it random number generator based? And how are questions / intentions able to affect it?

Thanks for reading --Pam

 

[hilary]

For full details you would need to ask the programmers, but basically, yes, random generation of each line in turn, with the probabilities of each kind of line copied from a traditional method.

And how are questions / intentions able to affect it?

Exactly the same way they affect coins or yarrow stalks. (Or in other words - no-one knows.)

 

[eowyndernhelme]

"...no-one knows." Ha, was kinda thinking I might get that answer Guess I'll just carry on with my exploration. Thank you!

 

[remod]

As Hilary said, you should ask a programmer, of which I happen to be one.

However, there is not a single answer. Each online site uses a different piece of code to cast hexagrams.

Let’s rule out the bad sites where they fake the casting process and just select an answer without even trying to relate it to one of the know casting process (i.e. using yarrow stalks or three coins). They should be simply avoided.

The "good ones" (luckily the vast majority of the ones I came across) mimic the casting process as closely as they can.

Here on Clarity, the casting script generates a random numbers between 0 and 15 every time you click on the "Cast" button and produce the results:

• 0: moving yin (6). Probability: 1/16
• 1 to 5: yang (7). Probability: 5/16
• 6 to 12: yin (8). Probability: 7/16
• 13 to 15: moving yang (9). Probability: 3/16

which closely resemble the results you would get using yarrow stalks.

The "yarrowfull" script on Lise's site goes even further and mimics the entire process of subdividing the yarrow stalks bundle more times to get the outcome, exactly following the same steps as the manual process.

Other "good sites" uses the same principles with the three-coins probability distribution.

The second key topic is how good the Random Number Generator is. When you toss the coins, divide the yarrow stalks, or shuffle the sixteen marbles, you use a TRUE random number generator: a process that is guaranteed to provide you an outcome that you have no conceivable way to know in advance.

A computer most of the times will use a PSEUDO Random Number Generator, a "formula" that produce outcomes that are not distinguishable from real random numbers but that could potentially be knonw in advance (by knowing the initial random generator state).

For the purpose of generating hexagrams there is no practical difference as long as the Pseudo Random Number Generator used by the script is half-decent as a poor one could favour certain results or even, in extreme cases, rule out certain outcomes. Luckily, the standard generators of most programming languages are good enough.

As for the question influencing the outcome, I've seen no example of such software. Yes, it may be done (for example using it to seed the RNG) but it should be threated carefully to avoid the outcome becomes predicatable.

It may be interesting, however, to add some real randomness by using a physical process like the white noise in a radio signal or the nuclear decay of a radioactive material but it would be much more complicated.

In the script I use on my site, I generate the lines according the yarrow stalks probabilities and the randomness directly depends on the user actions as they have to move (or click) the mouse on an image to cast a line. The script avoid using the javascript random numbers generator alltogether and is done so that you can't produce predictable outcomes (e.g. by clicking in the same spot all the time):



The full code is released as Open Source on Github, for those interested.

BTW, once you get the response, you can select Hilary's or Lise's translation to get the readings (by clicking on the hexagram numbers).

I can be much more technical if you want, just let me know.

Hope it helps.

 

[eowyndernhelme]

This does help, very much so. You are answering many of the questions I had that I didn't even get to ask yet

Also, per the 16 marbles, there was a site that used that method some time ago that I liked to use because of it supposed approximation to the yarrow sticks. But it was java based and either the site owner took it down or a browser "upgrade" did (maybe both).

It's good to know that the yijing sites that I currently use right now also aim for the yarrow method.

Will check out your github and.. wait, I think I know who you are now, you're kind of famous.

Gods, I am so embarrassed right now. Did you do the chatGPT AI I Ching?

I am really honoured to meet you.

 

[eowyndernhelme]

I don't have plus, just the standard.

 

[remod]

Gosh! I didn't know I was famous

Yes I did the GPT4 experiment too, you'll find the prompt on GH as well but I've done some modification online, let me know if your are interested and I'll update the prompt on GH as well. The most current version is Ji Yi, it uses yarrow stalks (sixteen marbles) probability distribution but follows its own process as it casts first the entire hexagram and then, for each line, determines if it's a changing line according to the yarrow stalks probabilities. You can see it done in the response.py script. Alas OpenAI only allows access to those with a premium account.

I'm happy to see someone is interested! Feel free to ask, sharing will be a pleasure for me.

