Latest Updates:
Very Hot Topic (More than 25 Replies) What is the best way to analyze your games? (Read 47642 times)
Pawnpusher
Senior Member
****
Offline


I Love ChessPublishing!

Posts: 273
Joined: 01/04/18
Gender: Male
Re: What is the best way to analyze your games?
Reply #66 - 12/24/19 at 11:18:32
Post Tools
I am very fond of Nibbler too, I think it is very useful in going over games.
  
Back to top
 
IP Logged
 
Zenchess
YaBB Newbies
*
Offline


I Love ChessPublishing!

Posts: 3
Location: Omaha
Joined: 10/09/18
Gender: Male
Re: What is the best way to analyze your games?
Reply #65 - 12/24/19 at 09:48:37
Post Tools
I have found the chess GUI Nibbler to be very effective to analyse your games in a totally new way.  It's a chess gui for leela.  What's unique about it is it will show about 5-10 arrows on each move with a percentage for each one.  You can just click on the square the arrow points to and it will make that move.  So it is very efficient to quickly look over some variations and get an idea what moves are good or not.
  
Back to top
 
IP Logged
 
LisaStew
YaBB Newbies
*
Offline


I Love ChessPublishing!

Posts: 15
Joined: 05/07/19
Gender: Female
Re: What is the best way to analyze your games?
Reply #64 - 05/08/19 at 13:03:16
Post Tools
When I was training I was recording all my games, and then by rewatching the games I was making notes! I think this method works very well as you analyze your own game, what worked well and where you messed up, so in future you will escape from making the same mistakes
  
Back to top
 
IP Logged
 
Sylvester
Senior Member
****
Offline


Chess is the ultimate
game!

Posts: 314
Location: Canada
Joined: 07/03/08
Gender: Male
Re: What is the best way to analyze your games?
Reply #63 - 09/07/16 at 11:36:49
Post Tools
Also Ch 12 of The Chess Instructor: The Apeldoorn Analysis Questionnaire
  
Back to top
 
IP Logged
 
Sylvester
Senior Member
****
Offline


Chess is the ultimate
game!

Posts: 314
Location: Canada
Joined: 07/03/08
Gender: Male
Re: What is the best way to analyze your games?
Reply #62 - 09/06/16 at 00:32:54
Post Tools
See The Zugzwang Method by Daniel Munoz (an inexpensive Kindle book)
  
Back to top
 
IP Logged
 
msiipola
Junior Member
**
Offline



Posts: 95
Joined: 10/04/09
Gender: Male
Re: What is the best way to analyze your games?
Reply #61 - 04/06/15 at 06:55:07
Post Tools
dfan wrote on 01/26/15 at 13:31:02:
ReneDescartes wrote on 01/26/15 at 05:05:03:
@hictenunc Agree. Heisman's recent collection of amateur games should be called Houdini Analyzes Amateur Games. Heisman is afraid to offer his own assessments. And this is a guy who wrote a book on evaluation of positions! You get the feeling that computers have corrupted his brain.

Yeah, that was really weird. I think Heisman has a lot to offer beginners, and I enthusiastically recommend A Guide to Chess Improvement to everyone under 1500, but The World's Most Instructive Amateur Game Book (nice title, yeesh) is a real missed opportunity that encourages thinking about the game in a really unproductive way.


I'm not sure what happened with Dan Heisman or if anything have changed in his writings last years. But he is one of the best free resources you can find on the net, if you are a low rated player. But I have also noticed when Dan analyse games on ICC, he is talking about "and the computer says ...". It's like he don't (anymore) trust his own judgement?! Or is it only laziness?

In contrast to this I have found Nigel Davies lectures on Tiger Chess be very good. Pupils can send in games for analyse by him and he always gives his own opinion on the moves. And not what the computer says. According to Nigel if you don't have access to coach, use computers for finding tactical errors, and nothing else. Because computer plays in a computer style, not in a human style.
  
Back to top
 
IP Logged
 
katar
Senior Member
****
Offline


look another year went
by

Posts: 454
Location: LA
Joined: 09/21/05
Gender: Male
Re: What is the best way to analyze your games?
Reply #60 - 04/01/15 at 18:07:41
Post Tools
Sylvester wrote on 04/01/15 at 16:49:25:
So, what is a club player to do when he/she does not have a stronger player around to insert the timely comment “Have you thought about this?” 

Such a person would not be a "club" player nor would he be reading this forum. 

I presume a "club" player belongs to a club that has other human chess players.  Pretty simple to approach a stronger player or a peer and ask, "Hey would you mind looking at my game with me?".  I also presume a person who has access to chesspub.com has the means to publish games online with comments and to invite human feedback esp. for losses.  I note that receiving feedback is far less important than the act of publication because publication promotes honesty/objectivity rather than ego-stroking, excuses, and other forms of wishful thinking.  Anyone can publish video analyses/re-caps of games in a blog or on YouTube.

The above is just my 2 cents based on my experience first setting foot in a chess club or seeing a chess clock at age 25 and making USCF "expert" in 3 or so years while working professionally full time.  I reviewed games with equally-rated peers (1700-2100) and published game analyses on a blog and video forum.  I engine-checked my games only after several phases of human review.  YMMV.  We mainly play to have fun, so my advice is to do whatever you find enjoyable.
  

2078 uscf
Back to top
WWW  
IP Logged
 
Sylvester
Senior Member
****
Offline


Chess is the ultimate
game!

Posts: 314
Location: Canada
Joined: 07/03/08
Gender: Male
Re: What is the best way to analyze your games?
Reply #59 - 04/01/15 at 16:49:25
Post Tools
This is an interesting conversation on a complex topic that has been discussed in books devoted to this process.

As is often the case, the OP does not include an accurate description of himself/herself. This is an important element to know to give an adequate response. JohnG, quite correctly, points this out.

“Well I disagree pretty strongly with the advice that's been given so far. These things depend upon one's current level of play, which the OP doesn't give us.”

And, JohnG makes another valid point as well.

“Regarding advice from Botvinnik and the other old masters, let's be realistic. They did not intend this advice for sub-2000 adult club players….Even if they were addressing us, there's no reason to believe that they had particularly good insight into how middle-aged club players can best improve. A given piece of advice from them may or may not be applicable to the play at our level...”

I respect ReneDescarte’s comments. He is able to bring isolated chess concepts, found from various sources and periods in chess literature, together. He adds quality to the Forum commentary and he is one of the reasons I come back to visit the Forum from time to time.

Still, Rene writes with assumptions that, I think, JohnG was addressing, at least partially.

I don’t think anyone can dispute the futility of letting the computer do all the “thinking” for the player.

I do think that ReneDescartes’ comments assume a certain level of proficiency on the part of the player who should analyze his/her games.

The short list that follows is not for Rene as I think he knows it. It is for those, interested in this topic, who may not know it. I am referring to Bloom’s taxonomy of learning domains:
◦Knowledge
◦Comprehension
◦Application
◦Analysis
◦Synthesis
◦Evaluation

In analyzing a position you, at some point, must make an evaluation.

JohnG could probably have posted from the following viewpoint. How can a club player who is deficient in knowledge and comprehension expect to be able to analyze effectively, let alone evaluate?

I also think that ReneDescartes is making the assumption that “discovery learning” is best. Rene would probably agree that “most growth occurs at the edge of one’s comfort zone”. I also agree with that concept but only up to a point.

In offering suggestions for help with analysis, I would want to know the size of the person’s comfort zone so I could determine where this “edge” was situated. Even then, the “discovery learning” often needs to be guided. That is, a platform (stronger player) has to be put in place so the ‘voyager to discovery’ doesn’t fall into the abyss!

So, what is a club player to do when he/she does not have a stronger player around to insert the timely comment “Have you thought about this?”  I recommend (as if this is novel) that this club player consult his/her chess engine, but with an important stipulation. 

The chess engine should be used to:
a. offer feedback
b. push the boundaries of your present level of comfort

The first point assumes you have made an initial attempt at analysis yourself and therefore agrees with much that has been already written in this thread and others on this topic. If JohnG uses his engine to see whether or not he has missed a tactic during his game, I see that as a legitimate use of the chess engine. He is getting feedback. If he was using the engine to find the solutions to tactics exercises, I would think differently.

