I'm not certain if this should be here or in beginners so please excuse me if it's misplaced.

I've been trying to get better at counting up settled points and making estimates based on the position. Ultimately, I'd like to improve at this to avoid over-play by staking out a larger area than I need to win and risk playing too thin. Also, I assume everyone that really plays Go can keep score because it seems quite necessary so I might as well get started. That being said, the estimator on the server I play on is quite unpredictable. Sometimes its very accurate and other times it'll be off by 30+ points. So I've began counting up my games as play and tonight I was in a game that I counted up and saw I was behind gy 35 points and there wasn't a whole lot of soil left to fight over so I resigned. I opened the sgf. with the Many Faces of Go which does a great job of counting and shows you what it does and does't count so you can tell what the score will be based on whether a group does or doesn't make life. I was disappoint to see the program put me behind by 8 1/2 points. I guess I never realized keeping score would really be hard, just time consuming. Has anyone come up with any good exercises for learning to evaluate the board better? My plan is to continue to count during every game until its natural (this only requires discipline, something any Go player can bring to their game at any level), and to brush up on my times tables so that I can count faster. Any thoughts? What was your rank (ballpark) when you started keeping score?

_________________
Thinking like a go player during a game of chess is like bringing a knife to a gun-fight. Thinking like a chess player during a game of go feels like getting knifed while you're holding a gun...

Did you actually try to count in the game? it's not that hard.

For starters don't count every single point one by one.. since as you said it is time consuming. Count by tens or fives.

Just go 2-4-6- meh around 10ish points here..

ok then around 15-20 points here.. few points here few there, komi... yes it will be pretty inaccurate but eventually once you practice it this way you will be able to count the score in 15 sec and as the time goes on you will learn to count every single point (even calculating the endgame in head).

For most general amateur play counting by comparison is easy, fast, and relatively accurate. It works by comparing regions of territory that are approximately the same size. Here is an example from one of my games. There was no komi, so scores are as they appear on the board.

Click Here To Show Diagram Code

[go]$$c
$$ ---------------------------------------
$$ | . . . . . . . . . . . . O a . . X . . |
$$ | O O O b . . O X . c X O . O O O X X X |
$$ | X X O . . . O X . . X X O O X X X O O |
$$ | X . X O O . . O . , X . X X X , X c . |
$$ | . X a X O . d . . O X O O O O O X X . |
$$ | . X . X O . . . O X . X O . O X X O X |
$$ | . O X X O O . . . . X X X . O X X O . |
$$ | . . O O . . O . O O X . . O . O X O . |
$$ | . O . X O O X O X X . X O . . O O X . |
$$ | . . X X X X X . X O X X X O . , O O . |
$$ | . X X O O O X . X O O . O . . . . . . |
$$ | . X O O X O X . . X O . . . . d . . . |
$$ | . X O X c O O . . X O . . . . . . . . |
$$ | . . X O . . . . . X O . O O . O O . . |
$$ | . . X O . . . . . X . O X O . O X . . |
$$ | . O O O . . X . . X . X X X . , X . . |
$$ | . . X X . X . . . . . . . . . X . . . |
$$ | . . . b . . . . d . . . d . . . d . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ ---------------------------------------[/go]

(a) both have small living groups surrounded by enemy stones, these about equal out. Just leave a mental ot the B's is a little bit bigger.
(b)W's top left corner is about equal to B's bottom left, so these cancel out.
(c)B's top right group - including the potential territory on the top should be about equal to W's left middle group. W has a bit more potential in the corner and may capture 2 more stones, so this makes up for B's (a) being larger than W's (a) and then a little more.
(d) That leaves W's top center and right side to make up for the entire bottom (minus the left corner). B looks to have maybe about 10 points more to me, but it is now B's turn, so he should get some sente to increase the gap.

What you are looking for is information on 'positional judgement' of territory, so search for related keywords. Forget about too approximative methods - the more accurate the more meaningful your counts are. The key is to know which points to count; once you know that, counting their number is learnt quickly. Note that a dead stone represents two points (the stone and its intersection).

Simply speaking, count those points the opponent cannot take away by peaceful reductions. This gives you two things: the territory count of Black's and White's difference, and the regions of "safe" territory intersections. Although there are more sophisticated additions for intermediate and advanced players, you can treat moyos in a simple manner: Once you have the count, you know how many points more a player needs to catch up on safe territory points in and outside the moyos.

When you can do this well, it is still time to learn additions.

Before the endgame, I would say that counting by comparing your territories with your opponent's should be fine. If one side has a significant lead in solid territory, it should detect it, and so help you decide on how to approach the rest of the game.

On the other hand, if using this method you find things are close, you might want to make a more precise count, and factor the part of the board that are claimed by one player but not yet solid territory.

Personally, I use the count by pair method when counting. The only tricky part is handling lone points, but once you find a way to do that (I keep track of them with my hand) it's as precise as you can get, and with practise it gets faster.

Also, I will point out that your opponent thinking time is the best moment to count, when you aren't in a complicated fight where you must use it as much as possible to out-read him.

