That logic assumes that the edge exists, and that the set of "what is not an edge" is finite and known, and that "what is an edge" is a knowable nonempty set.
Let's give an analogy. Imagine a class trying to define a parabola and a dramabola (dramabola does not exist). So the class says, to find a parabola and a dramabola, let us look at what is not a dramabola and what is not a parabola. So they go over the list: triangles, circles and etc. Then they try the combinations. What do you think will happen at the end? Bottom line is this: one needs to prove the existence of an edge, and, if it exists, work to find sufficient (not necessary) conditions of an edge.
Based on all of the live trading calls from traders on both this site and trade2win, over a span of about 5 years, I'd say practically nobody has an edge. There's about 3 traders on this site who make money, all who mainly only post in the pnl threads. I don't think they look at charts at all, since there is no 'edge' there. If there was, we'd all be making money.
What about market selection as a skill, experience,advantage, edge. Two traders trading the same signals and mm rules are allowed to make their own selections of stocks and assets to trade. One trader does well, the other fails. This happens time afte time, year after year. The only difference is market selecion as the system is automated/mechanical. Hypothetical of course but... Is market selection an edge?
Actually the thread is not about what is an edge. That's the other thread. This is about what isn't an edge. The existence or non-existence of edges would be a third thread, making 224.
I am starting to think that a valid edge must have significant and outperforming profit expectancy (positive expectancy, of course). Small positive expectancy (with alpha return close to 0) should not be classified as having an edge, that means claiming simply "positive expectancy is edge" could be basically wrong.