The second point also assumes an initial effort on the part of the player and the engine acts as the strong player asking, “Have you thought about this?” A few years back I studied the Art of Attack as intently as I could. I came to this book with a fairly good background in tactics (for my level) and an undeveloped understanding of attack. I went through the book using a chess engine to help me answer the questions that start with “What if Black/White had played …?” I don’t apologize to anyone for doing so. I was using the engine to push the boundaries of my level of comfort at that time. The engine allowed me to become an active reader of the material. Before computers, and I’m old enough to know, I was more of a passive reader. The chess engine allowed me to get answers to typical dedicated learner questions. When I finished the book I knew more than what had been written on the pages. Mind you, Vukovic continually led the way and, without him, my learning would not have taken place but I was transported safely beyond my comfort zone thanks to the chess engine.

I recall reading a post by GabrielGale which fits into the second point. Basically it was an episode where the computer was used to find a missing candidate move for a position.

This is only another aspect of this topic. It’s time to get off the soapbox.
  
Back to top
 
IP Logged
 
Jupp53
God Member
*****
Offline


be

Posts: 988
Location: Frankfurt/Main
Joined: 01/04/09
Gender: Male
Re: What is the best way to analyze your games?
Reply #58 - 01/27/15 at 18:58:35
Post Tools
hicetnunc wrote on 01/26/15 at 22:17:56:
Nigel isn't a crazy calculator anyway, so I found it extremely instructive to see how he coped with some of the messy positions that arose in my games. He didn't always find the best line per the engine, but he never let an advantage go, and that was usually with few and fast calculations  Roll Eyes. I found it way more instructive than seeing the first line of the engine, to be honest. It was also instructive to see when he was cutting the lines of the tree.


This is my experience with GM Khenkin too except he sometimes calculates on a high level I only can admire. If he would find always the best line he would be 2700+ or 2800+. This is not the point. The point is human thinking and how to improve it. Now I went in 3 1/2 years from 1820 to 1998 and I'm 61yo now. Younger players could get more out of it.

Tactics are sometimes impressive and probably every chess player loves them. But you don't need a trainer for tactics on my level. Solving training positions each day 15 min till 30 min and writing down the themes you didn't get is enough.

But the importance a GM lays on each single tempo, the reasoning over silent positions, this is something a trainer should show you.

When analyzing my games I have always two questions: Which tactics did I miss? And more important: What made this tactics possible? If I don't understand a position mistakes follow systematically.
  

Medical textbooks say I should be dead since April 2002.
Dum spiro spero. Smiley
Narcissm is the humans primary disease.
Back to top
 
IP Logged
 
ReneDescartes
God Member
*****
Offline


Qu'est-ce donc que je
suis? Une chose qui pense.

Posts: 1236
Joined: 05/17/10
Gender: Male
Re: What is the best way to analyze your games?
Reply #57 - 01/26/15 at 23:12:06
Post Tools
hicetnunc wrote on 01/26/15 at 14:48:10:
Provided the coach is strong enough (like your elo+300 and at least 2000), his tactics are good enough to provide new common patterns for you. At the same time, it's interesting to see how he analyzes : which line he goes for first, when he stops, how he assesses positions. Even if at the end of the day he misses a couple of ideas, or even if his evaluation isn't completely right, I believe it's worthwhile to see how the guy at the next level thinks about the game.[...]I believe human ideas that won't work in a specific position still have a lot of practical value for improvement, both because they are probably applicable in more positions than weird engine tactics, and because that's how your next opponent may think too.

I think that's it exactly. For the same reason, an unchecked book by Tarrasch is chock full of useful, typical, memorable ideas,  solid gold. If an engine finds some subtle, exceptional, unmemorizable sequence that refutes this particular application of the idea, what of it? If a superb teacher like Tarrasch thinks the idea is worth mentioning, then I want to see the idea, period. And I would gladly play like Tarrasch--mistakes and all!

Edit:Just saw the last post--very interesting about GM Davies.
« Last Edit: 01/27/15 at 13:23:49 by ReneDescartes »  
Back to top
 
IP Logged
 
hicetnunc
Full Member
***
Offline


"Do something scary every
day"

Posts: 233
Location: Paris, France
Joined: 01/04/03
Gender: Male
Re: What is the best way to analyze your games?
Reply #56 - 01/26/15 at 22:17:56
Post Tools
I had the opportunity to work some time with GM Nigel Davies, and I asked him explicitly to not use engines during the sessions, except on a few occasions when we wanted to check the result of our analysis (arguing with your teacher is good fun, because once in a while you may get lucky, not so with the engine !)

Nigel isn't a crazy calculator anyway, so I found it extremely instructive to see how he coped with some of the messy positions that arose in my games. He didn't always find the best line per the engine, but he never let an advantage go, and that was usually with few and fast calculations  Roll Eyes. I found it way more instructive than seeing the first line of the engine, to be honest. It was also instructive to see when he was cutting the lines of the tree.

My concern is when the teacher just let the engine take over. What if Houdini shows a nice tactical shot, and the teacher immediately shows it and post-rationalizes it as 'normal and obvious', when he may have missed it himself ? Isn't he showing his student chess in a distorted way ?
  

48 yo, 1920 elo
Back to top
 
IP Logged
 
Smyslov_Fan
God Member
Correspondence fan
*****
Offline


Progress depends on the
unreasonable man. ~GBS

Posts: 6902
Joined: 06/15/05
Re: What is the best way to analyze your games?
Reply #55 - 01/26/15 at 19:01:12
Post Tools
I believe it depends on what is being taught and how.

For me, a chess engine is just a calculator. I would expect my math teacher to do most of the math in her head, but occasionally, especially in higher level math classes, even the best Ph.D level math teachers will pull out the calculator. That doesn't diminish their ability to teach, as long as it's clear they're just checking to make sure the calculations add up.

If the topic is learning how to calculate in your head, you shouldn't use an engine. But if the teacher is just checking his or her math, I don't see why it's necessarily bad.

I enjoyed Heisman's early work, which I saw in Chess Life. His explanations, and even his examples were very similar to my own explanations to my students. But yes, it sounds as if he's gone off the deep end with a silicon weight tied around his neck in his latest book and lessons.
  
Back to top
 
IP Logged
 
hicetnunc
Full Member
***
Offline


"Do something scary every
day"

Posts: 233
Location: Paris, France
Joined: 01/04/03
Gender: Male
Re: What is the best way to analyze your games?
Reply #54 - 01/26/15 at 14:48:10
Post Tools
Bibs wrote on 01/26/15 at 13:09:12:
Coach = good. Coach + software = good, and checking for wacky tactics.
I play very rarely, but a clear example is from the last OTB game I played, a draw, versus a young IM. I missed a win at the end, we both missed in the postmortem, I found on my iPhone on the train home. Ah, oops!

Seems just bizarre to state that software should not be used by coaches.

Dunno, ask some super GMs, over 2750, see what they say....


Well, I guess when you coach a 2750 you need an engine to follow his train of thought in the first place  Wink

But okay, my point isn't that you shouldn't use engines in your analysis (like you did on your way home), but rather that a coach has IMO a better value by sharing his own thought process and even human calculations.

Provided the coach is strong enough (like your elo+300 and at least 2000), his tactics are good enough to provide new common patterns for you. At the same time, it's interesting to see how he analyzes : which line he goes for first, when he stops, how he assesses positions. Even if at the end of the day he misses a couple of ideas, or even if his evaluation isn't completely right, I believe it's worthwhile to see how the guy at the next level thinks about the game.

If he just goes : "oh Bxh7+ works here because there's this pawn on e5 and if blah blah blah blah blah (engine output)", then where's the added value compared to analyzing yourself with an engine ?

I believe human ideas that won't work in a specific position still have a lot of practical value for improvement, both because they are probably applicable in more positions than weird engine tactics, and because that's how your next opponent may think too.

