hemmed in

[CXUD]

I posted a thread in the amateur section about keeping pieces connected. This is the result of my attempt to put that information to good use but I got hemmed in badly anyways, I never see it coming until it's too late.

http://eidogo.com/#26eSREw

[emeraldemon]

A quick review, I think the main point to take away is about move 59. You had some weak stones, so your plan was to make more weak stones . Don't worry, with practice you'll improve.

[CXUD]

Ahhhh, I hadn't even thought of it that way.

g7 was to try to connect a line of stones from my upper base down into the lower part that never panned out. by threatening to connect with those trapped stones I thought I would give myself an extra turn to extend upward, invading the territory at the bottom with a safe connection and separating the right and left forces. Was it doomed, improbable, or just not useful enough?

[jts]

Hi CXUD! Given what you said last time about your strategy (you like to take empty points), I focused on identifying points where you played a big, empty point but should have played an urgent point, or vice versa.

7: *If* was a good move, or even an okay move, and white ignored it, then at q17 or r16 would be a great move. (In general it's true that if you make a threatening move and white ignores it, it's great to add a second move in the general area. But if it was a lousy threat to begin with, that complicates the calculations.) ------------ Also, in general when players are taking big points in an empty part of an empty board, regardless of what their strategy is, the point is on the 3rd or 4th line. (Why? Well, play more games and you'll get an intuitive feel for it. But the short answer is that if you loosely box out a bunch of territory on the 5th line, it's very easy for white to jump under and live in "your" territory.)

15: Adding another move around c7 is considered quite big, because the extra space on the side makes it nearly impossible to kill your group.

33: Here, I don't think this was urgent (especially because locally, you end in gote! Yuck.) The life of your stones wasn't at stake, since they are connected to the top side, and white can't improve the health of his stones much, either. Instead, R10 is the last huge point on the board, impossible to attack very effectively, and threatens to hem in Q15.

37: Tiny. Again, white doesn't have to make eyes in the corner... he can escape out to the left side. Any empty part of the board would be better.

39: Consistent with 37, but also small. If you're going to try to make good on this threat, start with B13, removing W's escape route. If he ignores that, then see if you can kill.

47: Not too bad, but this would be a good time to take R12, a crucial point to make R9 safe. (Tactically, W can still play K16 and then M16, or vice versa, so you have what we call "bad aji" - there is a cloud of doom hanging over this spot.)

51: You want to play closer to D11.

53&55: This may seem like a victory, but note that you had to play two moves which only barely rescue D11 (and do little else), whereas white got to play two moves that were good locally *and* solidified his control of the center.

59: The board is so open! Why would you play here?

65: This move is just far enough away to be useless.

83: If you want to keep some (potentially living) stones from getting boxed in, play L7.

87: As emeralddemon pointed out, this is just unnecessary.

Okay, past this point I think there are few open areas left, so our theme breaks down.

At this point, I don't necessarily think your biggest problem is keeping groups/strings connected, or even knowing when to take urgent points versus big points. The biggest problem is that you don't have a good sense of which groups are alive, which are dead, and where life and death hangs in the balance. Try doing these, for starters:

http://senseis.xmp.net/?BeginnerExercises

[CXUD]

Hi CXUD! Given what you said last time about your strategy (you like to take empty points), I focused on identifying points where you played a big, empty point but should have played an urgent point, or vice versa.

7: *If* was a good move, or even an okay move, and white ignored it, then at q17 or r16 would be a great move. (In general it's true that if you make a threatening move and white ignores it, it's great to add a second move in the general area. But if it was a lousy threat to begin with, that complicates the calculations.) ------------ Also, in general when players are taking big points in an empty part of an empty board, regardless of what their strategy is, the point is on the 3rd or 4th line. (Why? Well, play more games and you'll get an intuitive feel for it. But the short answer is that if you loosely box out a bunch of territory on the 5th line, it's very easy for white to jump under and live in "your" territory.)

15: Adding another move around c7 is considered quite big, because the extra space on the side makes it nearly impossible to kill your group.

33: Here, I don't think this was urgent (especially because locally, you end in gote! Yuck.) The life of your stones wasn't at stake, since they are connected to the top side, and white can't improve the health of his stones much, either. Instead, R10 is the last huge point on the board, impossible to attack very effectively, and threatens to hem in Q15.

37: Tiny. Again, white doesn't have to make eyes in the corner... he can escape out to the left side. Any empty part of the board would be better.

39: Consistent with 37, but also small. If you're going to try to make good on this threat, start with B13, removing W's escape route. If he ignores that, then see if you can kill.

