Master class

by admin on June 28, 2011

Last weekend I had an opportunity to attend a class for masters (and a few experts) that was led by grandmaster Varuzhan Akobian. Akobian was teaching kids at the Berkeley Chess School last week, and my friend Robin Cunningham, who also works at the Berkeley Chess School, asked an obvious question. “If we have him here giving classes to kids, why not a class for grownups?”

The price seemed a little bit steep at first. It would be $50 a person IF we could get 10 people to sign up. But when you compare it to, say, a chess tournament, it’s not that outrageous. A typical weekend tournament costs $100 or more for two days of play. Many times you can play a whole tournament without having any games where you learned something you will remember forever. Alternatively, for $50 you can get  a whole day (well, 6 hours) of instruction from a grandmaster, which is guaranteed to make a lasting impression on you. Which is a better deal?

Fortunately we did get ten people to sign up. I told Gjon Feinstein from Santa Cruz and Mike Splane from San Jose about the class, and both of them were eager to attend. Robin found six other people from the San Francisco Bay area.

Perhaps the most interesting participant was Uyanga Byambaa, who is originally from Mongolia but now going to college in the San Francisco area. She has been recruited by John Donaldson to play for the San Francisco Mechanics in the US Chess League! Her rating is still in the expert range (perfect for a fourth board in the US Chess League) but Robin tells me she is improving rapidly. Specifically, he said, she loves to attack. “If you ever play her, trade queens,” he advised.

So, what did Akobian cover in the master class? Well, I feel as if I should not tell you too much because, after all, I paid $50 for it. But I can give you the general outline. He started with two of his recent games. First was a victory against Kacheishvili in the last round of the 2011 National Open (held just a week earlier!) which gave him first place. (He tied with GM Loek van Wely but came in first on tiebreaks.) This game featured Akobian’s improvement on Black’s play in a recent game between Nakamura and Ponomariov; suffice to say that Akobian’s play rehabilitates the line for Black.

Next up was a game Akobian played in a very similar situation against Laurent Fressinet in the 2009 National Open, which he also won. This game especially impressed me because Akobian reached a position where I would have tried to play a flamboyant sacrificial attack. Instead, Akobian played a series of calm positional moves — all the while ducking a series of traps laid by Fressinet. The motto of the game would be “Keep it simple.” If you can win by quietly strengthening your position, that’s better than risking it all on an unclear attack. This game also showed that tactical mastery involves not just playing your own combinations, but also anticipating the combinations of your opponent. A deeply impressive game reminiscent of Capablanca.

Then came the most surprising part of the workshop. Akobian went over several endgame studies with us. Akobian said that solving studies formed an important part of the Soviet approach to training, and he personally solves about five or six studies a week. Mike Splane (I hope he doesn’t mind my saying this) seemed skeptical that such studies were really relevant to tournament chess, because the positions often look a little bit contrived. Akobian replied that studies build your imagination and creativity. They open your mind to new possibilities.

I would add that the motifs of checkmate and stalemate, which occur so often in studies, actually pop up more often than you would expect in actual tournament play. I can think of two games I’ve seen in the last month — one of my own, and one grandmaster game — where a reminder of those themes before the game would have been really useful. Perhaps I will give a ChessLecture on this topic.

Finally, after solving a few studies, we looked at one more game by Akobian. This was his game against Yuri Shulman that won a Best Game of the Year prize in the US Chess League. If the second game resembled a performance by Capablanca or Karpov, this one was more like Alekhine or Kasparov, with an intuitive piece sacrifice that gave Akobian complete domination over the board.

My only slight complaint about the workshop was not entirely the fault of Akobian. His plan originally called for a period where we would play rapid games against each other and then he would comment on them. I am not sure how this would have worked, with ten people playing five games. How could we have time to play the games and then have him make useful comments on each one? Still, I would have been very interested in getting grandmaster commentary on one of my games, even if it was a 25-minute game or a 10-minute game.

Unfortunately, what happened was that at the suggestion of one of the participants, we had a round of introductions. This was not a bad idea, but they went on way, way too long. Instead of a quick “My name is Dana Mackenzie and I live in Santa Cruz,” the introductions turned into a discussion of each player’s personal chess issues and foibles. Although Akobian did say some interesting things, I would have much preferred the time to be spent on playing games, as per the original plan.

All in all, I felt that it was worth going to Akobian’s workshop this time but I’m not sure that I would go to a second one. It would be wonderful to get some more personalized instruction from a grandmaster, with game analysis … but of course, that’s something you would generally have to pay much more for.

