simple mistakes

[robertg]

When I play blitz games or have to play fast I always make very simple mistakes like "I'll capture that group with my next move", it's his turn and he captures mine. That or similar mistakes (overlooking atari) happen in almost each game when I have to play fast. I always think there were some liberty left or he had one less. How can you reduce such blunders? I'm not a beginner. Just play a lot of blitz games?

[tchan001]

I also play a lot of blitz games. You can either play slower games or learn to think faster. As you do more go problems such as tsumego and tesuji, you will see patterns faster and this will also be very helpful.

[Bill Spight]

When I play blitz games or have to play fast I always make very simple mistakes like "I'll capture that group with my next move", it's his turn and he captures mine. That or similar mistakes (overlooking atari) happen in almost each game when I have to play fast. I always think there were some liberty left or he had one less. How can you reduce such blunders? I'm not a beginner. Just play a lot of blitz games?

Playing more blitz might help. Overlearning might help.

However, it is hard to believe that you have not overlearned atari. Perhaps it is a question of inattention or carelessness. If so, simply playing more blitz might not address the problem. What do you think about when you play blitz? Do you ask yourself what your opponent threatens? Do you take a second to look at the board? Do you play in the zone?

[daal]

I've read that teaching beginners could be vastly improved just by getting them to count liberties. I think that getting into the habit (particularly in your slow games) of counting liberties whenever stones come into contact would be a good idea for you. With a bit of practice, you will get faster at it so that in your blitz games you will be doing it at a glance and the blunders will become fewer. Also, you will start to recognize dangerous situations earlier. A string with three liberties in an enclosed group for example is the basis for many a shortage of liberty problem.

[robertg]

Bill Spight: I have the sequence read out, play it out and suddenly he captures my stones. What I think when I play blitz... Not much different from longer games. I might play more aggressive.

daal: Theoretically I should be good enough to see those things at a glance. I just somehow overlook it.

[Bill Spight]

Bill Spight: I have the sequence read out, play it out and suddenly he captures my stones.

And that does not happen in slower games? If not, what is different?

[robertg]

I'm normally very cautious and recalculate everything multiple times after each move. So even in long games I'll get to byoyomi at the (beginning) of the middle game most of the time.

[daal]

daal: Theoretically I should be good enough to see those things at a glance. I just somehow overlook it.

That's because of the difference between being able to do something and actually doing it - which is why I suggested trying to make it a habit.

[Bill Spight]

I'm normally very cautious and recalculate everything multiple times after each move.

Ah! That is something that Kotov warns against. His recommendation is to read each variation (that you do read) only once. And then trust your reading.

What may be happening is that, because you check and recheck yourself in non-blitz games, you can afford to be careless the first time, so you are. And then in blitz games, where you cannot check yourself, your carelessness shows.

Tommy Armour, the golf champion, said of the amateur golf practice of taking a Mulligan, "A cow could do it right the second time."

You might try, in regular games, Kotov's advice. Read each branch that you decide to read only once. Adding time pressure to that exercise would probably make it more difficult, so it is probably better to stick to regular games until you can do that.

----

Edit: BTW, that does not mean that once you have read a line of play and your opponent makes the next play on that line, that you just go ahead and make the next play without reflection. It is fine, and normally desirable, to read afresh at each turn. Even a variation that you have read before.

[robertg]

I'm normally very cautious and recalculate everything multiple times after each move.

Ah! That is something that Kotov warns against. His recommendation is to read each variation (that you do read) only once. And then trust your reading.

What may be happening is that, because you check and recheck yourself in non-blitz games, you can afford to be careless the first time, so you are. And then in blitz games, where you cannot check yourself, your carelessness shows.

Tommy Armour, the golf champion, said of the amateur golf practice of taking a Mulligan, "A cow could do it right the second time."

You might try, in regular games, Kotov's advice. Read each branch that you decide to read only once. Adding time pressure to that exercise would probably make it more difficult, so it is probably better to stick to regular games until you can do that.

----

Edit: BTW, that does not mean that once you have read a line of play and your opponent makes the next play on that line, that you just go ahead and make the next play without reflection. It is fine, and normally desirable, to read afresh at each turn. Even a variation that you have read before.

Ok I'll try. I should do the same with tsumego, right? When I do them, even very easy ones, I recalculate everything multiple times. Maybe because of the advice that one should be 100% sure about the solution.

[Li Kao]

When my opponent plays a self-atari I often marvel at how I missed that wondrous move and curse the dead spot in my reading. And only later I realize that the reason I didn't read the move was because it's a totally brain dead self-atari.

[illluck]

Ok I'll try. I should do the same with tsumego, right? When I do them, even very easy ones, I recalculate everything multiple times. Maybe because of the advice that one should be 100% sure about the solution.

Yeah, I think the advice that one should be 100% sure mostly refers to considering more candidate moves and responses (greater reading breadth) rather than reperforming things you've read out already.

[Uberdude]

I problem I have reading is I will read some deep sequence and then lose focus and forget where the stones are. And then read again to the same point again and again. Need more tsumego I suppose...

[peppernut]

daal: Theoretically I should be good enough to see those things at a glance. I just somehow overlook it.

This describes a problem in your mental game. I know, mine needs a lot of work as well.

There is a difference between knowing that you should be able to see those things at a glance and actually taking the glance on each and every move. It's the difference between knowledge and execution. This is a bigger difference than it seems.

There are a variety of mental mistakes that you can be making, such as trying too hard, rushing moves, moving without thinking, the so-called "analysis paralysis". It's different for each person and there's no single solution.

Experience helps, it's true. However, I'm not sure how much really. If you play more, you'll get better at Go, and you'll stop making these mistakes -- this is correct but it's almost a tautology. Who's to say that we don't stop making simple mistakes, and thereby improve at the game, and not the other way around?

I'm just about convinced that studying the game itself helps very little. Surgeons study for a decade before being allowed to operate on people, and surely one of the things they have learned is "don't forget anything inside the patient", yet it still happens. There is a class of errors that are not borne out of ignorance.

I wish I could say there was some particular thing that helped me, but I still have a lot of similar issues. I'm working on focusing on process. I think for this type of problem, the *how* you arrive at the decision is more important than the elements going into that decision. One advice I've heard is to practice your decision-making itself, one aspect at a time. For instance, you could play blitz games where your goal is to know the liberty count of all unsettled groups before every move.

[Bill Spight]

I'm just about convinced that studying the game itself helps very little. Surgeons study for a decade before being allowed to operate on people, and surely one of the things they have learned is "don't forget anything inside the patient", yet it still happens. There is a class of errors that are not borne out of ignorance.

Checklists -- I think in hospitals they call them protocols --, help.

Moi, I do not play go by checklist, but I know that some people use them. When I was playing bridge, I found a checklist helpful at the opening lead. Bridge is a fast paced game. That checklist took a while, and could even take half my time. In go, particularly blitz go, checklists might help, but maybe not on every move.