When playtesting the place
Some Test - 10 Nov.rbxl (73.4 KB)
, you’ll notice that the Conjurer turns to you and casts their beam. You can’t resist the force and your character is stuck, fixed in place.
Regardless of the PhysicsSteppingMethod being Fixed, Default or Adaptive, the outcome is always the same - you can’t move away / resist the BodyPosition’s force.
Originally, you were able to ‘fight back’ and resist this force. You’d attempt to move away and break free, for which the force was carefully adjusted.
When I made my post in September last year, it could be seen how BodyPosition’s force had dropped significantly., resulting in the platforms no longer holding up and ‘sinking’ a whole lot.
This was until the mover came in contact with a Humanoid type, after which it returned to its set power.*
*unless extreme force was assigned to the mover, you’d always had a subtle dip (first video 0:51), where it’s slightly below the desired spot
This very drop (not the subtle dip!) was likely because the PhysicsSteppingMethod’s Default state got changed from the Fixed state to Adaptive. After RadioGamer told me about this, I changed every place’s SteppingMethod to Fixed, essentially solving the issue 
The platforms were back to normal and the subtle dip was still there, allowing me to make inventive use of it and creating a force for this Conjurer enemy that the player could ‘fight’ and resist by moving.
Then November 9th, as the Adaptive method was updated and no longer featured this drop, the platforms all would indeed no longer sink. This surprisingly wasn’t only for the Adaptive (or Default) PhsyicsSteppedMethod, but also impacted the Fixed state, meaning there was no longer a big drop or loss in power anymore (this was already solved when changing to Fixed) , but the subtle dip also disappeared!
So… this change in Adaptive apparently also impacted the other SteppingMethod, wiping out this dip entirely. This resulted in the Conjurer’s beam no longer allowing for any resistance, breaking the mechanic. 
“Why not lower the force or power”, you may ask. Well, when in this place’s example the force is changed from powerGrabEffect.MaxForce = Vector3.new(100000, 0, 100000) (100,000 - Completely locked in place) to say 10,000 , you find that there is indeed a slight force present and the player starts to feel slowed down in their movement - which is a good thing.
This force isn’t enough however. I noticed 20,000 again completely locks my character from moving, so I opted for 19,000. This still felt off. As I described in one of my messages under the original post:
“Something to the movement or force put towards it seemed off however. It felt as if the force wasn’t constant and that with every movement direction I took, I was able to manipulate the force to a point where sudden movements affected the force too much.”
The force had become reactive to my movement. When moving in a straight line, the force exerted is the same, but when wiggling around and changing direction, you may notice that there’s moments at which there’s far less force present. This reactive, or maybe it’s rather ‘Adaptive’ force is present regardless of whichever PhysicsSteppedMethod you’re using.
This dynamic behaviour is further shown in the part video I linked November 10th in my message. On minute 1:22, the force of 1100 on the Y-axis is no longer sufficient, whereas on 1:13 it lifted the part back up.
In the end it may come down to bringing back different SteppingMethods, as they currently all see the same behavious. Both Adaptive (/Default) and Fixed.
- A few months before my original post there was a subtle, natural dip for a regular force
- At the time of my post creation the power was significantly less (for Adaptive/Default), which I by default my places were set on.
- Around October 17th, when changing the SteppingMethod to ‘Fixed’, the lack of power and big drop gone, but using
Fixed still kept the subtle dip (this because I never put the force too high, as otherwise the platforms would be too fast)
- Around November 9th the physics were changed for Adaptive, but somehow for Fixed as well, making them behave the same - without the loss in power, but without the subtle dip allowing for nifty uses like these as well
I’m not sure if the ‘Fixed’ method was inadvertently updated too, when updating Adaptive’s behaviour around November 8th/9th 
Giving both PhysicsSteppingMethods back their own respective behaviour and uses would solve the issue, allowing for dynamic and reactive physics uses in Adaptive, and keeping consistent force when opting for Fixed.
, but getting forces surrounding this staff to work again would resolve the current issue and probably bring this thought to rest as well.