Print Friendly, PDF & Email

{ 13 comments… read them below or add one }

Mike Splane June 29, 2011 at 12:43 am

“Mike Splane (I hope he doesn’t mind my saying this) seemed skeptical that such studies were really relevant…”

They weren’t relevant in that context. Hearing the insights that only a GM can offer is what makes a class like this so valuable and that is what I was paying for. Going over studies is something I can do on my own and added no value in that setting. Obviously my point was not clearly made when I asked why that activity was relevent.

The whole day was a lot of fun but I left feeling somewhat disappointed. I was expecting more tips and insights than the very few nuggets he offered us. He seemed like a great guy, I would love to hang out with him, but I wouldn’t pay for additional lessons.


Brian Wall June 29, 2011 at 7:16 am

In 2006 I had the greatest job in the world for a Chessplayer, ICC commentator. Imagine analyzing games online for the whole world with Varuzhan Akobian, Larry Christiansen, Gregory Kaidanov, IM Bill Paschall, Vaisich Railich, Dan Heisman, Greg Shahade, Eugene Perlelshtein, Danny Kopeck, Roman Dzindzichashvili, Robby Adamson, Cyrus Lakdawala. I thought I had died and gone to heaven. I got paid $25 an hour, they got paid more. How did I get the job? I was kind to William Chandler, a Denver Chessplayer with cerebral palsy. He moved up from just making moves to becoming a manager at ICC. I knew there were much better qualified people than me, I knew it couldn’t last forever. Akobian laughed when I said on air, It’s not what you know, it’s who you know. One time I guessed about 20 moves in a row and Varuzhan suggested I should be playing in the GM tournament. Varuzhan had a habit of maybe lifting hand weights on his breaks. The sound manager Sandy and I used to joke it sounds like he is having sex every time he is off the air. Varuzhan amazed me with both his tactical skill and his strategic mastery. Everything seems so smooth with him. He went to the Armenian School of Chess with Aronian and Sargissian. Varuzhan didn’t remember playing me at the Levy Memorial in Denver.. He played the French, it was an even position but he just tacked back and forth until I lost the thread. Some of the people who I thought could do the job better than me eventually became ICC commentators like IM John Watson and Yermolinsky. Watson is an old friend going back to my teen years and Yermo’s online diaries were just awesome Chess commentary. My Denver friend Dean Samsarovich is a two minute specialist and he has played thousands of blitz games with Varuzhan. There is something about Akobian that makes me really root for him to succeed in this country, he is very humble, talented, hardworking and recently married.


Ashish June 29, 2011 at 8:01 am

@Brian: Let’s hope that being recently married works better for Akobian than it did for Topalov. Not that I’m rooting for Topalov.


admin June 29, 2011 at 8:12 am

Great stories, Brian! You’re making me think twice about my stubborn resistance to subscribing to ICC. It sounds as if you can have a lot of fun on ICC even if you never play a game.

I didn’t mention it, but I drew against Akobian in our only game. It was in the US Open in 2003, when it was held in Los Angeles. I played my favorite unsound opening as Black, the Marshall Defense to the Queen’s Gambit (1. d4 Nf6 2. c4 d5?!). That variation has worked amazingly well for me against super-strong players. Besides Akobian, I also drew GM Serper with it. Those are the only two games I have drawn as Black against GMs in my entire life! (I have about ten losses.) So I’m going to keep on playing it, although GM Jesse Kraai says that it is an embarrassment.


Robin June 29, 2011 at 2:23 pm

I may not be objective since the class was my idea, but I found Akobian’s comments on how to move beyond master helpful. He described the necessity and reason for having solid openings, especially as Black, and he described a method for developing mastery of those openings. His concrete approach to a wide variety of positions from the opening seemed very ‘Russian School’ to me (I realize he is Armenian). I also appreciated his description of playing for a win as Black without taking undue risk.

The part of the class I would have changed is the part mentioned by Dana. Brief introductions morphed into an hour of semi-wasted time that would have been better spent on a training game and a few comments. This segment was unplanned and I think it developed from an effort to respond to a request by us.

At $50, I thought the class was a huge bargain and I would have happily paid $80. I can’t think of another forum where 10 of us can have a strong GM go over 3 quite different games, say what they are thinking and answer our questions at a level appropriate specifically for Masters and Experts. In fact, while the event was being set up, I joked with the BCS folks that even if it were just him and me or him and 2 of us, I thought the price would have been reasonable.

Now I just have to find the time to put in the kind of worked he described as necessary for improvement.


Mike Splane June 30, 2011 at 12:19 pm

I agree with Dana and Robin that the Q & A was the least successful part of the class.

I disagree with their opinion that this was a poor choice of actrivity. I think the problem was in the extremely short answers that Akobian gave to our questions which did not lead to further discusssion. This should have been the best part of the day.