I'm only a 2000 elo player, so maybe this way of thinking doesn't apply at higher pro levels (where it's more and more about accuracy and exceptions), but I firmly believe being exposed to a human thought process during the first stages of analysis is more beneficial for an amateur player
  

48 yo, 1920 elo
Back to top
 
IP Logged
 
dfan
God Member
*****
Offline


"When you see a bad move,
look for a better one"

Posts: 766
Location: Boston
Joined: 10/04/05
Re: What is the best way to analyze your games?
Reply #53 - 01/26/15 at 13:31:02
Post Tools
ReneDescartes wrote on 01/26/15 at 05:05:03:
@hictenunc Agree. Heisman's recent collection of amateur games should be called Houdini Analyzes Amateur Games. Heisman is afraid to offer his own assessments. And this is a guy who wrote a book on evaluation of positions! You get the feeling that computers have corrupted his brain.

Yeah, that was really weird. I think Heisman has a lot to offer beginners, and I enthusiastically recommend A Guide to Chess Improvement to everyone under 1500, but The World's Most Instructive Amateur Game Book (nice title, yeesh) is a real missed opportunity that encourages thinking about the game in a really unproductive way.
  
Back to top
 
IP Logged
 
Bibs
God Member
*****
Offline



Posts: 2338
Joined: 10/24/06
Re: What is the best way to analyze your games?
Reply #52 - 01/26/15 at 13:09:12
Post Tools
Coach = good. Coach + software = good, and checking for wacky tactics.
I play very rarely, but a clear example is from the last OTB game I played, a draw, versus a young IM. I missed a win at the end, we both missed in the postmortem, I found on my iPhone on the train home. Ah, oops!

Seems just bizarre to state that software should not be used by coaches.

Dunno, ask some super GMs, over 2750, see what they say....
  
Back to top
 
IP Logged
 
brabo
God Member
*****
Offline


Welcome chessfriend

Posts: 1068
Joined: 02/02/07
Re: What is the best way to analyze your games?
Reply #51 - 01/26/15 at 09:55:13
Post Tools
Just published on http://en.chessbase.com/post/wijk-aan-zee-the-pictures-and-their-stories-2-2

"Chess players are all the same. Right after his game, Giri rushed to see what the computer said
about his play. His trainer Tukmakov and girlfriend, Sopiko, had to drag him away to with the magic
word, "Dinner"."

Today it has become pretty standard to consult the computer. Now I do agree that coaches should provide more than what engines tell us. A student/ pupil must be emotionally strong to swap coaches when there is insufficient quality.
  
Back to top
WWW  
IP Logged
 
Jupp53
God Member
*****
Offline


be

Posts: 988
Location: Frankfurt/Main
Joined: 01/04/09
Gender: Male
Re: What is the best way to analyze your games?
Reply #50 - 01/26/15 at 09:18:55
Post Tools
A trainer using an engine while analysing with his trainee should be kicked at once. It's about human thinking otb. That's why a certain level of competence is absolutely necessary for the trainer.

Maybe I'm a bit harsh. 400 Elo points over the trainees level should be a rule of thumb till 2000 on the trainees side.
  

Medical textbooks say I should be dead since April 2002.
Dum spiro spero. Smiley
Narcissm is the humans primary disease.
Back to top
 
IP Logged
 
ReneDescartes
God Member
*****
Offline


Qu'est-ce donc que je
suis? Une chose qui pense.

Posts: 1236
Joined: 05/17/10
Gender: Male
Re: What is the best way to analyze your games?
Reply #49 - 01/26/15 at 05:05:03
Post Tools
@hictenunc Agree. Heisman's recent collection of amateur games should be called Houdini Analyzes Amateur Games. Heisman is afraid to offer his own assessments. And this is a guy who wrote a book on evaluation of positions! You get the feeling that computers have corrupted his brain.

Or maybe it was success. Heisman informs us that when a lower-rated player is doing well against someone stronger, he often entertains himself by wondering how the lower-rated player will manage to lose. This august form of entertainment may offer some quiet sadistic pleasure, but for a teacher to suggest it in print to his (low-rated) students is unforgivable.

A great contrast to Heisman's book is Fred Wilson's Simple Attacking Plans. Wilson reassures his students that a rating just means a player has had some good results recently. He tells his students that they can win and says, "everyone makes mistakes--don't even sit down at the board if you're not prepared to fight this guy" His very literate collection of all manner of games from amateurs and professionals has some extremely memorable insights--many attacks feature "a long strong queen move"--and concludes with a game from the 18th century. Oh--and Wilson doesn't rely on an engine.
« Last Edit: 01/26/15 at 22:14:28 by ReneDescartes »  
Back to top
 
IP Logged
 
brabo
God Member
*****
Offline


Welcome chessfriend

Posts: 1068
Joined: 02/02/07
Re: What is the best way to analyze your games?
Reply #48 - 01/25/15 at 20:47:01
Post Tools
hicetnunc wrote on 01/25/15 at 17:57:43:
Maybe I didn't get the joke  Smiley but I think this is a very important point.

I've seen numerous coaches using engines during sessions (notoriously the famous Dan Heisman), and I just don't believe it can add anything to the coaching. Actually, I'm pretty sure it susbtracts a lot, because you're not exposed to the thought process of the strong player, which is what this kind of analysis is all about...

I'm not even sure a good player can distance himself enough from the engine's output to filter the most relevant information : he might just get caught in it and present difficult engine lines as logical consequences of the position...  Sad

Also in post-mortems more and more smartphones, laptops are consulted even by grandmasters. Why losing time by analyzing something which an engine can clarify immediately while instead you can enjoy a beer at the bar.
  
Back to top
WWW  
IP Logged
 
hicetnunc
Full Member
***
Offline


"Do something scary every
day"

Posts: 233
Location: Paris, France
Joined: 01/04/03
Gender: Male
Re: What is the best way to analyze your games?
Reply #47 - 01/25/15 at 17:57:43
Post Tools
Maybe I didn't get the joke  Smiley but I think this is a very important point.

I've seen numerous coaches using engines during sessions (notoriously the famous Dan Heisman), and I just don't believe it can add anything to the coaching. Actually, I'm pretty sure it susbtracts a lot, because you're not exposed to the thought process of the strong player, which is what this kind of analysis is all about...

I'm not even sure a good player can distance himself enough from the engine's output to filter the most relevant information : he might just get caught in it and present difficult engine lines as logical consequences of the position...  Sad
  

48 yo, 1920 elo
Back to top
 
IP Logged
 
ReneDescartes
God Member
*****
Offline


Qu'est-ce donc que je
suis? Une chose qui pense.

Posts: 1236
Joined: 05/17/10
Gender: Male
Re: What is the best way to analyze your games?
Reply #46 - 01/25/15 at 17:05:42
Post Tools
That's his point, a cynical joke that nevertheless shows a failure to appreciate just what a coach can do that an engine can't, and undoubtedly also (though this is no fault) a lack of experience with in-person lessons with a strong titled player.
  
Back to top
 
IP Logged
 
hicetnunc
Full Member
***
Offline


"Do something scary every
day"

Posts: 233
Location: Paris, France
Joined: 01/04/03
Gender: Male
Re: What is the best way to analyze your games?
Reply #45 - 01/24/15 at 17:50:39
Post Tools
JohnG wrote on 01/24/15 at 00:14:14:
ReneDescartes wrote on 01/23/15 at 13:59:05:
However, I would point out that if one can show the game to a very strong player, the computer analysis is not really necessary.


Certainly, if you are a weak club player and you have just taken a lesson on ICC in which a strong player looked over one of your games then there is no reason to analyze with an engine later. After all, turning on the engine is what you just paid the titled player to do.  Roll Eyes


Oh man, I would consider a coach using an engine while reviewing your game pretty useless... If he doesn't provide his own analysis in the first place, you can as well use the engine yourself...
  

48 yo, 1920 elo
Back to top
 
IP Logged
 
Fromper
Senior Member
****
Offline


GrandPatzer

Posts: 378
Location: Chicago, IL, USA
Joined: 03/12/10
Gender: Male
Re: What is the best way to analyze your games?
Reply #44 - 01/24/15 at 15:09:04
Post Tools
MartinC wrote on 01/24/15 at 10:05:57:
Smiley There is of course a lot to be said for human insight where it can be had. In particular computers make absolutely no distinction in terms of how easy/hard a position is for a human to actually play.


And your phrase "where it can be had" covers the biggest advantage engines have over consulting a stronger player: availability.

The engines are available 24/7 for a small one time payment to buy them. Chess coaches cost at least three times as much just for a single session. And yes, discussing your games with stronger players at a local club is a good idea, if you have a local club you can attend regularly, and people aren't too busy playing their own games to look at yours, and assuming you're not already the strongest player at that club, ...

