Most advice about collecting is really advice about speculating, and the two want opposite things.
The investment frame
Buy what will appreciate. Prefer graded, sealed, canonical. Avoid anything idiosyncratic. Track the market. Sell at the top.
This works, sometimes, for a small number of people who treat it as a job. For everybody else it produces a cupboard of things chosen because somebody else might want them later — which is a strange reason to own an object.
The other frame
Buy what you want to look at.
It sounds soft and it has a hard edge: it is the only frame where you cannot really lose. If the thing never appreciates, you still have the thing. The downside of a bad investment is loss; the downside of a bad purchase you love is nothing.
What this changes in practice
You buy the odd one. The piece with the unusual colourway, the regional variant nobody indexes, the item with a history. These are the ones that are interesting and underpriced, precisely because the speculators avoid them.
You buy condition you can afford. A played-with example of something wonderful beats a sealed example of something you do not care about.
You stop at your number, because you are not calculating a return — you are deciding whether you want it.
You learn faster. People who collect what interests them end up knowing a category properly, which is also, incidentally, how you spot value.
The honest caveat
Some things do appreciate, and it is pleasant when something you bought because you liked it turns out to be worth more. That is a bonus and not a plan.
Anyone who tells you which category will appreciate is guessing with confidence.