Game Review: New Feature

Sort:
fluffytitbabbler
Wits-end wrote:

I don’t think I would upgrade for this alone. I think that chess.com means well with the new review, especially in light of the great number of youngsters. But then again, the young ones seem to advance much more rapidly so I doubt it will be useful to them either. 

No one should upgrade for this alone. There's plenty of new adult chess enthusiasts coming and it's easy to be under the impression, that a feature like this is worth the extra money. I used to be the same, I found chessdecode.com which is leaps ahead of this chess.com gimmick, have a few games analysed and went home not any smarter than I was before. There's no replacement for hard thinking, or maybe a proper chess coach.

Ziryab
n0time4chess wrote:
Wits-end wrote:

I don’t think I would upgrade for this alone. I think that chess.com means well with the new review, especially in light of the great number of youngsters. But then again, the young ones seem to advance much more rapidly so I doubt it will be useful to them either. 

No one should upgrade for this alone. There's plenty of new adult chess enthusiasts coming and it's easy to be under the impression, that a feature like this is worth the extra money. I used to be the same, I found chessdecode.com which is leaps ahead of this chess.com gimmick, have a few games analysed and went home not any smarter than I was before. There's no replacement for hard thinking, or maybe a proper chess coach.

 

Books.

 

fluffytitbabbler
Ziryab wrote:
n0time4chess wrote:
Ziryab wrote:
n0time4chess wrote:
Ziryab wrote:
n0time4chess wrote:

I agree that's quite bad on top of everything else, that's wrong with this "feature". I'm sure, there's going to be a lot of these.

Analysing the suggested engine lines is not a simple and easy task. In order to generate the advice text correctly, the software would need to go into variation starting with a bad player's move, understand why is it bad and generate text description. Then it needs to pick the best engine move, understand why is it best and generate the text that takes into consideration the consequences of the bad move.

It's not as simple as saying, you should develop a piece. Yes, Qd2 is a developing move, but it's doing so much more. I mean look at the position after 12. bxc3 Qxc3. Although the material is even, evaluation is almost -2 in blacks favour. Why? Putting it into words is hard even for experienced chess players, let alone some half baked software.

This feature might be somewhat useful for a complete beginner who makes a lot of basic mistakes, other than that it's just a shiny bling-bling to attract new diamond members, that's useless at best and confusing otherwise.

 

 

bxc3 is a blunder that loses a pawn and the right to castle

Is that supposed to prove that it's easy to come up with an explanation? 😂

 

Yup. 

Ok, grandmaster😂

What if I argued that loss of a pawn is compensated by material being equal, that if I didn't take that pawn, I'm down a piece. And lost right to castle by itself is not enough to cause the position to be losing?

And why Qd2 is the ONLY good move that keeps the position equal, even though I'm down material?

 

There are severe limits for a virtual coach. I’m not convinced that Chess.com is employing AI, but rather using code like ChessBase used for Fritz analysis two decades ago. Nonetheless, they could add substantially to the coach’s vocabulary. Maybe they will.

 

I agree that as it is, the new feature is more gimmick than helpful. I also think the confidence it gives to some could make them inflexible in their  preference for generic developing moves over making and assessing concrete threats (see John Watson, Secrets of Modern Chess Theory and Dan Heisman, Elements of Positional Evaluation).

 

The coach’s effort to highlight the game’s critical points is a step in the right direction, but this was done more easily in the old system because the game was always viewed from both sides and the evaluation slider was always prominent beside the board. It may be there still, but there are distractions seeking attention.

 

I’m not a GM, but I’ve coached many beginners up to the 1500 level, and some have gone on upwards from there, even exceeding my skills. I’ve been coaching youth chess players the entirety of the twenty-first century. I taught my own children in the twentieth.

I agree with what you say. My whole point was that it requires extensive knowledge and experience and teaching skills for a human to explain the moves and at this time afaik there's no AI to replace that. And even then, it's hard or impossible to stuff the explanation into a sentence or two.

Perfect engine moves are one thing, perfect human moves other and explaining both in human language yet another. It'll take a huge amount of time, effort and missteps for chess.com to get to the level of chessdecode.com, provided that's what they want, not just a shiny toy. Until then this feature is pointless and distracting, I think we both agree on that.

 

Chan_Fry
Martin_Stahl wrote:

I asked the same in a different topic/post. I could only get that to happen by making the browser window smaller than my screen (about 75% of the screen width).