For example, and in these examples I’m trying to give the gist of the Q&A from memory. These are not exact quotes

Q. How do you maintain consistency?
A. I wish I knew.

I was expecting something like this:

There are two ways. I always reach middlegme positions I know how to handle and, although I may not always play the best moves I always avoid making blunders.

I reach good positions by careful opening preparation of solid openings that bring me into middlegame positions that suit my style.

I avoid bad moves with a two step process. First, I always ask myself why my opponent made his last move. I don’t consider my own candidate moves until after I understand his plans and threats. Second, after I select the candidate move I intend to play I ask myself “what am I missing?” to make sure I don’t accidentally hang something.

Another example
Q. How do you evaluate a position?
A. I compare my pieces to his to see which are better.

I actually considered this a useful answer, as short as it was, but I was hoping for something like this:

I start by counting the material to see who would win the endgame.

Then I look at king safety to see if one side has a good chance of playing for mate.

After that I look at the pawn structures and how they affect space, center control, and open lines. I identify any weaknesses – color complexes, weak squares and weak pawns.

Finally, I move on to consider piece placement. Are any pieces out of play? If one of mine is poorly placed I try to improve it. If one of my opponent’s is poorly placed I try to take advantage of this. If my opponent has a particularly strong piece, I try to eliminate it.

Finally, when I completely understand the key factors in the position, I form a plan to take advantage of the imbalances that favor me.

But for me, the key factor is always piece activity.

Ths is the kind of dialog I was hoping for. Answers like this can lead to further questions that can expand our understanding. We could ask, “Which openings do you consider to be solid openings and why?” Or “How do I identify middlegames that fit my style?” Or “What do I look for when I’m trying to figure out if a king is going to be vulnerable to attack? ”

He could reply to the latter question:
A successful kingside attack against a castled king requires a stable center, more attacking pieces than defenders in the king’s vicinity, and at least one of the defenders pawns to be advanced.

Then we could go on to discuss how advancing pawns in front of your castled king creates weaknesses and how to exploit those weaknesses.

This kind of dialog and structured thinking that can be applied in numerous positions is what I was hoping to get from the lessons.

A couple more examples.

Q.Why didn’t you play Bf3 here?
A. I looked at it but it didn’t feel right. I trusted my intuition.

Instead I expected an answer like this:”For two reasons. First, I think I will need to play f4 soon, so I don’t want the bishop blocking the f pawn. Second, although Bf3 consolidates my center control it also gives my opponent an extra tempo to consolidate his position. Right now his pieces are uncoordinated so I’m only looking for aggressive moves. Then we could have discussed the importance of pace in a game, when to press and when to consolidate.

Another example.
Q. How would you compare your thinking process to Silman’s recommendations? What do you do differently?
A. I don’t follow Silman’s thinking process.

I would have liked an explanation of the flaws in Silmans process, more about when it applies and when it doesn’t, how we could improve on it,
and so on.

BTW I completely agree that Akopbian has a great personality, and I also agree that the demonstration of how to use chessbase to build up your opening repertoire was quite useful, and I thoroughly enjoyed the games discusssions.


Jason July 3, 2011 at 5:16 pm

Interesting discussion. We are experiencing similar tensions in our higher-level classes as well. There is always someone who wants their chess lessons to be wordy nuggets of chess generalities. I think this is due to Silman’s popular books.

I personally think this is a misguided approach. Chess problems that appear over the board, more often than not, are very specific entities. What’s needed is concrete calculation that is aided by having familiarity with the positions. I personally think it is more useful to go over problems with a strong player– as many problems as possible– to get ideas about their thought process. Yes, you can do those problems at home, but seeing them worked on by a strong player is very useful. Even knowing what is intuitive vs. concrete to a strong player is quite informative.

Instead, many people want to sit around and talk about how their emotions over the board, how they get upset when someone plays the Caro-Kann, and how they get bogged down by theory. And can they get advice on how to deal with that? Really vague general stuff that I believe is just a way to avoid the real hard work of chess improvement.


Brian Wall July 4, 2011 at 9:01 am


I kept posting-“Crush the rusty Honeymooner!”on Kamsky’s Facebook during his second Topalov match and I’m glad it worked out.


Mike Splane July 5, 2011 at 10:19 pm


Just curious what is your rating?

You stated: “What’s needed is concrete calculation that is aided by having familiarity with the positions.”

My first reaction when reading that was to think it was a great description of how to get good at checkers.

But wehen you are dealing with a more complex game I think there are big problems with that approach. First, you’re never going to be familiar with even a small portion of the possible positions that can occur in a chess game. Second, most chess positions, I’d estimate about 90%, are not solvable by concrete calculation. If they were, computers would have been beating humans decades earlier.