  

GrandPatzer!!!

1777 peak USCF rating - currently 1620 from coming back rusty
Back to top
WWW  
IP Logged
 
MartinC
God Member
*****
Offline


I Love ChessPublishing!

Posts: 2073
Joined: 07/24/06
Re: What is the best way to analyze your games?
Reply #43 - 01/24/15 at 10:05:57
Post Tools
Smiley There is of course a lot to be said for human insight where it can be had. In particular computers make absolutely no distinction in terms of how easy/hard a position is for a human to actually play.
  
Back to top
 
IP Logged
 
ReneDescartes
God Member
*****
Offline


Qu'est-ce donc que je
suis? Une chose qui pense.

Posts: 1236
Joined: 05/17/10
Gender: Male
Re: What is the best way to analyze your games?
Reply #42 - 01/24/15 at 04:38:56
Post Tools
Maybe on the ICC.
  
Back to top
 
IP Logged
 
JohnG
YaBB Newbies
*
Offline


I love spending time with
ChessPublishing...

Posts: 32
Location: USA
Joined: 05/18/14
Gender: Male
Re: What is the best way to analyze your games?
Reply #41 - 01/24/15 at 00:14:14
Post Tools
ReneDescartes wrote on 01/23/15 at 13:59:05:
However, I would point out that if one can show the game to a very strong player, the computer analysis is not really necessary.


Certainly, if you are a weak club player and you have just taken a lesson on ICC in which a strong player looked over one of your games then there is no reason to analyze with an engine later. After all, turning on the engine is what you just paid the titled player to do.  Roll Eyes
  

"The move is the idea." -John Watson
Back to top
 
IP Logged
 
Jupp53
God Member
*****
Offline


be

Posts: 988
Location: Frankfurt/Main
Joined: 01/04/09
Gender: Male
Re: What is the best way to analyze your games?
Reply #40 - 01/23/15 at 21:21:40
Post Tools
Fwiw

Bologan's way writing down at the end of the game a little summary seems meanwhile very usefull to me. B.e. Selected games 1985-2004, Russell enterprises 2007, p. 25:

"1. Don't be afraid to study current opening variations, or to use them, even against a stronger opponent. I abstained from my safe 3.Bb5+ to play an Open Sicilian for the first time in a long while, and won the game."

He adds to further points. This gives a structure and content to the further work.

  

Medical textbooks say I should be dead since April 2002.
Dum spiro spero. Smiley
Narcissm is the humans primary disease.
Back to top
 
IP Logged
 
Smyslov_Fan
God Member
Correspondence fan
*****
Offline


Progress depends on the
unreasonable man. ~GBS

Posts: 6902
Joined: 06/15/05
Re: What is the best way to analyze your games?
Reply #39 - 01/23/15 at 21:18:07
Post Tools
Let's put it this way: no modern author should publish without first doing a basic computer check of his lines.

Every modern GM that I've seen uses computers to check their lines, and some use computers to supply the lines!
  
Back to top
 
IP Logged
 
Uhohspaghettio
God Member
*****
Offline


I Love ChessPublishing!

Posts: 514
Joined: 02/23/11
Re: What is the best way to analyze your games?
Reply #38 - 01/23/15 at 19:26:20
Post Tools
MartinC wrote on 01/23/15 at 19:05:48:
There's no one on earth whose analysis will routinely stand up to modern computer checking, so no shame in doing it Smiley


Rubbish.
  
Back to top
 
IP Logged
 
MartinC
God Member
*****
Offline


I Love ChessPublishing!

Posts: 2073
Joined: 07/24/06
Re: What is the best way to analyze your games?
Reply #37 - 01/23/15 at 19:05:48
Post Tools
There's no one on earth whose analysis will routinely stand up to modern computer checking, so no shame in doing it Smiley

Whether what they point out is that often genuinely useful for humans trying to use/understand the position might be more interesting to question.
  
Back to top
 
IP Logged
 
Uhohspaghettio
God Member
*****
Offline


I Love ChessPublishing!

Posts: 514
Joined: 02/23/11
Re: What is the best way to analyze your games?
Reply #36 - 01/23/15 at 18:37:13
Post Tools
ReneDescartes wrote on 01/23/15 at 13:59:05:
Now in annotating a game for publication, an engine is essential today--as self-defense against embarrassment.


For me that would rank among the worst reasons to use an engine. That would suggest the author cannot stand by his inadequacies but is instead attempting to mask them. If you can't stand by your analysis without needing a computer then why should other people pay any attention to it? I think the standard reason isn't to protect the author but because if something is known to be better/worse by any method about a position it is generally stated. Agree with all you've said otherwise.
   
  
Back to top
 
IP Logged
 
MartinC
God Member
*****
Offline


I Love ChessPublishing!

Posts: 2073
Joined: 07/24/06
Re: What is the best way to analyze your games?
Reply #35 - 01/23/15 at 15:48:08
Post Tools
Wouldn't disagree much Smiley Just being pedantic, I suspect that by now there must be the odd GM out there who is mostly database/computer taught....
  
Back to top
 
IP Logged
 
ReneDescartes
God Member
*****
Offline


Qu'est-ce donc que je
suis? Une chose qui pense.

Posts: 1236
Joined: 05/17/10
Gender: Male
Re: What is the best way to analyze your games?
Reply #34 - 01/23/15 at 13:59:05
Post Tools
No one here has advocated not using a computer after analyzing by hand. However, I would point out that if one can show the game to a very strong player, the computer analysis is not really necessary. The trainer will point out the main tactical and positional points, and model the sort of analysis that one might do if one were a better player. Who cares if a fully-armed and operational John Nunn can find something that Capablanca or Lasker or Karpov wouldn't?

Many people have become grandmasters and world champions without engines, but no one has become a grandmaster or world champion with an engine but without discussing his games with strong human players.

I use the engine eventually, even when I have shown my games to a grandmaster, but it is by no means essential. And do not think for one moment that it's useful to see all the tactics (whatever that means). Even the engines miss tactics! And even if it were possible, it would not be particularly effective training. The point is to be able to do better yourself next time.

If there is no much stronger player to look your game and spot important points, then an engine is an acceptable partial substitute. Otherwise, it's a very mild help and a great pleasure.

Now in annotating a game for publication, an engine is essential today--as self-defense against embarrassment. But that doesn't mean you acquire more skill from analyzing with it. The status another would gain by pointing out a very hidden tactic has a great deal to do with human frailty and not much to do with learning. Reading Vukovic without corrections, you look at the variations, see them, and absorb patterns, blissfully unaware of any corrections. Reading Vukovic corrected, you look at the variations, see them, and absorb patterns, blissfully unaware of any corrections to the corrections. All the chess classics were written without engines, and the corrected versions are, say, 0.1% more educational? Or less, because they cause you to discount reasonably memorable patterns in favor of less memorable ones?
« Last Edit: 01/25/15 at 04:00:00 by ReneDescartes »  
Back to top
 
IP Logged
 
MartinC
God Member
*****
Offline


I Love ChessPublishing!

Posts: 2073
Joined: 07/24/06
Re: What is the best way to analyze your games?
Reply #33 - 01/23/15 at 10:23:35
Post Tools
You do have to be very careful drawing opening conclusions from how blitz games go though, because there's plenty of position types that are very much easier to play at one time control than the other.

If you're explicit about using it to rehearse your slow play openings to avoid early traps, reinforce the early ideas etc then OK.

For middlegame play you'd really want something a bit longer than typical blitz, but even longish quickplay tends to reduce to people hanging material Smiley
  
Back to top
 
IP Logged
 
CanadianClub
Senior Member
****
Offline


Greetings from Catalonia!

Posts: 416
Joined: 11/11/12
Gender: Male
Re: What is the best way to analyze your games?
Reply #32 - 01/23/15 at 08:58:17
Post Tools
Agree with Fromper in every point he wrote about blitz games.

In blitz games (5' + 3''/move in my case) you don't see some moves because of the dynamics of time controls. But sometimes you don't see some moves because that moves are not in your brain. There are ideas not assumed, not learned properly in some cases; even if the time control was 2hours your would not emerge with that manouvre/retreat/defensive pattern already.

