In general... (I prefer rapid over speed chess) I think I lean slightly toward the strength of prep over theoretical soundness. But I also think most chess players lean way too far in the direction of theoretical soundness, to a point they're all way too predictable.
Engines are just very powerful and are most utilized when you're in prep and the opponent isn't.
Though lately I shifted a little back toward theoretical soundness because... I used to play the Veresov, but at some point in the game prep wears off and then you want to be left with a fundamentally sound position, that wasn't really happening with the Veresov.
I realized then there's two ways of getting the opponent out of prep - sidestepping the prep, and going deep to where you escape their prep. And ideally I think you must use both approaches. If you try to just sidestep the prep - you will probably end up compromising your position too much. But if you make no attempt to sidestep you will just be playing the super-theoretical lines everyone knows and wind up playing 20 moves of theory, and there will be no venom to your repertoire either.
So I try to find an ideal mix. But this is all kind of abstract, I try to look at the whole picture of a given line and think about it in all its different aspects. Things like...
- I like it when the line has an idea behind it that's novel
- I don't want there to be a simple and straightforward way of responding to it (i.e. the Smith Morra - that opening doesn't throw me off at all, it's common and it's the easiest line to prepare against, it's just a bad line)
- is there a strong refutation or just a slightly reduced engine eval but very playable? With leela I try to keep the engine eval above +0.25 for black, and for white... even or just slightly in favor of black, like -0.10ish at most. Beyond that I try not to place too much weight on engine eval, though I do need some reason to ignore the engine... but I like to know that if things got sharp I could buckle down and memorize something and be fine. Though sometimes there are continuations that leela evaluates outside that boundary but they're super-sharp and never will get played so I just ignore that
- how sharp will I need to play the line? Could I take over and play it intuitively once the prep wears off?
- How does the position "feel" to play? Are there any compromises in the position? What are the final winrates in the line, and what are they throughout the line? This is some rudimentary attempt to get the pulse of the position
- What percentage of players play the opening? I do somewhat insist on this number being relatively low, or at least whatever continuation I'm playing
- how much pressure will it actually put on my opponent? Could he just play it intuitively, or probably not? How punishing is it if he screws up?
- I try to find lines that are complex and dynamic, and prefer them over alternatives... but if the position is sharp it either needs to be very sharp without alot of deviations, or it needs to settle on something that can be played intuitively within like 10-12 moves
Hello everyone,
At shorter time controls (3 minutes or less), does a technically unsound but very familiar opening perform better than mainline theory?
The success of the Banned Gambit, successfully deployed by Brandon Jacobson against Danya, seems to have pushed the envelope far more significantly than previously thought. Despite it being an objectively terrible opening, it worked due to thorough preparation in a fast time control.
How much can one sacrifice in objective soundness in the name of shock value before it begins hurting more than helping?
I’m curious to hear your opinions, regardless of your rating. I see a lot of sniper setups even at the 2200+ bullet/blitz level, while mainline theory often doesn’t help in obtaining any meaningful advantage.
Also, note that even in classical time controls, shock value is real, like when Pragg defeated Vidit with the objectively lower-tier Deferred Schliemann in the Candidates.