 

[hilary]

Also, per the 16 marbles, there was a site that used that method some time ago that I liked to use because of it supposed approximation to the yarrow sticks. But it was java based and either the site owner took it down or a browser "upgrade" did (maybe both).

That was here! Unfortunately, I used the Wilhelm/Baynes translation, and someone reported the copyright infringement. I received a very scary email, promised them I hadn't made any money from it, and took the script down in a hurry!

 

[eowyndernhelme]

Really, here? Fascinating, as Spock would say. So that's what happened, eh? Wow, what a huge tiny world we live in. I know many who decry it (one of my sons included; how is he even related to me ) but imho internet is one of the best things to happen in my lifetime.

 

[Liselle]

Also, per the 16 marbles, there was a site that used that method some time ago that I liked to use because of it supposed approximation to the yarrow sticks. But it was java based and either the site owner took it down or a browser "upgrade" did (maybe both).

It's good to know that the yijing sites that I currently use right now also aim for the yarrow method.

Just in case you don't know, those odds are heavily in favor of hexagram 2 as the relating hexagram. You'll get 2-relating a lot more often than anything else.

I can't explain it pithily and don't know if it's good, bad, or indifferent. I used the marble method for a while years ago and the readings seemed just as comprehensible as coin odds (or, you know, not - I don't understand them a fair amount however I cast them ).

 

[eowyndernhelme]

Hello Liselle,

Are you saying the odds for the marble method, or the yarrow, or both?

Yikes, I use the yarrow online and the 2-coin manual method (which I read also mimics the yarrow) for most all my readings and yes, the stats for a few months back show relating hexagram 2 much more than any other besides 23.

I just thought Yi was telling me to stop being so bossy and to listen more (and 23 to strip away the deadwood).

 

[eowyndernhelme]

Gosh! I didn't know I was famous

Yes I did the GPT4 experiment too, you'll find the prompt on GH as well but I've done some modification online, let me know if your are interested and I'll update the prompt on GH as well. The most current version is Ji Yi, it uses yarrow stalks (sixteen marbles) probability distribution but follows its own process as it casts first the entire hexagram and then, for each line, determines if it's a changing line according to the yarrow stalks probabilities. You can see it done in the response.py script. Alas OpenAI only allows access to those with a premium account.

I'm happy to see someone is interested! Feel free to ask, sharing will be a pleasure for me.

I appreciate this and will be asking many questions, please stand by

 

[Trojina]

Never heard of a 2 coin method.

 

[Liselle]

Are you saying the odds for the marble method, or the yarrow, or both?

I think they're the same odds, but that's off the top of my head.

 

[Liselle]

And also in this article of Hilary's!
https://www.onlineclarity.co.uk/learn/ways-to-consult-the-i-ching/16-beads/

 

[hilary]

Never heard of a 2 coin method.

2 coins

In addition to reproducing the yarrow probabilities, this method has all the advantages of the three coin method: practical, simple, and no special equipment required. Of course, you can still opt to light a candle or burn incense, or use special Chinese coins - whatever helps your focus on what yo

www.onlineclarity.co.uk

 

[hilary]

I think they're the same odds, but that's off the top of my head.

Or approximately the same... the maths is beyond me for yarrow. @remod ?

 

[remod]

Or approximately the same... the maths is beyond me for yarrow. @remod ?

Hilary knows I can't resist

Yes, the 2-coins method is fully equivalent (odds-wise) to the 16-marbles one.

The first time you throw the two coins you have 1/4 chances of getting 2 and 3/4 chances of getting 3, the second time you will get 1/4 chances of getting 4, 1/4 chances of getting 6 and 2/4 chances of getting 5.

If you put them all together you will get:



which are the same odds as the 16-marbles one.

 

[eowyndernhelme]

Yes, and I use 2-coin because I lost my marbles but still have 2 pennies to rub together.
[applause]
Thanks, folks, we'll be here all week.
Seriously, will check out your odds table. The ghost of Tom Lehrer (my statistics teacher at UCSC) is watching.

 

[Trojina]

2 coins

In addition to reproducing the yarrow probabilities, this method has all the advantages of the three coin method: practical, simple, and no special equipment required. Of course, you can still opt to light a candle or burn incense, or use special Chinese coins - whatever helps your focus on what yo

www.onlineclarity.co.uk

Why ? For when you don't have 3 coins available ? Anyway never heard of it till now.

 

[eowyndernhelme]

Why ? For when you don't have 3 coins available ? Anyway never heard of it till now.