Revisiting some games played with your heroes in similars positions is VERY usefull. And how they handle typical middlegames. Playing 10 blitz games in the nimzoindian would help A LOT in playing better your next nimzo in a 90' match for sure. That piece of information can't be underestimated. You have to knows that it was a blitz game but... they are your own games also.

You gain the better analyzing your slow games. Of course! Blitz games are interesting in some aspects (especially in the opening and how to play typical middlegames)? I absolutely think so.

Salut,
  
Back to top
 
IP Logged
 
Fromper
Senior Member
****
Offline


GrandPatzer

Posts: 378
Location: Chicago, IL, USA
Joined: 03/12/10
Gender: Male
Re: What is the best way to analyze your games?
Reply #31 - 01/22/15 at 17:27:28
Post Tools
Interesting thread, with a lot of conflicting opinions. Much like a chess game, the tension makes things more interesting. Smiley

I agree with bits and pieces of what everyone has said. I seem to be the only one who agrees with JohnG that different advice applies to different levels of players. Us class level patzers shouldn't be blindly following Botvinnik's training advice for masters, especially since there were no engines back in his day, so he didn't account for them in his advice.

I also disagree with the point that analyzing blitz games is a waste of time. I'm not saying you should spend 2 hours analyzing a game that took 5 minutes to play. But why not look it over quickly afterwards to see what you can learn? Spend as much time analyzing it as you spent playing it, and make a point of learning at least one thing from doing so.

At the very least, EVERY game has a point at which you got past your book opening knowledge, so you can spend 2 minutes looking up the opening and learning what the next book move is. Because it relates to your own game, this will cement the book knowledge in your head more concretely than just reading about the opening outside the context of your own games.

I also find that looking at my blitz games afterward to see what tactical shots I missed helps cement those tactical patterns into my brain. It's like doing tactics puzzles that way, with blitz exposing me to more positions in general, due to quantity of games played, and analyzing afterward helping to make the patterns more memorable.

I theoretically try to analyze every game of chess I play, regardless of speed. But I agree that I'll get a lot more doing deeper analysis on the slow games rather than the blitz games.

As for my general method, I do agree that checking with an engine should come last, but I disagree with anyone who says to skip the engines altogether. Having that computer blunder check to see what you missed is an essential step. I tend to look through my games on my own first, then consult reference books, then check an engine, and then publish my written analysis online. I try to do those first two steps on every game (even the blitz games), with the first step taking much longer and including written analysis only for the most important games (usually slow losses). The last two steps only come up on games that I'm spending some extra time on.

  

GrandPatzer!!!

1777 peak USCF rating - currently 1620 from coming back rusty
Back to top
WWW  
IP Logged
 
MartinC
God Member
*****
Offline


I Love ChessPublishing!

Posts: 2073
Joined: 07/24/06
Re: What is the best way to analyze your games?
Reply #30 - 01/11/15 at 10:43:27
Post Tools
Not boring positions, more badly one sided ones Smiley The sort where it does make sense to ask a computer just to confirm they really are dead and then leave them buried.

For some reason I reach very few even remotely equal endings.
  
Back to top
 
IP Logged
 
GeneM
Senior Member
****
Offline


Tournament winner gets
two fun filled knights!

Posts: 303
Location: near Seattle WA USA
Joined: 01/12/08
Re: What is the best way to analyze your games?
Reply #29 - 01/11/15 at 06:20:58
Post Tools
.
ReneDescartes wrote on 12/26/14 at 16:55:18:
Practice in long analysis without a computer ...

In the January 2015 issue of Chess Life, Andy Soltis wrote about the perils of misusing Fritz during your at-home post-mortem analysis.

Soltis' column said:
Quote:
But today the computer does all the work. There is little incentive to analyze games on your own, Svidler said. You can rely on an engine instead. "It's hard convincing yourself" to study your own games, he said.

.
  

GeneM , CastleLong.com , FRC-chess960
Back to top
WWW  
IP Logged
 
ReneDescartes
God Member
*****
Offline


Qu'est-ce donc que je
suis? Une chose qui pense.

Posts: 1236
Joined: 05/17/10
Gender: Male
Re: What is the best way to analyze your games?
Reply #28 - 01/10/15 at 22:36:12
Post Tools
It sounds like you are looking to analyze only nearly equal positions with interesting tensions, what Crouch calls "fighting chess" as opposed to technical chess. If so, you're missing a lot.

I love searching for new resources that I missed during the game in "boring" positions, endgames, and defense. Even after someone goes "significantly wrong for no particular reason," since a Lasker or Karpov would have a good chance to wrest half a point extra in this position, why shouldn't I  try to think more like them even if I have to do it in extremely slow motion? When I discover a new idea myself, I not only remember it, as has been remarked; I remember how I found it (this benefit can only be obtained from analyzing without an engine) and remember the sensations that led me not to find it during the game, e.g., thinking too much about such-and-such (this benefit can only be obtained from analyzing one's own, slow games). In other words, I find out what it is like to stretch myself a bit beyond my previous range of motion. Next time I play, I can do a little more, like a person who has just wiggled his ears or whistled for the first time, and now knows where to find the muscle.

Is this work less useful in positions without a certain kind if tension? --Well, wouldn't it be nice to be able to convert more like Fischer, defend more like Lasker, strike  more from boring positions like Reshevsky or win more equal endames like Carlsen?
« Last Edit: 01/11/15 at 02:21:25 by ReneDescartes »  
Back to top
 
IP Logged
 
MartinC
God Member
*****
Offline


I Love ChessPublishing!

Posts: 2073
Joined: 07/24/06
Re: What is the best way to analyze your games?
Reply #27 - 12/27/14 at 22:45:35
Post Tools
About most games not being worth analysing 'properly' - I've had 14 so far this season but probably only 4 games with sustained tension/a contextually reasonable standard of play.

Most of the rest of the games, while not without interest in the odd place, have one side or the other going quite significantly wrong for no particular reason. Mostly worth checking for the odd thing, but really not worth detailed analysis.
  
Back to top
 
IP Logged
 
Jupp53
God Member
*****
Offline


be

Posts: 988
Location: Frankfurt/Main
Joined: 01/04/09
Gender: Male
Re: What is the best way to analyze your games?
Reply #26 - 12/27/14 at 20:09:56
Post Tools
Uhohspaghettio wrote on 12/25/14 at 23:46:50:
I find Dana Mackenzie's examples a little strange. When I'm looking at the computer analysis, I'm trying to take away a general rule or principle. I'm not trying to look over tactical shots from a computer that humans shouldn't even consider (even if 0.0001% of the time it would have been worth it). For example I'm a pawn up here as white and I really wanted to see what the computer said about it, I also saved this position that surprised me based on static considerations.

2r2rk1/pb3ppp/1p6/n2p4/3Pn3/P3PNB1/1P3PPP/R3KB1R

* * * * * * * *
* * * * * * * *
* * * * * * * *
* * * * * * * *
* * * * * * * *
* * * * * * * *
* * * * * * * *
* * * * * * * *


I played b4 with the idea of making his knight move and half-opening the b file for the rook. Turns out b4 is a bad mistake and Rb1 is best. White is barely winning if he plays Rb1, significantly losing if he plays b4.


Thank you for this example and your comments.

For me it is showing how necessary it is to analyse with another player(s) if you haven't reached master level at a stage of the analysis.

(In the diagram I would think first about the lack of development and the c2 square. What would a master do?)
  

Medical textbooks say I should be dead since April 2002.
Dum spiro spero. Smiley
Narcissm is the humans primary disease.
Back to top
 
IP Logged
 
Smyslov_Fan
God Member
Correspondence fan
*****
Offline


Progress depends on the
unreasonable man. ~GBS

Posts: 6902
Joined: 06/15/05
Re: What is the best way to analyze your games?
Reply #25 - 12/27/14 at 00:24:39
Post Tools
Well, Nunn's criticisms of Lasker's play isn't the first to point out his tactical failings. Even before the advent of computers, Lasker's play was being criticized for this.

Nunn's point is that the hagiographies that elevate the players of the past are not based on concrete analysis. Today's players are far superior to those of the distant past. But that's another topic.

