People arrive at dealer positioning data wanting a directional edge, and the data does not have one. This is not a limitation of a particular vendor; it is a property of what is being measured.
Why the mechanism is symmetric
Dealer hedging responds to price rather than anticipating it. In a damping regime, an upward move brings selling and a downward move brings buying — the same resistance in both directions. In an amplifying regime, both directions extend.
There is no asymmetry in the mechanism, so there is no directional information in the output. Asking a symmetric model which way price goes is asking it to produce information it does not contain.
What you get instead
The expected size of moves. Whether ranges are likely to hold or break. Where the resistance to movement is strongest. Roughly how far a move is likely to run before it meets something.
These are volatility statements, and they are genuinely useful. They tell you what kind of session to expect, how to size, where to place a stop so it is not sitting in the middle of a level that will be defended.
How to combine it with a view
Positioning is a second input, not the first one. Form a directional view from whatever you normally use. Then use the map to decide how to express it: how much room the move has, where it is likely to stall, whether the environment rewards patience or speed.
A bullish view in an amplifying regime and the same view in a damping regime are different trades with different targets and different stops, even though the view is identical.
The tell for bad analysis
Any commentary that reads a positioning figure as bullish or bearish has made a category error. Read what it says about the size of moves and ignore the direction attached to it.