Ah! Thanks for this, Martin. When I widen the browser window, then the buttons expand and I can finally see them. It still seems like a design oversight, or does everyone else keep their browsers open at full width all the time? (I rarely do, which is why I missed that simple trick.) I'll just have to remember to widen my browser window whenever I want to do a game review...

Martin_Stahl
Chan_Fry wrote:
Martin_Stahl wrote:

I asked the same in a different topic/post. I could only get that to happen by making the browser window smaller than my screen (about 75% of the screen width).

Ah! Thanks for this, Martin. When I widen the browser window, then the buttons expand and I can finally see them. It still seems like a design oversight, or does everyone else keep their browsers open at full width all the time? (I rarely do, which is why I missed that simple trick.) I'll just have to remember to widen my browser window whenever I want to do a game review...

 

I normally keep the browser full screen. The layout is responsive, so if you make it too small, something has to give, so you lose some of the text in that case.

 

If you make it small enough that the side panel drops below the board it will remove the text:

Martin_Stahl

I was asked to pass along the message that staff are aware of the concerns with the new Review and associated issues, agree with most of the feedback, and are working on fixes.

Wits-end
Martin_Stahl wrote:

I was asked to pass along the message that staff are aware of the concerns with the new Review and associated issues, agree with most of the feedback, and are working on fixes.

Thanks for the report @Martin_Stahl. You are one of the ModStars here in the forums, keep up the good work. You always have good input and must have the patience of Job! 

Ziryab

Okay, yes I’ve played the Four Knights once or twice, but this was the first and so far the only time I played the Halloween Gambit. The stats offered here are less useful than they might be (again, I don’t care, as I download my games and put them in a database with games played at other sites).

mrtb411

Beardy congratulated me on booting my opponent's bishop, then admonished them for moving it instead of capturing an undefended pawn with one of their pawns. Hanging their knight was a blunder while hanging their queen was an inaccuracy.

Martin_Stahl
mrtb411 wrote:

Beardy congratulated me on booting my opponent's bishop, then admonished them for moving it instead of capturing an undefended pawn with one of their pawns. Hanging their knight was a blunder while hanging their queen was an inaccuracy.

 

I can see hanging a queen being an inaccuracy if they were losing no matter what move was made.

Martin_Stahl
Mirjana_k wrote:

Just this moment, November 22, I got automatically the “ game review” instead what we had before, game report. For me as basic member was it one game analyse per day. I would like to know how does it works now? Can I analyse all games and every day? Unlimited?

 

It works like Game Report. You get one full Game Review a day, and all others get a low depth analysis with no accuracies and just a count of moves classified (excluding brilliant and great).

https://support.chess.com/article/364-how-does-the-game-report-analysis-work

Signal25

I no longer seem to have basic analysis. Why is chess.com chasing non paying members onto other sites which offer great features for free? I used to be premium on here but can't justify cost atm. 

Martin_Stahl
Signal25 wrote:

I no longer seem to have basic analysis. Why is chess.com chasing non paying members onto other sites which offer great features for free? I used to be premium on here but can't justify cost atm. 

 

Basic analysis still works. You get one full Game Review a day. After that, you'll get an upgrade dialog but when it's closed you'll get a depth 10 analysis that looks like this:

YoshBear

I use the Game Review after every game and I honestly don't get it. Most of the "Best Moves" suggested seem absurd to me. Most involve throwing away pieces away (e.g. trading a queen for a pawn and the like) just because you can. Maybe I'm too noob of a player to see 10 moves ahead and how throwing away a piece will benefit me, but even games I've won, I would have lost if I did some of the "Best Moves".

Martin_Stahl
YoshBear wrote:

I use the Game Review after every game and I honestly don't get it. Most of the "Best Moves" suggested seem absurd to me. Most involve throwing away pieces away (e.g. trading a queen for a pawn and the like) just because you can. Maybe I'm too noob of a player to see 10 moves ahead and how throwing away a piece will benefit me, but even games I've won, I would have lost if I did some of the "Best Moves".

 

I've never seen anything like that but if you have your depth set too low, it may have some moves that are marked best at that depth, but may not actually be best.

YoshBear

Thanks, I will look into this.

YoshBear
Mirjana_k wrote:

I don’t think that the computer would say that someone should trade a queen for a pawn. It must be mistake.

Yes, it's crazy. Maybe I'm too noob to see how it would be a benefit several moves in the future.