47: Not too bad, but this would be a good time to take R12, a crucial point to make R9 safe. (Tactically, W can still play K16 and then M16, or vice versa, so you have what we call "bad aji" - there is a cloud of doom hanging over this spot.)

51: You want to play closer to D11.

53&55: This may seem like a victory, but note that you had to play two moves which only barely rescue D11 (and do little else), whereas white got to play two moves that were good locally *and* solidified his control of the center.

59: The board is so open! Why would you play here?

65: This move is just far enough away to be useless.

83: If you want to keep some (potentially living) stones from getting boxed in, play L7.

87: As emeralddemon pointed out, this is just unnecessary.

Okay, past this point I think there are few open areas left, so our theme breaks down.

At this point, I don't necessarily think your biggest problem is keeping groups/strings connected, or even knowing when to take urgent points versus big points. The biggest problem is that you don't have a good sense of which groups are alive, which are dead, and where life and death hangs in the balance. Try doing these, for starters:

http://senseis.xmp.net/?BeginnerExercises

The stuff until 59 was on my mind but I get nervous and mess up my priorities. Except 47, I'd have been to chicken to not respond.

59 I just like to try to get or ruin as much territory as I can but I'm rethinking that strategy (not making weak stones).

65 useless and going to get cut off as well

L7 definitely, another misstep.

The end was just me messing around although I did slack on memorizing those patterns when I was learning the game. I've started reading them and I'm getting most of them right.




http://eidogo.com/#2guwEWF

My attempt at applying the new information, not creating lots of weak groups of stones for instance.

Unfortunately I got a lower score than I usually do (I only got half the computer's score). I was on the ropes a lot less but the computer usually goes light on you if you are staking out the losing share of the board so I can't tell how well it worked. It did feel good though. I'm usually getting cut to ribbons with no idea why, now I'm able to not get hemmed in. Any ideas on what I did wrong with grabbing territory?

[topazg]

Some comments

[CXUD]

The opening moves you suggested were definitely better options, you can see the computer agrees. I was kicking myself for not taking the bottom center, it would have allowed me to do the same thing to the computer's territory that it did to mine.

On move 73 I had thought about taking that move (I didn't know if it would get the group on not, it just seemed damaging) but I was afraid it would just allow the group to extend out the side and get even more territory. I decided to take the better to live small than risk havoc tradeoff.

[emeraldemon]

A few suggestions:

(1) Play humans.
(2) If you want to play gnugo, take a 9 stone handicap.

Score doesn't really mean much (i.e. that you lost by more points this game). But if you are losing by 20+ every game, it's time to play handicap.

[Bill Spight]

A few more comments.



Two main points:

First, playing next to an opponent's stone, or diagonally adjacent to one, tends to strengthen it (when he answers). Beware of needlessly strengthening your opponent's stone.

Second, in general avoid playing too many stones near each other. Often two stones more than the opponent's stones in the same area is too many, especially in the opening.

[CXUD]

A few more comments.



Two main points:

First, playing next to an opponent's stone, or diagonally adjacent to one, tends to strengthen it (when he answers). Beware of needlessly strengthening your opponent's stone.

Second, in general avoid playing too many stones near each other. Often two stones more than the opponent's stones in the same area is too many, especially in the opening.

That's good to think about, I think it'll help me get more territory. I get worried and go overkill. I like rules of thumb, they keep me out of trouble.

A few suggestions:

(1) Play humans.
(2) If you want to play gnugo, take a 9 stone handicap.

Score doesn't really mean much (i.e. that you lost by more points this game). But if you are losing by 20+ every game, it's time to play handicap.

I do learn more from playing people but that's at my go club where they can teach me what I missed. Do you get more out of playing people?

I'd like to take a handicap but then I wonder if it will hide my mistakes rather than making it cost me.

[jts]

You can play humans easily on KGS.

[CXUD]

I've been meaning to get around to it. I also play on panda glgo.

What level is the gnu program at it's highest and at it's medium? Who can beat it?

[emeraldemon]

I'd like to take a handicap but then I wonder if it will hide my mistakes rather than making it cost me.

A lot of beginners have an aversion to handicap. Short answer: play handicap! Yes, it will hide some of your mistakes. This lets you focus on basics. Basically, I think you will improve most quickly if you are winning about half the games you play. At that level, improvement will be clear to you, because you will win more games! I made an older post about the same topic:

http://www.lifein19x19.com/forum/viewto ... 327#p24327

[CXUD]

can't do it, feels too weird.