[ALittleLight]

Whenever I have to mentally rotate shapes (only on tests of mental rotation / IQ) I do it by just comparing the features of the original shape to each of the candidates and ruling out options. e.g. noticing that one option has too many vertices or not enough. It's tedious but pretty reliable.

Your comment made me Google "aphantasia mental rotation" and the summary of the first result I got makes me think I'm probably right. As I read the summary, aphantasia people were slower but more accurate. In other words, the aphantasians know they don't have visual imagery and therefore do the slow but steady thing whereas the visualizers think they have visual imagery and are fast but wrong.

Granted, I haven't read the paper, but at first glance I'm counting it as evidence for me.

https://jov.arvojournals.org/article.aspx?articleid=2700106

[vidarh]

Or we've spent a whole lifetime training spatial recall, and so we can do that part better. I can spatially "manipulate" scenes I remember well with great precision, and I don't visualise anything. As such I don't think you can reasonably say anything at all about the ability to visualise based on this.

[thewakalix]

But why would it take visualizers longer to just pretend to rotate a shape over a greater angle? I think the best explanation is that something is really being rotated, however inaccurate, simplistic, or vague.