There's a quiet frustration in the screenshot that says it all: you freeze columns to keep your bearings, and then the spreadsheet punishes you with a double-vision echo every time you scroll left. The user who posted this did everything right, marked the column, hit freeze panes, even tried split to see if that would stop the bounce. And still, the tool lingers on a duplicated view for half a second before snapping back. That's not a workflow problem. That's a design flaw hiding in plain sight.
The core issue here is that the freeze feature is doing its job in the most literal sense, but not in a way that respects how people actually work. You lock columns to anchor your context, not to watch them reappear on the wrong side of the screen while you're mid-scroll. The fact that split mode keeps the duplication permanently, while freeze only makes it temporary, tells us the software knows these two states are different, but it doesn't give the user a way to say, "Stop at the edge and don't show me the ghost." The user shouldn't have to know the internal mechanics of panes versus splits to get a clean view. They just want the scroll to stop where the frozen column ends, full stop.
What this reveals is a deeper truth about how spreadsheet tools treat their most basic features: they assume users will adapt to the quirk, rather than the tool adapting to the user. The person who posted this isn't asking for a radical new feature or a complex macro. They're asking for a simple boundary, a hard stop at the edge of the frozen region so their eyes don't have to process redundant information for even a fraction of a second. That's not a niche request. Anyone who works with wide datasets and frozen headers has felt this exact moment of cognitive friction. It's the kind of thing that makes a tool feel clever in theory but clumsy in practice.
The good news is that this is fixable, and it doesn't require rethinking the entire spreadsheet paradigm. It just requires listening to the people who actually hit these walls. When a user says, "I don't want to see double for half a second," that's not a complaint about performance, it's a request for intentionality. The tool should let the user decide whether the frozen column acts as a hard boundary or a soft reference point. Until that choice exists, features like freeze will always carry a hidden tax: the mental cost of constantly reorienting yourself after every scroll. The user who posted this deserves better than a workaround. They deserve a scroll that respects their focus.