Three things I don't mean with super-risky:
- Tal was a risky player but he could calculate extremely well and had reasons to believe that if a refutation of his sacrifices exist, the opponent wouldn't find it (only today's engines).
- In a lost position anything is fair game, so no point of thinking about the objective centipawn value of a move.
- In rapid games you can risk far more anyway.
I thus define a super-risky move as a trap in about equal position in a tournament game that can be seen through by the average opponent (say, GM level) and ends up to the disadvantage of the trapper. (That's surely not 100% objective but shall suffice.)
Have there ever be any GMs (especially including GM-strength players before the title existed) playing consistently super-risky? (I throw at least one suggestion into the room: Dawid Janowski.)