game review - best move questionable
Sort:
Jul 24, 2023
0
#2
If you leave review tab and go to analysis tab (right next to it) it’ll show you the evaluation difference and the lines.
thanks for the quick reply! i will dig into that and take a closer look. but in general those suggestions by the tool (computer, stockfish?) should be reliable, right?
Jul 24, 2023
0
#4
Yes. The position is just so lost that the computer suggests that giving up the queen might actually be the lesser evil
Schachmafix wrote:
in general those suggestions by the tool (computer, stockfish?) should be reliable, right?
The recommendations of the engine (in the Analysis tab) are reliable, yes. Especially in sharp tactical positions like this one.
On the other hand, anything you read in "Game review" is not.
Jul 24, 2023
0
#6
magipi wrote:
Schachmafix wrote:
in general those suggestions by the tool (computer, stockfish?) should be reliable, right?
The recommendations of the engine (in the Analysis tab) are reliable, yes. Especially in sharp tactical positions like this one.
On the other hand, anything you read in "Game review" is not.
Game Review uses Stockfish 16 NNUE so unless the engine didn't get deep enough in a complex position, they should be comparable.
Martin_Stahl wrote:
magipi wrote:
Schachmafix wrote:
in general those suggestions by the tool (computer, stockfish?) should be reliable, right?
The recommendations of the engine (in the Analysis tab) are reliable, yes. Especially in sharp tactical positions like this one.
On the other hand, anything you read in "Game review" is not.
Game Review uses Stockfish 16 NNUE so unless the engine didn't get deep enough in a complex position, they should be comparable.
They "should be", but they are not, are they? There are dozens of topics that show the myriad of bugs that game review has.
The current example is a good example. The engine prefers Ne4 to Be2 at all depths, except the ultra-shallow depth that Game review uses. Furthermore, the totally pointless and bad b4 (??) move is labelled as "Good". Furthermore, look at the explanation for the Be2 move: "This breaks a pin, enabling a queen to provide needed defense". In what world does that sentence make any sense, either grammatically or chess-wise? Especially that the queen is gonna get captured now, so it is an absolute mystery what Game review is talking about.
Jul 24, 2023
0
#8
magipi wrote:
They "should be", but they are not, are they? There are dozens of topics that show the myriad of bugs that game review has.
The current example is a good example. The engine prefers Ne4 to Be2 at all depths, except the ultra-shallow depth that Game review uses. Furthermore, the totally pointless and bad b4 (??) move is labelled as "Good". Furthermore, look at the explanation for the Be2 move: "This breaks a pin, enabling a queen to provide needed defense". In what world does that sentence make any sense, either grammatically or chess-wise? Especially that the queen is gonna get captured now, so it is an absolute mystery what Game review is talking about.
They are comparable. I'm on mobile right now, so can only load the engine without NNUE but Stockfish 16 does not prefer Ne4 and in fact give a much higher evaluation for black in that line.
Ne4
Jul 24, 2023
0
#10
Finally, using server side NNUE and depth 40
As to the played move classification, when one side is considered completely lost, pretty much any move is as good as any other.
The move explanations, are an attempt by the code to give human readable explanations of the why of a move. That's a hard thing to do programmatically and something the site is continuously working on to improve. That's something that not even people are always good at explaining correctly.
hi, i am new to this, and i am playing around with the analysis tool. in a game i just played the white queen got pinned by the black rook, and it is whites move. the computer tells me that bishop to e2 is the best move, but would knight to e4 not save the queen?