Replaying Games

Higher level discussions, analysis of professional games, etc., go here.
User avatar
gogameguru
Lives in gote
Posts: 477
Joined: Sun Jan 09, 2011 9:18 pm
Rank: 5d
GD Posts: 0
Has thanked: 192 times
Been thanked: 357 times
Contact:

Re: Replaying Games

Post by gogameguru »

hailthorn011 wrote:Right. I started trying to memorize pro games myself. So far, I've memorized 23 moves of one game. I have the SGF but I play it out on an actual board from memory. I haven't been able to work on it very much because a lot has been happening lately. But hopefully I'll be able to continue working on it tomorrow.


Optionally, play all the way through the game quickly, trying to get a feel for the 'shape' of the game. Playing through on a real board using a printed game record is harder, because you have to find the next move yourself. This can be frustrating at first, but working out the next important area and looking there trains your intuition. While you are doing this, you are thinking about the moves, but not too deeply. You will get faster and faster at finding the next move and get stronger at the same time.

In contrast, studying on a computer is faster and easier, but also spoon feeds you to some extent and can encourage you to click through the moves too quickly. Computers are better if you use the setting many programs have to display a numbered diagram. I know other people won't share my opinion on this, these comments are just based on trying different things, asking many strong players and finding the most effective study method for me.

After that, you can replay through the game and try to remember the shapes and important areas move by move, as Dusk Eagle suggests. This time, try to play through more slowly and think about why the pro might've played each move and what the meaning of it is. Also think about what you would play at each move and why. Compare your reasoning with the pro's. This will help you learn a lot, but will also help you memorise the game, if that's what you want to do.

You can do a third pass over the same game if you still have time, to test how much you've remembered.

When you're replaying on a real board, use one hand for black and one hand for white. This makes replaying faster, and also activates more parts of your brain, possibly creating more relationships. This is my theory, but is based on reading extensively about neuroscience. I know some very strong Chinese players who agree with this method, though they don't feel any need to explain how it works. Again, I know other people may disagree with or laugh at this idea. That's fine, but I would recommend trying it first.

You can alternate this method with sometimes skipping the initial play through, so you don't have any prior knowledge of the moves. View each move as a problem position and think about how you would play, before seeing what the pro does.

This advice, along with doing lots of life and death problems, has helped several players who I've taught become dan level players quickly. On the other hand they may have become dan level in a short time anyway, because they really liked Go.

Lastly, don't pay too much attention to what I, or anyone else tells you. Experiment with all advice and find the method that works best for you. In all things related to Go, try to follow the advice of Bruce Lee, To paraphrase: "Study everything, take only what works, reject what is useless and add what is uniquely your own". (There are various versions of this quote, but this captures the sentiment well enough)
User avatar
Toge
Lives in gote
Posts: 313
Joined: Tue May 18, 2010 11:11 am
Rank: KGS dan
GD Posts: 0
KGS: Toge
Has thanked: 36 times
Been thanked: 63 times

Re: Replaying Games

Post by Toge »

Interesting skill this memorizing. I can't remember my own games, save the guessing based on common openings. In which game did I play that opening? I only play a couple of common josekis too. I doubt I could remember the complex ones like taisha and avalanches.

If professionals and strong players can reproduce positions just out of memory, what is the 'game' part that they play? Do they read or do they mostly recall from memory? I know I recall some principles and choose among them. Studying improves my game because I can recall more principles. Benefit of tsumego is to memorize shapes. I'm pretty bad in actual reading.
User avatar
gogameguru
Lives in gote
Posts: 477
Joined: Sun Jan 09, 2011 9:18 pm
Rank: 5d
GD Posts: 0
Has thanked: 192 times
Been thanked: 357 times
Contact:

Re: Replaying Games

Post by gogameguru »

Toge wrote:If professionals and strong players can reproduce positions just out of memory, what is the 'game' part that they play? Do they read or do they mostly recall from memory?


Reading is greatly assisted by having patterns stored away in your memory, because it speeds things up and helps you decide which moves you should be reading. Basically experience makes your reading more efficient.

At one point I did some extensive reading about this sort of thing. I'll try to tell you what I can recall right now, but if anyone knows more or wants to add in specific details, please do.