I read Hertan's book and really disliked it. Hertan's examples were quite often just bad. I thought Mackenzie's examples were interesting, and he explained the specific weaknesses his examples highlighted. I found MacKenzie's blog to be much more informative than Hertan's Forcing Chess Moves.

A further aside: I have used some of Hertan's exercises with my students. There are many useful diagrams in the book. The organization, claims, and writing style are what turned me off the book.
  
Back to top
 
IP Logged
 
ReneDescartes
God Member
*****
Offline


Qu'est-ce donc que je
suis? Une chose qui pense.

Posts: 1236
Joined: 05/17/10
Gender: Male
Re: What is the best way to analyze your games?
Reply #24 - 12/26/14 at 16:55:18
Post Tools
Mackenzie has a nice blog, but in this case I don't agree with him. He uses some computer-analyzed blitz games to point out certain human blind spots in seeking candidate moves. Fine--we knew that (and Charles Hertan who wants us to have "computer eyes" ?! told us if we didn't already know)--but what is needed is some practice finding surprising candidate moves in the context of a real game, where the situation looks perfectly ordinary.

Practice in long analysis without a computer with special attention to new ideas and candidate moves is the best exercise I can think to train this skill. Of course, analysis of any game whatsoever is useful, but such analysis is much more fruitful when applied to positions one has already considered carefully, as a corrective to one's previous idea of the game. In blitz one has a shallow idea of the game to start with, so there is not much thinking process to correct. Analyzing one's own blitz game is probably less useful than analyzing a master game in my view, whereas analyzing one's own slow game is at least as useful (I think one should do both the latter two).

But even with a blitz game, analyzing first without a computer is preferable. Immediate computer analysis mostly satisfies my curiosity and wish to say I have analyzed the game (maintenance), except for the opening, to look up the unusual opening ideas people throw at you in blitz. Maybe others learn more, though.

It is useful to do exercises with unusual candidates, but I cannot see how one can aspire to having "computer eyes." It is notable that Nunn, in his recent chess course, repeatedly chastises Lasker for missing tactical possibilities through being focused on a logical, human chain of thought. Lasker!!
« Last Edit: 12/26/14 at 22:24:59 by ReneDescartes »  
Back to top
 
IP Logged
 
Uhohspaghettio
God Member
*****
Offline


I Love ChessPublishing!

Posts: 514
Joined: 02/23/11
Re: What is the best way to analyze your games?
Reply #23 - 12/25/14 at 23:46:50
Post Tools
I find Dana Mackenzie's examples a little strange. When I'm looking at the computer analysis, I'm trying to take away a general rule or principle. I'm not trying to look over tactical shots from a computer that humans shouldn't even consider (even if 0.0001% of the time it would have been worth it). For example I'm a pawn up here as white and I really wanted to see what the computer said about it, I also saved this position that surprised me based on static considerations.

2r2rk1/pb3ppp/1p6/n2p4/3Pn3/P3PNB1/1P3PPP/R3KB1R

* * * * * * * *
* * * * * * * *
* * * * * * * *
* * * * * * * *
* * * * * * * *
* * * * * * * *
* * * * * * * *
* * * * * * * *


I played b4 with the idea of making his knight move and half-opening the b file for the rook. Turns out b4 is a bad mistake and Rb1 is best. White is barely winning if he plays Rb1, significantly losing if he plays b4.   
 
  
Back to top
 
IP Logged
 
Smyslov_Fan
God Member
Correspondence fan
*****
Offline


Progress depends on the
unreasonable man. ~GBS

Posts: 6902
Joined: 06/15/05
Re: What is the best way to analyze your games?
Reply #22 - 12/24/14 at 15:56:24
Post Tools
Gabriel, that's a great link you provided!

I agree wholeheartedly with Dana Mackenzie. Study without a computer first, then use the high-class abacus to check your work. I especially like his point that blitz games are a notable exception to that rule.

I strongly recommend anyone reading this thread to read the link that Gabriel provided!
  
Back to top
 
IP Logged
 
chk
God Member
*****
Offline


a pawn is a pawn

Posts: 1063
Location: Athens
Joined: 10/26/06
Gender: Male
Re: What is the best way to analyze your games?
Reply #21 - 12/23/14 at 16:08:47
Post Tools
Maybe it's just me, but as I mentioned above, the distinctive difference for me is that I tend to remember my games pretty well (including some of the "details to be found in the notes" or discussions/comments on them).

GM games are of better quality but not as memorable to me (even when I analyse them in detail).

Maybe it is the competitive conditions of OTB chess or just me, I don't know really.
  

"I play honestly and I play to win. If I lose, I take my medicine." - Bobby
Back to top
 
IP Logged
 
MartinC
God Member
*****
Offline


I Love ChessPublishing!

Posts: 2073
Joined: 07/24/06
Re: What is the best way to analyze your games?
Reply #20 - 12/23/14 at 14:03:51
Post Tools
Humm, yes analysing long games by hand is a good way to try and improve, but that's a slightly different matter from analysing your own games.

There's a 'maintainance' portion to doing that, which is nicely done by computers.

In fact I'd have thought that analysing game collections etc might be more worthwhile in general. For most people (definitely me!), the considerable majority of their games simply aren't worth analysing in any detail.

Either one side or another totally loses track of what plan(s) they should be following, someone randomly blunders horribly or similar.
  
Back to top
 
IP Logged
 
chk
God Member
*****
Offline


a pawn is a pawn

Posts: 1063
Location: Athens
Joined: 10/26/06
Gender: Male
Re: What is the best way to analyze your games?
Reply #19 - 12/23/14 at 12:15:56
Post Tools
I am 100% with Rene on that one.

My reasoning: Whenever I think or work with my OTB (long) games, I tend get these ideas or workings imprinted in my memory.

On the contrary, if I read a book and do not apply what I have learnt in real games, I tend to forget most of the material studied.

I propose putting the hard work in analysing your games - it may seem more time consuming at first, however it may be actually a short cut for improving given the long run.

And (Rene also touched on this) you do not need quantity here, but quality instead (i.e. long games or deeply analysing master games).

My 2c
  

"I play honestly and I play to win. If I lose, I take my medicine." - Bobby
Back to top
 
IP Logged
 
ReneDescartes
God Member
*****
Offline


Qu'est-ce donc que je
suis? Une chose qui pense.

Posts: 1236
Joined: 05/17/10
Gender: Male
Re: What is the best way to analyze your games?
Reply #18 - 12/23/14 at 09:38:56
Post Tools
@JohnG:Well, we disagree there. I think analyzing your own games has the benefits of correspondence chess (without a computer)--practice in finding your opponent's replies, calculating variations, etc.; practice in analyzing positions in depth. These benefits are very valuable even if your answers are quite inaccurate from the standpoint of an engine. Then there are the great benefits of self-knowledge--eradicating your typical mistakes. You do turn on the engine eventually (or the coach or strong player, or the publication audience); and at that point you learn a lot more if you have explored the position to the best of your ability. But even if you leave pieces en prise every game, playing slow games doing this sort of long analysis is one of the best ways to stop hanging pieces. Your accuracy improves steadily, which is all one could ask for.

I do have a soft spot for Botvinnik, but it's not just Botvinnik--see Yusupov's answer here, and he is certainly addressing class players.

http://www.qualitychess.co.uk/blog/?p=2152
« Last Edit: 12/23/14 at 12:49:35 by ReneDescartes »  
Back to top
 
IP Logged
 
GabrielGale
Senior Member
****
Offline


Who was Thursday?

Posts: 471
Location: Sydney
Joined: 02/28/08
Gender: Male
Re: What is the best way to analyze your games?
Reply #17 - 12/22/14 at 23:27:15
Post Tools
Christmas came a bit early with Dana Mackenzie explaining about engines analysis and blitz games , see http://www.danamackenzie.com/blog/?p=3320.

Before anyone diss Dana even though he is "only" a NM, he knows engines with his maths background.
  

http://www.toutautre.blogspot.com/
A Year With Nessie ...... aka GM John Shaw's The King's Gambit (http://thekinggambit.blogspot.com.au/)
Back to top
 
IP Logged
 
JohnG
YaBB Newbies
*
Offline


I love spending time with
ChessPublishing...

