If you can’t decide between the Marshall, the Najdorf, the Dragon, the Sveshnikov and the Winawer, you should consider the workload of each of these opening. The payoff on the other hand doesn’t matter, because all of openings lead to drawn positions and they are all sharp enough for your opponents to go wrong. What follows is a list of variations that you have to be prepared for. Since those are only the starting points, expect a huge amount of work required in some of these lines.
Judging from the pure workload the Sveshnikov is pretty interesting, and if you consider the amount of forced draws that help to cut down variations, it becomes even more interesting. The Rossolimo spoils it a bit, but so does the Ruy Lopez Exchange for the Marshall.
The problem with the Sveshnikov is that it is a very concrete opening where you have to know exactly what you are doing. Superficial knowledge or concepts like “play on the dark squares” are not sufficient. Here is an example from my own practice:
At the first glance this looks like a pretty decent game and if two GMs had played it, you would probably assume that it was computer preparation leading to a draw. There is a rook sacrifice, a queen sacrifice and an exchange sacrifice, white accepts neither one of them and after all of these fireworks it all fizzles out to nothing. It is all cool until you switch on your engine, then you start to cringe. The truth is that I totally botched the opening, ended up in a lost position and simply managed to swindle my opponent.
One final reminder: If you want to play the Sveshnikov because you like the dynamic positions after 9. Bxf6, forget it. About 80% of your games will feature 9. Nd5 where you will be faced with the task of defending a stable rather passive position. Same thing with the Dragon: If you seek sharp counterplay after 9. Bc4, forget it! About 80% of your games will feature 9. 0-0-0 which usually leads to static structures. In a way it’s like picking your poison. It depends on how comfortable you are with the positions when white doesn’t go all-in.