There have been many studies done of elite people in various fields like chess, tennis, football, motor racing etc... They were exploring the often quoted 10,000 hours of practise that it takes to create 'genius'.

Chess masters can easily remember many games, and play blindfold, too. Do they have better memories than people in other fields? Not significantly. Do the skills transfer universally? Generally not.

In tennis people have analysed Roger Federer playing Raphael Nadal. It seems like he has super-human reflexes. Does he? Well when they tested his reflexes they found they were pretty average.

So what some Scientists think is going on is that specialisation enhances the brains ability to process patterns and 'chunk' information in that specific field.

A person of average intelligence can only remember a certain number of things at once (I think it's generally quoted as 6-7 max, but see... I can't every remember that with certainty right now :)). The point is that a skilled chess or go player doesn't remember more pieces of information than the average person, they just chunk it more efficiently into bigger pieces, so that they can store the whole picture. That's the theory anyway.

Likewise, Roger Federer actually starts moving sooner than most other tennis players when playing against Nadal's serve (and that of other players). But this isn't reflexes, he starts moving before Nadal's racket even even makes contact with the ball. Because of intensive training his brain has developed the ability to process the patterns in the serve and predict quite reliably where the ball is going to go. Also, he is incredibly fit. Watching it on TV though, it just looks like he's got super-human reflexes :).

If you get a master chess player to play go (I have taught one), they will pick it up faster than other people and be quite strong at reading, because of the transferable skills. However, they won't necessarily be good at tennis. Likewise, Roger Federer is probably pretty good at table tennis, but not necessarily at go.
User avatar
kitanifan
Dies in gote
Posts: 27
Joined: Tue May 24, 2011 1:29 am
Rank: KGS 2 kyu
GD Posts: 0
Has thanked: 1 time

Re: Replaying Games

Post by kitanifan »

I'm 4 kyu and I can't memorize whole game, but just after game, I usually remember at least first 40 moves. But what I found interesting is that the more time passes after a "nice", interesting, tough match, the more i remember the moves. I actually find myself dreaming about game from a week ago. I guess my unconsciousness is mulling over the game.
Mef
Lives in sente
Posts: 852
Joined: Fri Apr 23, 2010 8:34 am
Rank: KGS [-]
GD Posts: 428
Location: Central Coast
Has thanked: 201 times
Been thanked: 333 times

Re: Replaying Games

Post by Mef »

gogameguru wrote:Chess masters can easily remember many games, and play blindfold, too. Do they have better memories than people in other fields? Not significantly. Do the skills transfer universally? Generally not.

In tennis people have analysed Roger Federer playing Raphael Nadal. It seems like he has super-human reflexes. Does he? Well when they tested his reflexes they found they were pretty average.

So what some Scientists think is going on is that specialisation enhances the brains ability to process patterns and 'chunk' information in that specific field.

A person of average intelligence can only remember a certain number of things at once (I think it's generally quoted as 6-7 max, but see... I can't every remember that with certainty right now :)). The point is that a skilled chess or go player doesn't remember more pieces of information than the average person, they just chunk it more efficiently into bigger pieces, so that they can store the whole picture. That's the theory anyway.

Likewise, Roger Federer actually starts moving sooner than most other tennis players when playing against Nadal's serve (and that of other players). But this isn't reflexes, he starts moving before Nadal's racket even even makes contact with the ball. Because of intensive training his brain has developed the ability to process the patterns in the serve and predict quite reliably where the ball is going to go. Also, he is incredibly fit. Watching it on TV though, it just looks like he's got super-human reflexes :).



I think I remember reading something about similar studies...One of the telling results with the chess masters was that they were able to quickly memorize and reproduce positions taken from real games, however fared no better than an average person when reproducing positions with random or nonsensical piece placement. I would imagine part of the abilities with players like Federer is that in tennis (much like in chess and go) there are certain shots in a a given position that make more sense than others (e.g. they will be easier to transfer more power or generate more spin). If you know how your opponent is trying to set up the point, you better interpret their body mechanics, anticipate their next shot, etc.

Perhaps one of the big differences is becoming better able to read "intention" from subtle cues and developing a better expectation of what should "naturally follow." It's something we do all the time in everyday life (a fact often exploited by illusionists and other entertainers), but for what I guess could be called artificially constructed activities (games, sports, etc) we have to build these expectations from scratch.
Post Reply