The Blindfold King vs The Machine

The Blindfold King vs The Machine

By GM Timur Gareyev

This year I am aiming to set a world record playing 47 games blindfold simultaneously. I embrace the body, mind and spirit connection as a holistic approach to developing my craft. The ‘show’ will only get better as I integrate the arts, sports, and science behind what I do.

Working with neuroscientists is particularly exciting considering the mystery and myths behind blindfold chess. I was recently excited about an opportunity to get to learn more about my brain as I continue along my blindfold chess journey. The crew of UCLA (University of California in Los Angeles), neuroscientists and graduate students guided by Dr. Jesse Rissman, designed a challenge to find out what is happening inside my brain as I play blindfold chess.

Timur Gareyev and the team from UCLA. From left to right: Micah Johnson, Timur Gareyev,
Dr. Jesse Rissman, Nicco Reggente, and Jennifer Vallens. It was great to meet these guys.

As we got together over a picnic before our experiment was to begin, Dr. Rissman shared the findings from our first round of testing which examined the structure and activity of my brain. Among many parameters reviewed within a sample of 60 participants tested, I found I had distinctive advantages in my visualisation ability evident in the parietal lobe. Meaning: the electrons in my brain fire away much quicker and effectively as I navigate along the avenues of the ‘chess world without sight’.

My friend and partner in my blindfold chess journey, Jennifer Vallens, connected with Dr. Rissman’s lab to schedule the second round of testing. This time we took our experiment to a whole new level. Not only we were to get to see the structure and processes happening naturally in my brain, we were to find out how my brain responds to the blindfold chess stimuli. As such, I was set to play four blindfold games simultaneously while my brain was examined inside an MRI chamber. This allowed us to see the structure of my brain and how it operates under normal circumstances, before we were able to find out how my brain responds to the stimuli of blindfold chess.

Despite the clear idea and subsequently the desired outcome of the experiment, the work it took to design the specific program for testing and execution must have been almost insane! Micah and Nicco worked together on programming codes as well as facilitating manually the transmission of moves as I played the four games. Then the second part of the challenge included a position reconstruction memory challenge which I performed easily with a perfect score.

Timur Gareyev is being prepared for his 90-minute brain scan in the MRI machine, during
which time he played four games of blindfold chess against a computer, simultaneously!

During the first hour of the 1.5 hour-long MRI scan I played four games. The completion of the games was not necessary to collect the necessary samples. Just for kicks an hour after the scan finished we picked up the games where we had left off around about moves 10-15. Still blindfolded, I took another hour to finish them as Micah guided the process and conveyed the machine’s moves.

Micah and Nicco working on move transmission as Timur is contemplating his
next move (or even moves!) inside the MRI chamber while taking on the machine.

Interview with Timur Gareyev

Finally, here is the interview I gave ahead of the challenge with Dr. Jesse Rissman and his crew of UCLA neuro-science graduate students.

Could you please describe to us, in as much detail as possible, your strategy for playing blindfold chess? How is it different from when you are playing regular chess?

My strategy for playing a single game of blindfold chess is similar to playing a regular game over the board. When I close my eyes, I still retain the pattern in my mind. At times it is easier to visualise how the game will evolve playing blindfold as opposed to looking at the board in front of me.

When playing multiple players at a time, aside from visualising and acting upon each one of the games, I must see how let’s say the 35 games I am playing relate to my strategy and each other. The beginning of the game is the most critical and challenging in achieving flow. The positions are not as distinct and do not ‘light up’ the memory as easily as later on.

I must use special techniques to be able to remember the games. I may assign an avatar, a distinct symbol such as ‘N’ for a knight move to a game. I then notice the relationships between the boards numbers and opening lines. At times the formations on several boards are identical for 2-3 initial opening moves.

After each new move that you make, how much time (in seconds) do you typically need to think about that move/board before receiving the next move of the next game? We are curious to know, during this ‘rehearsal’ time after each move you make, what are you thinking about specifically?

Generally, after I make the move in my last game, I move on to visualise the next game before I announce the next player’s turn to play. At times if I have just played a move, I may need time for the last position to sink in, especially in the opening. At other times, I may still be working on anchoring the last game as to how it relates to all the rest of the games within the simul, especially at large events.

Once I am done recollecting the position in the next game, which generally comes in combination with seeing the image or sensing the voice of the person who I am playing, I announce the player’s turn to play. If I am focused and clear-headed, I am able to recollect the next game instantly within two seconds or less. Occasionally, though, I lose my train of thought and take up to 20-30 seconds to remember the position in the next game.

At times after I am done with the move on the last board, I jump on to announce the next player’s turn to move. I may do it sporadically without prior visualisation of the position ‘at hand’ and then I have to piece together the opponent’s move I am hearing with the position I am trying to remember at the same time. At times the voice of the player or the kind of move being announced will suggest what the position is.

