Worst moves in a position

Sort:
Avatar of half_guard

Sometimes when analyzing positions, I think it would be useful for the computer to tell you what the worst moves in that position are. So like you'd flip what the calculation is looking for with the idea being learning how to best punish blunders which sometimes aren't so obvious and may be useful when you're playing against people who aren't going to be playing a perfect move on every move.

Avatar of chadnilsen

Hmm...

Avatar of rcu_21

ummmmm....

Avatar of Chessbovine

You probably want results of natural or logical moves that would be a mistake if played. What you'd get is hanging the queen or other pieces on purpose and destroying your pawn structure in front of your king, so I don't think it would have any instructive value.

But maybe it could have some sort of statistics from the games played, especially in the opening, when a certain bad move is often played in games on chess.com and have it as something like "common opening mistake to avoid", though I don't know how feasible it is. It makes probably more sense that if you do that opening mistake and analyze your game, you stop doing it

Avatar of poptran

,,,,,,,,,,

Avatar of will_buck
Chessbovine wrote:

You probably want results of natural or logical moves that would be a mistake if played. What you'd get is hanging the queen or other pieces on purpose and destroying your pawn structure in front of your king, so I don't think it would have any instructive value.

But maybe it could have some sort of statistics from the games played, especially in the opening, when a certain bad move is often played in games on chess.com and have it as something like "common opening mistake to avoid", though I don't know how feasible it is. It makes probably more sense that if you do that opening mistake and analyze your game, you stop doing it

exactly

Avatar of Fighter200
Chessbovine wrote:

You probably want results of natural or logical moves that would be a mistake if played. What you'd get is hanging the queen or other pieces on purpose and destroying your pawn structure in front of your king, so I don't think it would have any instructive value.

But maybe it could have some sort of statistics from the games played, especially in the opening, when a certain bad move is often played in games on chess.com and have it as something like "common opening mistake to avoid", though I don't know how feasible it is. It makes probably more sense that if you do that opening mistake and analyze your game, you stop doing it

 

Avatar of half_guard

I'm thinking something like how you can select up to 5 lines from the computer's analysis. It would be interesting to see like 20 and look at the ones that are in the bottom 10. Seeing the bottom 10 that are natural looking might be instructive. Especially with openings. I'm not talking about obvious things like hanging a queen but maybe moves that lose the queen in 2-3 moves later.

 

Avatar of Fighter200
half_guard wrote:

I'm thinking something like how you can select up to 5 lines from the computer's analysis. It would be interesting to see like 20 and look at the ones that are in the bottom 10. Seeing the bottom 10 that are natural looking might be instructive. Especially with openings. I'm not talking about obvious things like hanging a queen but maybe moves that lose the queen in 2-3 moves later.

 

 

Avatar of Chessbovine

@ElwenZhao please stop quoting without writing anything, that's just spam

@half_guard, you could use the opening explorer to see what lines in your openings usually lead to a win or loss for the colour you're playing and see which one have the best results.
I'm no expert but I believe if you trapped your queen in 2-3 moves no matter what moves you play, it's the same evaluation is if you just hang it

Avatar of jas0501

I think using analysis report and viewing the top 5 lines presented is a more productive approach. You can always make some reasonable looking but ultimately weaker or bad move from any position and review the top 5 response lines to get a picture of why they are bad.

With these settings in Analysis

you get a wealth of ideas, for example Ncxe5

 

Avatar of jdcannon

 I think it would be interesting or amusing, but I doubt it would be all the helpful. 

 

Usually, the worst moves aren't that hard to punish. Its the ones that are only just barely bad that are tougher to punish. 

Avatar of Sengdao

Maybe for April Fools, everyone's analysis becomes just a troll. For every blunder you make, you're given a "Brilliant!" response.

Book moves become mistakes. Good moves become blunders. Excellent moves become missed wins...

Now that's what Chess should be all about *cough*

Avatar of rcu_21

LOL

Avatar of chadnilsen
Sengdao wrote:

Maybe for April Fools, everyone's analysis becomes just a troll. For every blunder you make, you're given a "Brilliant!" response.

Book moves become mistakes. Good moves become blunders. Excellent moves become missed wins...

Now that's what Chess should be all about *cough*

I agree. That would be awesome.

Avatar of edr0nt
Sengdao wrote:

Maybe for April Fools, everyone's analysis becomes just a troll. For every blunder you make, you're given a "Brilliant!" response.

Book moves become mistakes. Good moves become blunders. Excellent moves become missed wins...

Now that's what Chess should be all about *cough*

+1 lol

Avatar of half_guard

Yeah it's hard to have silicon imitate dubious human moves. 

@chessbovine That's probably the best suggestion in the explorer. Especially in my games or the general online population as opposed to the titled players. 

In general, my thought was that if chess.com can score a move from 0-100, I want to see what the ones below 10 look like. I'd bet they aren't as obvious as some people here think or would at least take some thought to see why they are bad.

Avatar of Chessbovine

I'm not sure that's the best approach in the first place. Sure sometimes, you may not have any good moves and need to eliminate the worst ones, but otherwise you should look at checks, captures, threats, and if none of those offer any viable options, improve your pieces or restrict your opponent's.
The 10th stockfish move might not be a position a human would like to play.
What I mean is, if you try to learn to identify bad moves and avoid them, you might still find even worse moves or not play the best. If you learn to find the best moves, well, you won't do bad ones

Avatar of half_guard

I can play around with different move and see what evals do also.