No screenshot?
And a brand new account. You must be talking about another account you had.
You haven't played any games here with this one, so we can't find what you were talking about in particular.
It is easier to discuss these things when you can use the particular analysis that you ran.
Don't rely to much on the engine. And yes best move tend to shift with dept in most positions. Use game review to look for mistakes, blunders and interesting ideas - and learn from that.
...The engine is recommending the obvious Re8. It's third choice is Qg5 (apparently saying that Qxg2 is so strong I can sac the rook)....you can see the recommended "Qg5 is best" ... but it was the third choice? Actually if I let it run to depth 32, Qg5 drops off altogether.
So the "is best" move, is just low depth recommendation?
This is likely because the candidate moves are fairly close in evaluation and the lower depth engine prefers one choice (like Qg5) and a deeper depth drops off this option because now a better move is evaluated (like Re8).
Similarly, sometimes the engine (even on deeper depths) will claim one move is "best" but then you inputting another move will change the "best move." This is because your inputting of the other line "forces" the engine to calculate deeper into this line and if it updates the "best" to this, then all it means is that upon closer inspection, the engine now believes this last line is best.
The deeper one runs an engine then the less frequently this happens, but this occurs even when super computers evaluate chess positions. Better is to not stress over a few centipawns and simply try to reason WHY certain moves are played. Even if an engine line isn't "perfect", you picking a move with a few centi-pawns less is better if you can try to reason why this line is good. The computer "best" move does little to help you if you don't know why this is best and what plans go in alignment with that move.
Hello!
I use chesscom's analysis / game review feature a fair bit after interesting games. Something which annoys me and I don't fully understand is:
In the left hand side of the image is the position after move 13. The engine is recommending the obvious Re8. It's third choice is Qg5 (apparently saying that Qxg2 is so strong I can sac the rook https://omegle.onl/ ).
In the game I played Re8.
On the right hand side of the image you can see the recommended "Qg5 is best" ... but it was the third choice? Actually if I let it run to depth 32, Qg5 drops off altogether.
So the "is best" move, is just low depth recommendation? Even if I've let the engine run? Also super annoying if you're trying to just click through it, you can't actually trust what it's telling you?
It always behaves like this for me. It's still good enough to point out your mistakes, but I've remembered quite far into some of the lines I play: memorizing a sub optimal move is not ideal. In this case opponent hasn't played very well so memorizing is irrelevant but you get my point.
Is the takeaway that I just have to be more patient and let it run while ignoring the engine for dummies features? Or am I misreading?