On average, about how many seconds does it take you, after hearing a new move by the opponent, to decide your next move in: a) a single blindfolded game, b) when playing two games at once, c) when playing four games at once and d) when playing more than four games?

Depending on the part of the game and the type of position for all of the above-mentioned formats, I can go as fast as instantly and take up to 10 seconds per move. In critical moments when I need to calculate or reassess the position, I may take up to several minutes to move.

All of these questions must be considered too in terms of opposition. My average opponent at a blindfold simul is around 1500 USCF strength. Playing against stronger opposition or a particular opponent, say, if it is a computer, may require additional thought as to how I want to structure the game at any point.

For you to play four simultaneous games from start to finish (playing against a computer, which only takes about two seconds on average to make each move), how many minutes on average would that full game take you to play?

To play four simultaneous games it would take me 40 minutes to one hour, 10 minutes depending on the strength of the opposition and my desired outcome.

And how many minutes for six simultaneous blindfold games against a computer and what about too 8 games?

Six games would take one hour to one hour, 45 minutes. For 8 games more like two hours to two hours, 45 minutes.

Does it take about twice as long to play four full games (simultaneously and blindfolded against a computer), as it does to play just two full games in that same way, or even longer?

It takes incrementally longer to play, the more games there are due to the transition time. In the case of computers, the machine can be assigned a certain strength it will play to. In any case, when it comes to playing a computer, humans must exercise a particular style to avoid complications and outplay the computer positionally. That takes longer due to the extra strategic thinking required, while computers never resign until they are checkmated. Thus the time for each game on average will be about one a half times longer, which helps to show why I generally don’t play computers in blindfold simuls.

When you have a large blindfold simul scheduled, how do you prepare for it?

In preparation for a 35-player blindfold simul I visualise my opening moves for about an hour a day for several weeks. I also play several sessions with my chess friend who represents the 35 players as we play through the initial 2-3 moves. This the only extra training I do for mastering the initial stage of a large blindfold simul.

I do not prepare for smaller 10-player exhibitions, aside from following a generally healthy lifestyle. I like to do some yoga stretching and light exercise before, during, and after my games. As I started using a spin bike during my blindfold exhibition matches, my results have improved 100% in terms of my ability to focus, playing strength, and how refreshed I feel as I finish the session.

If we were to ask you to associate a given game-board with a particular person or context (e.g. Micah’s Chess Board vs. Nicco’s Chess Board or the Purple Chess Board vs the Green Chess Board) would that affect your strategy? If yes, why?

My initial moves will follow my pre-designed system. On move 2 or 3 if I know that a particular opponent is a stronger or a weaker player, I may avoid riskier approaches or be deliberately aggressive. In regards to using memory techniques in associating names, colours and numbers, I will rarely do that with simuls of 20 players or less. However, when it comes to exhibition matches of close to 30 boards, it is what I need to do to anchor to aid my memory. I mostly do that with numbers, voices, and players.

Certain numbers, voices, and players may be memorable on their own, as well as a subgroup. When I am facing 35 players blindfolded, my job is to break up the whole group into smaller groups I can assign certain themes to (if I can). Typically if I play with black pieces, I will face 1 e4 more than 50% of the time. What often times may end up happening is boards 2, 4, 6, and 8 all play 1 e4. Even better boards 12, 14, 16, and 18 may also play 1 e4. There you go, perfect correlation!

Another example with number association is if 3, 9, 13, and 21 are special numbers. Number 3 would stand for my birthday and I may play my favourite system of sorts on later moves (move one will always be the same according to my opening system.). Number 9 would stand for the Brazilian striker Ronaldo – I may build him in if I play a gambit line. Then numbers 13 and 21 may represent a wacky play or a gamble I may choose to play to help anchor the positions.

Would it alter your strategy if we had you play multiple moves on one board before moving on to the next game?

I have played multiple moves in a row on several boards and then gone back to the other positions before. To be precise I did that in my last 35-player blindfold simul where after about six moves the players on boards 30-35 wanted to finish earlier so that they could leave. After I completed my regular round of boards 1-29, I went on to play at 10 seconds or less per move on boards 30-35. I went on to finish the match against those six players within an hour. After transitioning back to the regular simul, I was able to continue smoothly.

The above article appeared in the July 2016 of the British magazine CHESS

CHESS Magazine was established in 1935 by B.H. Wood who ran it for over fifty years. It is published each month by the London Chess Centre and is edited by IM Richard Palliser and Matt Read. The Executive Editor is Malcolm Pein, who organises the London Chess Classic.

CHESS is mailed to subscribers in over 50 countries. You can subscribe from Europe and Asia
at a specially discounted rate for first timers here, or from North America here.

Product: ChessBase 60 Minutes