Georgina,
thanks for your reply. Let's now make the consistency-test.
Let's label the axis of flight of the particles with "y". Let's label the vertical axis with "z" and the remaining axis with "x".
Let us first look at the case where both magnets have a relative angle of 0 degree around the y axis so that they are oriented in space like depicted in figure 6.2 of that paper (although there is just one magnet scribbled)
https://physics.mq.edu.au/~jcresser/Phys301/Chapters/Chapter6.pdf .
Let us now analyse one particle pair. According to your scheme, every pair send out from the source has a shared orientation of their (gyroscopic) axis. So for each of the two particles of that pair its axis points in the same direction in our coordinate system as the partner's axis does.
Let us now assume that our particle pair's both axis' are in alignment with the z axis when sent out from the source (oriented vertically).
Since in your scheme the particles coming form the source have opposite spin direction, the test series will give anti-correlated results. As you wrote, each individual particle is acted upon by its local environment.
Let us now assume that for this test series the magnets hadn't been oriented as we have defined it above (called SCENARIO A), but both had been oriented 90 degrees relative to what we defined above (called SCENARIO B) whereby maintaining their relationship of field orientation. So in scenario B, the magnets are in alignment with the x axis, and hence have a 90 degree angle to the z axis - although we always assume that the particle pair's orientation and spin direction is left unchanged.
Now according to your scheme, that difference between the original angle of 0 degree and the alternative angle of 90 degree for both magnets does not alter your statement that "each individual particle is acted upon by its local environment.". So the local conditions for each particle in this alternative case are
"At 90 degrees each particle will be effected by its local environment, according to such things as where it entered the field and its orientation on doing so. That experience of field is not necessarily but could be by chance matched by the partner"
Consequently, according to your introduction of chance when one locally changes a 0 degree angle situation to a 90 degree angle situation at one magnet, the local outputs at that magnet now should come about by chance. Since in scenario B both sides have been altered by 90 degree, consequently your rule of chance is realized for both sides and that alternative scenario should give 50% anti-correlation and 50% correlation. But that is a contradiction to what you predicted when both magnets have the same orientation - what is exactly the case in my alternative scenario. Don't bother about me ignoring your rule of anti-correlation or your rule that only one magnet is allowed to be turned 90 degree for "activating" your rule of chance, i will soon come to that issue.
I now have to cite myself as i wrote above
"Consequently, according to your introduction of chance when one locally changes a 0 degree angle situation to a 90 degree angle situation"
Take care of what is meant here by me: DON'T CONFUSE the angles in my citation (0 and 90 degree) with the relative angles between the two magnets. The angles in my citation are the angles when ONE magnet's LOCAL output for 0 degree RELATIVE to the source IS COMPARED to its output if that one magnet's angle RELATIVE to the source is changed by 90 degree (NOTE that the orientation of axis and spin direction of that incoming particle stays exactly what we assumed it to be for the 0 degree relative angle to source). It DOESN'T matter how the other magnet is oriented, since we are examining LOCAL behaviour at one magnet - independent of the orientation of the other magnet:
the magnet we examine cannot know how the other magnet is oriented - even if the other magnet is oriented identical - what scenario B covers. But that other magnet could also well be oriented in a variety of angles and that's the reason why one should not confuse the angles in my citation with the relative angles BETWEEN the two magnets. And that is also the reason for why one cannot apply your anti-correlation rule for my alternative scenario.
As a result we have a scenario where the magnets have 0 degree relative angle (SCENARIO A) to each other and your anti-correlation rule should apply. And we have an alternative scenario where we compared this rule for the case when your rule of chance should apply (90 degree). That scenario (SCENARIO B) was the change of orientation of both magnets relative to the source by the same amount (90 degree) and in the same direction, whereby we assumed for both scenarios that the test particle's orientation of axis and spin directions have in no way altered in scenario A compared to scenario B and vice versa. The result is that your two rules are inconsistent with each other since they predict different results for scenario B.
KEEP IN MIND that scenario A and scenario B AREN'T to be understood that these are TWO runs that should factually be conducted one after the other in an experiment. Since it is clear that for such a test sequence we would need TWO particle pairs - that could well have different orientations and spin directions compared to each other. We assume instead that we have ONE particle pair with well defined axis of orientation and spin directions before measurement, no matter whether we then apply scenario A or scenario B to that pair. With that we examine what would happen to that particle pair if we had measured it differently than scenario A would have done. It doesn't matter that we cannot predict the outcome of just one particle pair being tested. The statistics does matter and the statistics that is produced by your rule of chance is different than the one produced by your rule of anti-correlation.
So, if you do not focus on your anti-correlation rule but instead focus on ONE magnet oriented in one scenario (scenario A) at 0 degree relative angle to the scource and in the second scenario (scenario B) the same magnet oriented 90 degree different from scenario A - but for both scenarios with the same particle with the same initial orientation and spin rotations unchanged entering the magnet of scenario B (so as if the first scenario hadn't happened but instead scenario B was applied to that identical particle) - then you hopefully will grasp that your anti-correlation rule and your introduction of chance at 90 degree are inconsistent with each other. They simply predict contradictory results for scenario B, since:
if you only focus on ONE magnet as just described, you may say that your rule of chance should apply for scenario B. But if you also focus on the other magnet in scenario B(that is oriented identical to the first magnet you focused on, namely with a 90 degree change compared to scenario A), then your rule of anti-correlation should apply. So two rules that predict different outcomes for one and the same scenario (scenario B) should apply for scenario B. That's the inconsistency i spoke of.