Posts: 32
Location: USA
Joined: 05/18/14
Gender: Male
Re: What is the best way to analyze your games?
Reply #16 - 12/22/14 at 21:03:55
Post Tools
Well I disagree pretty strongly with the advice that's been given so far. These things depend upon one's current level of play, which the OP doesn't give us. I'm guessing below 2000, but feel free to correct me.

I once subscribed to the belief that you should analyze your games by yourself, without an engine. I remember when I was 1700ish I played a game of G/90 and analyzed it extensively. All told I probably spent 8-10 hours looking over the game, using just a wooden board, a pad of paper, and a pen. Obviously this was over a few days time. What a colossal waste! I looked back over the analysis later and it was all utter nonsense. Entire variations rendered moot by a missed tactical shot right at the beginning. Sure "you probably learned something from the process," but I would have learned much more from 1/2 hour of engine analysis and then 7 1/2 - 9 1/2 hours of studying tactics and reviewing master games.

Strong players and weak players think and analyze differently. Strong players can look at a position and generate reasonable plans pretty easily. They are often analyzing to determine which plan is best. Weak players are struggling to find any half-way sensible plan at all. Strong players make basic tactical oversights rarely; weak players do so constantly. All of the advice about analyzing your games without engine assistance is fine if you are a master looking to become an IM, but for some 1700 player it is not particularly good advice.

Below 2000 try this:
1. Give the game to an engine for automated analysis. Use at least 5 seconds per move. Try a blunder threshold of 1 pawn and make it lower if you aren't getting enough comments.
2. Look over the engine analysis and see what makes sense and what doesn't. Spend a little time trying to make sense of things that seem confusing.
3. Turn the engine back on in infinite analysis mode and try to make sense of any lingering questions you have (whether about points the engine raise or your own concerns from the game). Really be stubborn and make the engine prove things to you. Do not accept something merely because the engine is stronger than you.
4. Look up the opening in books/database and find at least one early improvement on your play (sometimes this won't be possible, but usually it will be).

Regarding advice from Botvinnik and the other old masters, let's be realistic. They did not intend this advice for sub-2000 adult club players. They probably would not regard us as real chess players at all and they were not speaking to us. Even if they were addressing us, there's no reason to believe that they had particularly good insight into how middle-aged club players can best improve. A given piece of advice from them may or may not be applicable to the play at our level, but there's no reason to assume it is. This sort of advice gets passed around primarily because it is flattering to imagine ourselves "taking advice from Botvinnik." But really, we improve more when we admit that we aren't particularly good yet, forget the advice from Botvinnik and company, and look for more practical ways to improve. Engines are extraordinarily useful tools and we should make use of them. Yes, if you stop thinking for yourself and let the engine think for you you won't gain much. But you could just as easily make the same mistake analyzing with a coach (in fact, I find it easier to ask "stupid questions" with an engine, because there's no reason to be embarrassed). Just be stubborn and make the engine prove things to you and you will learn a lot.
  

"The move is the idea." -John Watson
Back to top
 
IP Logged
 
ReneDescartes
God Member
*****
Offline


Qu'est-ce donc que je
suis? Une chose qui pense.

Posts: 1236
Joined: 05/17/10
Gender: Male
Re: What is the best way to analyze your games?
Reply #15 - 12/22/14 at 19:38:42
Post Tools
No. What will benefit you is analyzing about one game a week for chess development, as Botvinnik and the Soviet school recommended. In order for this to make sense, the game would have to be played at at least a 90-minute time control for each side--preferable would be three hours for each side. Then spend a lot of your study time during the week analyzing it. If you analyze carefully a blitz game, you'll see what a waste of time both blitz and its analysis are: your mistakes are made because you don't have time to do a decent job.

The real trick is to see what you do in a slow game, when you have time to do much closer to your best, on every move holding yourself responsible for investigating everything dangerous for either side that might happen for the next few ply, concentraing hard for four to seven hours. When you analyze that game moving the pieces, you will see things that you miss not primarily because you lacked the time to see them, but because you lacked the capacity.

Bobby Fischer (former World Blitz Champion): "Blitz chess kills your ideas."
Nigel Short:"Blitz rots the brain."
Alexander Grischuk (former World Blitz Champion): "Anyone who analyzes a blitz game is an idiot."

  
Back to top
 
IP Logged
 
Michael Cobb
YaBB Newbies
*
Offline


I Love ChessPublishing!

Posts: 20
Location: Fairview Heights IL
Joined: 11/05/14
Gender: Male
Re: What is the best way to analyze your games?
Reply #14 - 12/22/14 at 12:52:41
Post Tools
I see my biggest problem.  Most the time I open up fritz immediately after playing a game without any self analysis.

I appreciate all the comments.  I have been playing anywhere from 10 to 15 games a day but don't review most of them. 

I am really thinking of only playing a few games a day and reviewing all of them for a while. 

Thanks again for all the great comments!
  
Back to top
 
IP Logged
 
brabo
God Member
*****
Offline


Welcome chessfriend

Posts: 1068
Joined: 02/02/07
Re: What is the best way to analyze your games?
Reply #13 - 12/20/14 at 16:48:29
Post Tools
MartinC wrote on 12/20/14 at 11:53:52:
True - the fun is definitely the main reason I analyse my games nowadays Smiley Modern computers are great for that of course - they find all sorts of incredible things so quickly.

Definitely analyzing can be fun. An example of some incredible stuff engines find, can be found in my article http://chess-brabo.blogspot.be/2014/10/interferences.html
  
Back to top
WWW  
IP Logged
 
MartinC
God Member
*****
Offline


I Love ChessPublishing!

Posts: 2073
Joined: 07/24/06
Re: What is the best way to analyze your games?
Reply #12 - 12/20/14 at 14:51:51
Post Tools
The best time to do something like that is directly after teams matches when there's a group of players around the place anyway.

I don't remember anywhere I've been doing formal training or such like. Group post mortems can be useful and interesting.
(Although obviously they do tend towards the light hearted Smiley)
  
Back to top
 
IP Logged
 
Pale Horse, Pale Rider
Senior Member
****
Offline


I Love ChessPublishing!

Posts: 287
Joined: 12/26/12
Re: What is the best way to analyze your games?
Reply #11 - 12/20/14 at 13:28:22
Post Tools
brabo wrote on 12/20/14 at 10:54:58:
katar wrote on 12/19/14 at 21:56:05:
You want others to review your published notes and hold you accountable for objectivity. 

I have published hundreds of my notes and very rarely got any comments. Only when i talk about general chess related stuff, some people start to react.

Also you need a lot of readers before you get any reaction at all. On my blog I notice averagely per 100 views, 1 reaction. So just a warning that you shouldn't expect much feedback from publishing notes.

I don't consider this strange. I have already lots of problems to find sufficient time to review my own games so surely reviewing somebodies else games will get very low priority (except for the opening when it involves an interesting idea which I can use in my repertoire).


For those playing in chess clubs: Prepare your game for discussion at the chess club. Work out some analysis and try to explain what's going on during the game. I don't know about better organized clubs but in my experience most chess clubs play blitz most of the time during "training". It's much more useful to take a demonstration board and explain your game to others.  You will get a lot of input. Of course the ideas will be very unfiltered but there will plenty of plans and different assessements of the position.
This is of course not really helpful without a club or without stronger players in the club.
  
Back to top
 
IP Logged
 
MartinC
God Member
*****
Offline


I Love ChessPublishing!

Posts: 2073
Joined: 07/24/06
Re: What is the best way to analyze your games?
Reply #10 - 12/20/14 at 11:53:52
Post Tools
True - the fun is definitely the main reason I analyse my games nowadays Smiley Modern computers are great for that of course - they find all sorts of incredible things so quickly.
  
Back to top
 
IP Logged
 
brabo
God Member
*****
Offline


Welcome chessfriend

Posts: 1068
Joined: 02/02/07
Re: What is the best way to analyze your games?
Reply #9 - 12/20/14 at 11:09:13
Post Tools
MartinC wrote on 12/20/14 at 09:13:50:
I think I have to dissent slightly. The primary reason to analyse your games is to get an objective assessment of how well you/your opponent played. That's fairly essential to doing anything.

You can get a good version of that from a computer, although obviously you do have to filter out some of the stuff depending on really computer style tactics etc!