I didn't do much counting until 10k-9k when I had a string of losses in games I thought I had easily won. I was using some type of rough territory estimation prior to then but I tend to overestimate black territory and underestimate white when I do that.

I use the pair counting method with a slight twist (which was suggested by someone on life in 19x19 - dr straw ).
1. Count all white's pairs of points and add on Komi and captured stones. Remember that captured stones still on the board count as 1 pair each. (For argument's sake say you count 23 pairs + 3 pairs for Komi = 26 pairs)
2. Subtract the result from 50 (50-26=24)
3. Count blacks pairs starting from the result in 2 (25,26,27 ...)
4. If you get over 50 black is ahead by that many pairs, or twice as many points.

Edit: forget my poor explanation and click on Dr. Straws link below

The advantage is that you only have to remember 1 number at any point in time (and pair counting makes counting dead stones easy).

I've found the method to be pretty accurate - more of a problem is remembering to count and responding appropriately to the answer. Too often I'll count that I'm ahead and play a pointless invasion anyway.

_________________
Still officially AGA 5d but I play so irregularly these days that I am probably only 3d or 4d over the board (but hopefully still 5d in terms of knowledge, theory and the ability to contribute).

This is an advanced topic.



At the 4th International Conference on Baduk in 2006 Lukasz Lew presented a paper in which, among other things, he reviewed the development of evaluation functions for go since Zobrist's program in the 1960s. There was no consensus among published functions, not even close. IOW, nobody knows how to evaluate go positions accurately before it is so near the end that the player with the move cannot do much about it. The programmers who developed these functions certainly had access to the methods that pros use to assess positions and knew how to program something like them.

The basic practical approach is to assess fairly stable and easy to count regions of the board and then go after necessary points in the more amorphous regions. IMHO, the best material on evaluation for humans is that by DrStraw, which he links to in his previous note.



Of course. It sucks, like all the others.



It sucks, too.



For your own games I would suggest counting at two points: first, around move 100 and second, around move 200. The second count will probably be too late in the game for you to change the result, but it will give you practice, and it will be close enough to the end for you to see how well you have estimated.

To improve your counting you need feedback to tell how well you are doing. For this I would recommend counting professional games. Start at the dame stage. That should be easy, right? Well, not really. Kyu players cannot always count the game accurately at that point because they cannot always tell whether a protective move is needed or not. (DDKs, in particular, should hardly ever resign, but should play the dame out, because of the appreciable chance that the opponent will make a big mistake at the very end.) If your count differs from the official result, try to figure out why.

Practice example:

The position at the end of the game, before the dame were filled. Komi = 6.5

Click Here To Show Diagram Code

[go]$$Bc Black has 13 more prisoners than White
$$ ---------------------------------------
$$ | O O O X X . . . X X O O O O O . O . . |
$$ | . O X X . X . X . X O O X X O O . O O |
$$ | . . O O X X X . X O X X . X X O O X O |
$$ | O O O O O X X X O O . X . X . X O X X |
$$ | O X O X O O X O . O X X X O X . X . X |
$$ | X X X X O O O O . O O X O O . X X X X |
$$ | . . . X O . O . O . . O O . O O O O X |
$$ | X X . X O O X X O X X X O . . O X X X |
$$ | X . X . X O X . X X O O O O O X X O X |
$$ | . X X , X X . . X , X X X X O X O O O |
$$ | . . O X . . X X . . . . X O O O . . . |
$$ | . . X X . X X X X X . X . X O . O X . |
$$ | X X . X X O X O O O X X O X . O O O X |
$$ | O X X X O O O . . O . . O O X X O X . |
$$ | O O X O . . X X O O X X O X X . X . X |
$$ | O . O O . X X O O X X O O X . X X X . |
$$ | . . . O . X O O . X O . O O X X . . . |
$$ | . . X O O X X O . X O . O O O X . X . |
$$ | . . . . . . . . O O . O . X O O X . . |
$$ ---------------------------------------[/go]

Then practice counting the game at move 200 or so. The count should differ somewhat from the final score. This game went 292 moves before the dame stage, so here is a practice example after move 250.

Click Here To Show Diagram Code

[go]$$Bc Black has 11 more prisoners than White
$$ ---------------------------------------
$$ | . . O . X . . . X . O . O O O . O . . |
$$ | . O X X . X . X . X O O X X O O . O . |
$$ | . . O O X X X . X O X X . X X O O X O |
$$ | O O . O O X . X . O . X . . . X O X . |
$$ | O X O X O . X O . O X . X . X . X . X |
$$ | X X X X O O O O . O . X . O . X X X . |
$$ | . . . X O X . . O . . . O . O O O O X |
$$ | X X . X O O X X O X . . . . . O X X X |
$$ | X . X . X O X O X X O O O . O X X O X |
$$ | . X X , X X O O X , X X X . . X O O O |
$$ | . . O X . . X . . . . . X O O O . . . |
$$ | . . X X . X . X X X . X . X O . O X . |
$$ | . X . X . O . O O O X . O X . . O O X |
$$ | O X . X O . O . . . . . . O X X O X . |
$$ | . O X O . . X X O O X X O . X . X . X |
$$ | O . O O . X X O O X X O O . O X X X . |
$$ | . . . O . X O O . X O . O O X X . . . |
$$ | . . X O O X X O . X O . O O O X . X . |
$$ | . . . . . . . . O O . O . X O O X . . |
$$ ---------------------------------------[/go]

After move 200.

