Guidance on using chess engines effectively

Hi,

I am using the free Lucas Chess using the also free Stockfish to analyze games, so that I can:

• Get an idea about the game level (Lucas Chess provides an elo rating played per player that I find pretty accurate, even for short move times);
• Get to know openings / positions people play and learn from them;
• Let it extract puzzles from my own games and other games to practice with.

I am analyzing games with a (very) good PC (i7-7700 cpu, 64GB RAM DDR4, SSD drive) and I found little difference whether I use 2 seconds per move or 10 seconds per move when analyzing the 2018 world championship matches, let alone lower rated matches. This seems logical, since Stockfish analyses millions of positions every second and it is quite powerful, but I am clearly no expert on the matter.

So I wonder: how many (m)s per move of my engines analysis would be already somewhat reliable? And at what point would it become superior to a human GM cracking a position for 20 minutes in a classical match? And would there ever be a reason to let it work on a position any longer than that (besides expanding opening theory)?

Thanks,

I had this discussion years ago with a guy from the local chess club.  He would always let a position he was wondering about run overnight, or longer.  I have always followed Steve Lopez's (chessbase guru) advice that 10 seconds per move is optimal.  I was proven correct.

Stockfish is amazingly strong even at blitz speeds, recently crushing other computer engines to convincingly win the computer chess championship.  No one would bother with a time handicap match anymore, because the computer would easily crush the human.  For a 1050 player like yourself, there is just no scenario where you'd need to let stockfish run for more than a few seconds.  Even for an master level correspondence player that spent hours on a position, 30 seconds on the game moves (per move) and 30 seconds on each important "what if" position would likely be sufficient.  The live feed on chess.com of the World Championship had 30 seconds of stockfish displayed, and it did differ some from an independent web site running stockfish on a very powerful set up without the 30 second limit (Sesse.org I believe it is; it is considered "the" place for live computer analysis of high level chess games.)  The differences were important to world championship level chess, but not really to the rest of us.