If you then want to non trivially improve you'll have to train your analysis of course, which will involve not using a computer. Doesn't have to be specifically linked to your games though.

I don't know if the primary reason is to get an objective assessment of how good the game is played. I do know it is surely for a lot of people a valid reason to make an analysis. Sure analyzing your own games is an important tool to improve but at some point in your chesscareer there will be no improvement anymore possible. Still analyzing can be something interesting if improvement is unlikely to happen anymore. Getting an evaluation of the played game or just discovering some hidden elements can be very enjoyable by itself.

Also objective assessments from (very strong) engines can be afterwards incorrect. I wrote about that an article: http://chess-brabo.blogspot.be/2014/01/the-fake-truth.html
  
Back to top
WWW  
IP Logged
 
brabo
God Member
*****
Offline


Welcome chessfriend

Posts: 1068
Joined: 02/02/07
Re: What is the best way to analyze your games?
Reply #8 - 12/20/14 at 10:54:58
Post Tools
katar wrote on 12/19/14 at 21:56:05:
You want others to review your published notes and hold you accountable for objectivity. 

I have published hundreds of my notes and very rarely got any comments. Only when i talk about general chess related stuff, some people start to react.

Also you need a lot of readers before you get any reaction at all. On my blog I notice averagely per 100 views, 1 reaction. So just a warning that you shouldn't expect much feedback from publishing notes.

I don't consider this strange. I have already lots of problems to find sufficient time to review my own games so surely reviewing somebodies else games will get very low priority (except for the opening when it involves an interesting idea which I can use in my repertoire).
  
Back to top
WWW  
IP Logged
 
MartinC
God Member
*****
Offline


I Love ChessPublishing!

Posts: 2073
Joined: 07/24/06
Re: What is the best way to analyze your games?
Reply #7 - 12/20/14 at 09:13:50
Post Tools
I think I have to dissent slightly. The primary reason to analyse your games is to get an objective assessment of how well you/your opponent played. That's fairly essential to doing anything.

You can get a good version of that from a computer, although obviously you do have to filter out some of the stuff depending on really computer style tactics etc!

If you then want to non trivially improve you'll have to train your analysis of course, which will involve not using a computer. Doesn't have to be specifically linked to your games though.
  
Back to top
 
IP Logged
 
katar
Senior Member
****
Offline


look another year went
by

Posts: 454
Location: LA
Joined: 09/21/05
Gender: Male
Re: What is the best way to analyze your games?
Reply #6 - 12/19/14 at 21:56:05
Post Tools
Play rated tournament games against players at your level or slightly better than you, then go over each game with your opponent or a stronger player or a group of players.  Then annotate your own games and publish them somewhere.  Where?  It doesn't matter.  Maybe at a blog or in videos.  Publishing forces you to be honest and not to pretend that you played better than you actually did.  You want others to review your published notes and hold you accountable for objectivity.  This is an ego-crushing process because you need to learn why you suck and tell the world about it when publishing.  We all suck.  Nothing personal.  Your moves are just trivial decisions about small pieces of wood or plastic that relate to nothing in the real world.

Kasparov quoted Botvinnik on the importance of self-annotation, publication, and peer-review in the front material for his early book "Test of Time."  Perhaps someone with the book will provide the quote.

I don't think using an engine is necessary, but if you choose to do so, it should be the last step in your annotation process.
  

2078 uscf
Back to top
WWW  
IP Logged
 
brabo
God Member
*****
Offline


Welcome chessfriend

Posts: 1068
Joined: 02/02/07
Re: What is the best way to analyze your games?
Reply #5 - 12/19/14 at 19:19:38
Post Tools
On my blog I wrote many articles about this subject.
I recommend to start with http://chess-brabo.blogspot.be/2013/09/which-games-to-analyze.html and click on the included links to know more (unfortunately quite some are not yet translated to English)
  
Back to top
WWW  
IP Logged
 
ReneDescartes
God Member
*****
Offline


Qu'est-ce donc que je
suis? Une chose qui pense.

Posts: 1236
Joined: 05/17/10
Gender: Male
Re: What is the best way to analyze your games?
Reply #4 - 12/19/14 at 17:40:58
Post Tools
I do it in phases. First write down as soon as possible everything you can remember you thought during the game--variations you considered, reasons for your decisions. If you have a postmortem (and do if you can), write down your opponent's thoughts, too, and variations you reach in the postmortem. Then go to sleep. No engine!!!!

Next, wait until you have a few hours and go through the variations yourself in more detail, testing your conclusions, looking for new candidate moves, and writing down as many variations as you can see,  while moving the pieces in Chessbase and thinking hard as if in a (pre-computer) correspondence game. Extend this phase as long as you have the patience: this is some of the the best training material in the world for you, because it is customized for the very sorts of situations you tend to get into. If you can do it, extending it over a few sessions is great. No engine!!!!

Next, on another day (to remove temptation), turn on the engines and see what you missed and show the game to a more experienced player if you can. Of course, if a more experienced player offers to look at the game at the club, do it and include this material in the first run-through.

Then try to categorize your mistakes.  Hold yourself responsible (and therefore believe in yourself) and devise some training to address them. That's what Fischer and Botvinnik did. If you turn on the engine before phase 3, your curiosity will be satisfied, but you will not make use of the great customized training material.

I periodically look through my past games to see if I can see new patterns of mistakes. I concluded after one such review that I become too emotional and lose objectivity when queen trades are possible. I still have strong feelings when queens are on offer, but now I go on "manual override."
« Last Edit: 12/20/14 at 16:29:19 by ReneDescartes »  
Back to top
 
IP Logged
 
Fllg
God Member
*****
Offline


I Love ChessPublishing!

Posts: 647
Joined: 05/30/09
Gender: Male
Re: What is the best way to analyze your games?
Reply #3 - 12/19/14 at 15:20:41
Post Tools
Michael Cobb wrote on 12/18/14 at 23:53:22:
What are you guys and gals using? 


My own brain, honestly. Engines only after this!
  
Back to top
 
IP Logged
 
Göran
Senior Member
****
Offline


ChessPublishing is great!

Posts: 454
Location: Sweden
Joined: 02/13/08
Gender: Male
Re: What is the best way to analyze your games?
Reply #2 - 12/19/14 at 02:04:05
Post Tools
Take your score sheet and your board and pieces and start going though the game and analyse every move. Write down on paper your analyses with variations and verbal comments and mark every questionable move with one or two ?-marks as well. At the second walk through you could use your reference litterature (open, mid, endgame books) to check your analyses result and make new ?? and add to the verbal notes, elaborating on the previous ones.  Now it is early in the morning so you must go for a a couple of hours of good sleep.

When you wake up after 3 hours good sleep it is time to mail your analyses and verbal descriptions to your GM friend and then pick up the phone and call him to discuss the game and your analyses.

If you still have some issues (e.g. you don't really believe him/her) you could fire up latest version of Houdini, Stockfish and Komodo on separate computers and while sipping on your drink see the fascinating wonder when the computers are spitting out moves and numbers. (I never remember if it is the numbers or the moves I must learn).

Next day you make a list of all your errors (??-marks) and classify them and start to make a studyplan. And start study.

If not the entire process is feasable I'll pick steps 1 and 4.
  

What kind of proof is that?
Back to top
 
IP Logged
 
Michael Cobb
YaBB Newbies
*
Offline


I Love ChessPublishing!

Posts: 20
Location: Fairview Heights IL
Joined: 11/05/14
Gender: Male
Re: What is the best way to analyze your games?
Reply #1 - 12/18/14 at 23:53:35
Post Tools
Thanks in advance!
  
Back to top
 
IP Logged
 
Michael Cobb
YaBB Newbies
*
Offline


I Love ChessPublishing!

Posts: 20
Location: Fairview Heights IL
Joined: 11/05/14
Gender: Male
What is the best way to analyze your games?
12/18/14 at 23:53:22
Post Tools
Hello everyone!

I was wondering what everyone's advice is on analyzing your games.  I have tried chessmaster blunder alert and more recently using Fritz from Chessbase server. 

What are you guys and gals using? 

I know having a GM do it would be best!  Smiley
  
Back to top
 
IP Logged
 
Bookmarks: del.icio.us Digg Facebook Google Google+ Linked in reddit StumbleUpon Twitter Yahoo