Well, I just maybe thought in my own mind that yarrow was the way to go because that's the way it was done in the beginning? But I only read this. Maybe it really started with kiln-dried tortoise shells. And make turtle soup?

 

[eowyndernhelme]

You are probably correct; in the end it's just probability. I have a pair of dowsing rods that actually worked to find us a well, but now I live in an apartment that I really love

 

[Liselle]

I think I read somewhere that yarrow stalks aren't actually the oldest? Could be mistaken.

(Oh - I don't mean not the oldest non-tortoise-shell method, either. Tortoise shells were older but I don't think that was casting a Yijing reading the way we think of it. They read cracks in the shells, but I don't remember ever seeing an explanation of exactly how it worked. Unless the cracks always lined up in stacks of 6 broken -- and solid __, it's not a Yijing reading I don't think.)

 

[remod]

I think I read somewhere that yarrow stalks aren't actually the oldest? Could be mistaken.

I never put together all the sources but from what I've gathered, yarrow stalks-based divination was in use for a long time before using coins, sometimes as an alternative method to confirm/enlight oracles got with scapulomancy (using tortoise shells or ox bones). The exact method used, however, has been lost to us.

Then the three-coins method became common and only much later the yarrow stalks method we know has been "reconstructed" based on previous commentaries.

So, it's true that yarrow stalks were in use before coins but it is also true that the yarrow stalks method we know is more recent than the three coins.

Sooner or later I will muster the energies and will required to diligently read and report all the sources. For now, this is what I can remember on top of my mind .

 

[hilary]

So, it's true that yarrow stalks were in use before coins but it is also true that the yarrow stalks method we know is more recent than the three coins.

Spot on, at least as far as I know. Either that predominance of changing yang (and hence Hexagram 2 as relating) reveals a deep understanding of the asymmetric nature of yin and yang, or else it shows a blithe disregard for statistics. Take your pick.

 

[Liselle]

Either that predominance of changing yang (and hence Hexagram 2 as relating) reveals a deep understanding of the asymmetric nature of yin and yang

Yes, an explanation like that that makes sense could convince me to use it, but otherwise...well, I don't know. Explanations either way would be good, but maybe they don't exist. Meanwhile I'll just keep my head in the sand; it's reasonably happy there.

 

[surnevs]

I just don't understand how the algorithm / AI program etc. for the I Ching casting can know that you are the Questioner and not anybody else who pushes the button?

 

[hilary]

About the same way the coins 'know' that you are the questioner. Either way, you're setting in motion a random process to generate a hexagram.

 

[surnevs]

Well, yes. Seen from the perspective that it went on from turtle shell- to plant- and to coin divination... why not?

Add.: maybe except that humans did those ceremonies contra algorithms based on randomized algorithms - again based on theories concerning randomness.

The theory will go that when you throw a die six times, one of those six throws will result in one eye. Try it out...

1. six throws: Oh, 1 eye occurred
2. six throws again: Oh, 1 eye occurred three times
3. six throws again: Oh - Hmmm. No occurrence of 1 eye - and so on.

This is the formula given, roughly, by the French counterpart to the Italian Leonardo da Vinci, Blaise Pascal. If you ask me, Pascal made it as a finger pointer to what in many cases could be expected, but suppose the case where there was no occurrence of 1 eye through those six throws, and in an allegorical way compared with what you achieved at this very moment where you urgently needed advice?

 

[remod]

Well, yes. Seen from the perspective that it went on from turtle shell- to plant- and to coin divination... why not?

Add.: maybe except that humans did those ceremonies contra algorithms based on randomized algorithms - again based on theories concerning randomness.

The theory will go that when you throw a die six times, one of those six throws will result in one eye. Try it out...

1. six throws: Oh, 1 eye occurred
2. six throws again: Oh, 1 eye occurred three times
3. six throws again: Oh - Hmmm. No occurrence of 1 eye - and so on.

This is the formula given, roughly, by the French counterpart to the Italian Leonardo da Vinci, Blaise Pascal. If you ask me, Pascal made it as a finger pointer to what in many cases could be expected, but suppose the case where there was no occurrence of 1 eye through those six throws, and in an allegorical way compared with what you achieved at this very moment where you urgently needed advice?

I'm not clear what you mean. The probability of not getting any 1 in six throws is 33.49% (which is a little bit higher than 1/3).
For an event to have probability 1/6 doesn't mean that if you repeat the experiment 6 times you will get that even at least once.

Probably I missed your point.