Click Here To Show Diagram Code

[go]$$Bc Black has 3 more prisoners than White
$$ ---------------------------------------
$$ | . . O . X . . . . . . . O . . . O . . |
$$ | . O X X . X . . . . . . X X O O . O . |
$$ | . . O O X X X . X . . . . X X O O X O |
$$ | . O . O O X . X . , . . . . . X O X . |
$$ | . X O X O . X O . . X . . . X . X . X |
$$ | . X X X O O O O . O . . . . . X X X . |
$$ | . O O . O X . . O . . . O . O O O O X |
$$ | . X O . O . X X O X . . . . . O X X X |
$$ | X . X O . O X O X X . O O . O X X O X |
$$ | . X X O . X O . . , . . X . . X O O O |
$$ | . . O X . . . . . . . . X O O O . . . |
$$ | . . X X . . . X X X . . . X O . . . . |
$$ | . X . X . . . . O O X . . X . . O . . |
$$ | O X . X . . . . . . . . . O X X O . . |
$$ | . O X O . . X X O O X X . . X . X O . |
$$ | O . O O . X X O O X X O O . O X X X . |
$$ | . . . O . X O O . X O . O O X X . . . |
$$ | . . X O O X X O . X O . O O O X . X . |
$$ | . . . . . . . . O O . O . X O O X . . |
$$ ---------------------------------------[/go]

Also try evaluating the position after move 100. At that point Black should gain a few points before the end. If your count differs by a lot from that, try to figure out why. Was it because of better play by one player or other? Or was it because of inaccurate assessment by you? The answers to these questions may not be obvious, but trying to find the answers is a good exercise.

After move 100.

Click Here To Show Diagram Code

[go]$$Bc
$$ ---------------------------------------
$$ | . . . . . . . . . . . . . . . . O . . |
$$ | . . . X . . . . . . . . . . . O . O . |
$$ | . . O . . X X . X . . . . X X O O X O |
$$ | . O . O . . O X . , . . . . . X O X . |
$$ | . X O . O . . . . . . . . . X . X . X |
$$ | . . X X O . . . . . . . . . . X X X . |
$$ | . . . . . . . . . . . . . . O O O . . |
$$ | . . . . O . . . . . . . . . . . . . . |
$$ | . . X . . . . . . . . . . . . . . . . |
$$ | . . . , . . . . . , . . . . . , O . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . X X . . . X X X . . . . O . . . . |
$$ | . . O X . . . . O O X . . X . . O . . |
$$ | . . O . . . . . . . . . . O X X O . . |
$$ | . . . . . . X X O O X X . . X . X . . |
$$ | . . . O . X X O O X X O O . O X X . . |
$$ | . . . O . X O O . X O . O . . X . . . |
$$ | . . . . O X X O . X O . O . O . . . . |
$$ | . . . . . . . . O O . O . . . . . . . |
$$ ---------------------------------------[/go]

Here is the game record.

_________________
The Adkins Principle:
At some point, doesn't thinking have to go on?
— Winona Adkins

Visualize whirled peas.

Everything with love. Stay safe.

So it seems that as you play more and more and see the same moves occur frequently, you ought to develop a since of what particular moves are worth. Such as basic joseki. I feel like if I'm going to spend sente on a joseki, I should be aware of what I'm actually getting for it. Likewise, I've been told that the monkey jump is worth 8 points (though it does seem like it would be more). So when I look at making this move, I often try to assess whether there's a move bigger than 8 points on the board. Is there some sort of library that goes through different basic moves and sequences and explains their value? For instance, I feel like I should know what a basic 3x3 invasion under a star point is worth (assuming my opponent responds and cramps me into the corner).

_________________
Thinking like a go player during a game of chess is like bringing a knife to a gun-fight. Thinking like a chess player during a game of go feels like getting knifed while you're holding a gun...

A library of a few moves is available but above your level.

Endgame move values (8 points of a big monkey) are not the same as calculating a territory count (which predicts the expected score) in positional judgement, which assesses boundaries and the territories behind.

Right. I was talking about a separate practice to only aid in counting. I've read that top-players know the value of every move they make. Sounds a little over-the-top but hey, people get good at this game.

_________________
Thinking like a go player during a game of chess is like bringing a knife to a gun-fight. Thinking like a chess player during a game of go feels like getting knifed while you're holding a gun...

They don't know the value of every move they make, not even in the endgame.

With no time limits, top players in the 19th century played nearly perfect endgames. In early 20th century games, you start to see errors by top players with plays that gain three points or more. With today's short time limits, who knows?

_________________
The Adkins Principle:
At some point, doesn't thinking have to go on?
— Winona Adkins

Visualize whirled peas.

Everything with love. Stay safe.