Somebody (probably Tartakower, he was famous for his many witty sayings) once said “It is easy to figure out what to do when there is something to do. It is harder to figure out what to do when there is nothing to do.” What you call wordy generalities and avoiding hard work is how you develop the skills you need when tactics are not present.

I would also argue that it is harder to develop those positional skills than it is to learn how to analyze tactics, and it requires a lot more work.

BTW, After reading your comment I stumbled across a column on that argues the exact opposite of your point of view.

Amatzia Avni is an Israeli psychologist, who is a FIDE Master in both game and composition. He writes:

“The general public is under the impression that top chess players devote most of their allotted thinking time to calculating variations.”

“Contrary to this widespread belief, strong players usually invest little time in analysis. They tend to examine fewer lines, concentrating on general considerations and assessments of resulting positions. Russian psychologist Viktor Malkin, cited in Secrets of Chess Intuition by Beliavsky & Mikhalchishin (2002) states that the higher a player’s rating, the less he actually calculates.”

“Let there be no misunderstanding: in critical positions, when analysis is absolutely essential, stronger players will indeed perform thorough calculations. When they do that, their analysis will be superior to that of club players: wider, deeper and more accurate. But in the vast majority of positions, champions engage in less calculation than average players.”

“In summation, strong players naturally analyze more accurately than their lesser brethren. But they check fewer lines and pursue them only as far as they deem necessary. In approximately 95% of cases they will make decisions based more on intuition, judgement and their knowledge reservoir, than on concrete analysis.”

Here’s a link to the complete article


Jason July 6, 2011 at 2:07 pm

My rating is in the 2100 range. I held a 2450+ to a draw and beat a couple of masters a few weeks ago if you need to judge or dismiss my ideas based on my rating.

I think we must have a better understanding of what it actually means for a high level player to solve a position without calculation. Your summary quote says it best: “based more on intuition, judgement, and their knowledge reservoir.”

Let us dissect what this means. First, I do not believe that the “knowledge reservoir” is a matter of looking up generalities like, “two bishops are better on the open board” or “rooks belong behind passed pawns.” or, “my knight needs a good outpost, how do I create that?” No, the knowledge reservoir is filled with familar games and positions. A GM is looking at position X and, really without even consciously thinking necessarily (this is where intuition and judgement come from!) are comparing the pawn structure to some famous Kasparov game, where the plan was P,Q,R,and S, and oh, yes, that game that was played just a month ago where their bishop looked active but really was useless– better avoid that in this game.

This happens very quickly for these high level players. To take an example from this discussion– when Akobian says he didn’t play f4 “because it didn’t feel right,” He is in part saying, “Subconsciously, I knew that, in the thousands of positions I have seen that are similar to this one, this is not a very good plan.” Of course, normal players like ourselves would need to decide f4 is no good based on other determinations. But, and this is why I think even cursory comments in a GM lesson like this are invaluable, by working out on my own what Akobian’s intuition worked out instantly, I too am taking a step towards a deeper understanding. And I am learning what material I should be internalizing and what material I should be calculating. Furthermore, to ask Akobian to explain his intuition on this matter, a’ la Silman’s word-solutions, is kind of like asking you to explain why you know instantly that 6×3 = 18 or how you were able to read this sentence with comprehension. This is very difficult to explain.

I think the fMRI studies of chessplayers are informative to this point. GMs access their long term memory far more than regular masters, who spend a lot of time in short term memory (doing calculations).

But (and I say this from working with 2400+ players on a regular basis), they are calculating deeply and quickly, even in positions where “nothing is happening.” It is different than tactical calculation, but it is still calculation– assessments of late middlegame and endgame positions that will arise from various plans, for example. Call it positional calculation, if you will. And I still think the generalities can be mostly discarded for an examination of specfiic needs of the position. The thought process shouldn’t be, “OK, the 2 bishops are great in open positions, and I have an open position, so the two bishops will be great.” It is more like, “Are the 2 bishops better in THIS position? Let me examine a few lines and see if the bishops are really going to have the power and utility that I want based on concrete features of the game before me…..”


Jason July 6, 2011 at 2:10 pm

Oh, I should add that the fact that computers beat the pants off of humans only shows that pure calculation is a pretty good way to be good at chess…..

It only took computers so long because computer programmers had to learn a lot about programming chess.


Jason July 6, 2011 at 2:17 pm

Finally, finally…

I would also say that some of this “GMs don’t calculate” dogma is a little like the kids in college who “never studied” but aced all the exams. It might not be the “if X than Y” kind of conscious calculating, but deep analysis is still happening.


Bonuses August 2, 2012 at 8:42 pm

I’m still learning from you, as I’m trying to achieve my goals. I absolutely enjoy reading everything that is posted on your blog.Keep the information coming. I loved it!


Leave a Comment

{ 1 trackback }

Previous post:

Next post: