Endgame proverbs

[RedStick]

My endgame is definately the part of my game I've put the least amount of thought into. I'm wondering if people have general rules of thumb or proverbs that they use to help them sort different endgame plays and more importantly when to play them. Does "endgame" always have to wait until the "middle game" in over?

All I have for now is: keep sente.

[kokomi]

If i'm ahead, i'll just do endgame very soon. I'm bad and endgame as well, because i often resign and don't get many practice. So possibly just play every game until the end, even if you are 100 moku behind...well, then there's another thread discussing about how annoying people are when they refuse to resign.

[Solomon]

I think relying on proverbs to help your endgame is a bad idea. There is simply no substitute to solving endgame tesuji problems and applying the tactics you learn from those problems into your games.

[Kirby]

It is best to calculate the endgame exactly, if you can do it. However, there are some heuristics that can help you to make some good moves.

I believe it was Herman that initially rewrote the SL page on basic endgame theory. Regardless of who wrote it, it currently provides a pretty clear explanation of the basic ideas of endgame. It's right here: http://senseis.xmp.net/?BasicEndgameTheory

There's more depth to this, I think, but it should get you started in the right direction.

[Chew Terr]

Something that I'm trying to work on: use your clock time. If you're in endgame and have time left, try to use it to read sente and points. All the practice in the world won't help you if you just play endgame blitz. So I'm trying to actually make use of leftover time reading sente values in endgame, because that seems to be the best thing to practice. If I lose by a small amount and have lots of time left, I probably should have spent more time on endgame analysis.

[Kirby]

Something that I'm trying to work on: use your clock time. If you're in endgame and have time left, try to use it to read sente and points. All the practice in the world won't help you if you just play endgame blitz. So I'm trying to actually make use of leftover time reading sente values in endgame, because that seems to be the best thing to practice. If I lose by a small amount and have lots of time left, I probably should have spent more time on endgame analysis.

Sente is very important in the endgame, but there are at least a couple of other tricky aspects:
1.) Is your sente really sente? Even though a move appears to be sente, you can evaluate the value of the board position if the opponent ignores the move. A sente move is only sente at certain times. It depends on the other things on the board.

2.) What is really tricky is figuring out when you should play the reverse-sente moves. Your opponent can supposedly play these moves in sente when he wants to. But when is it more useful to play them first yourself?

There is a lot of complication in trying to play things optimally, but I think that the endgame theory that was written up on SL is a great place to start. Still, though, as you've pointed out, sente moves are probably the most important in general.

An interesting thing to note is that, at a Kim Myungwan workshop that was held recently, he told us that scientists believed that pro go players used more of the right side of their brain. They seemed to suggest that creativity was really important in go strength in the middle game.

Myungwan said that endgame skills were really the primary place where "left brain skills" were necessary. He also noted, though, that the endgame was trivial compared to the other parts of the game... :-p

[Bill Spight]

Endgame proverb:

The endgame is fighting strength.

[Bill Spight]

Sente is very important in the endgame, but there are at least a couple of other tricky aspects:
1.) Is your sente really sente? Even though a move appears to be sente, you can evaluate the value of the board position if the opponent ignores the move. A sente move is only sente at certain times. It depends on the other things on the board.

2.) What is really tricky is figuring out when you should play the reverse-sente moves. Your opponent can supposedly play these moves in sente when he wants to. But when is it more useful to play them first yourself?

There is a lot of complication in trying to play things optimally, but I think that the endgame theory that was written up on SL is a great place to start. Still, though, as you've pointed out, sente moves are probably the most important in general.

Rule of thumb 1): Play the largest move, as a rule.

Rule of thumb 2): In general, sente is preferred to gote and gote is preferred to reverse sente.

[kokomi]

why gote is preferred to reverse sente?

[Bill Spight]

why gote is preferred to reverse sente?

In general, to get the last play at the level of the plays,

For instance, suppose that the last two plays are a gote that gains 1 point and a reverses sente that gains 1 point. If you take the reverse sente, you gain 1 point, but then the opponent takes the gote to gain 1 point, as well. Net result: 0. But if you take the gote, you gain 1 point, and then the opponent takes his sente, for no gain. Net result: 1 point for you. Note that in the second line of play you get the last play (the reply to the sente).

[karaklis]

I think relying on proverbs to help your endgame is a bad idea. There is simply no substitute to solving endgame tesuji problems and applying the tactics you learn from those problems into your games.

This is a valuable advice. I remember that Jonas suggested to me to go through "501 Tesuji Problems" that includes many endgame tesujis. It is on my to-do list.

then the opponent takes his sente, for no gain.

I don't understand this. If the opponent has no gain from a sente move, was it really sente then?

[Bill Spight]

then the opponent takes his sente, for no gain.

I don't understand this. If the opponent has no gain from a sente move, was it really sente then?

This SGF flie shows what I mean:

(;CA[ISO8859-1]ST[2]SZ[19]GM[1]AP[GOWrite:2.2.21]FF[4]AB[aa][bb][bc][ac][cb][db]PM[2]FG[259:]PB[ ]C[The corner is counted as 2 points for Black.]PW[ ]GN[ ]AW[ad][bd][ce][dc][ec][eb][ea][cc]
(
;C[White plays sente.]W[da]
;C[Black replies. The local score in the corner is 2 points, for no gain.]B[ca]
)
(
;C[Black plays a reverse sente. The local score is 3, for a gain of 1 point.]B[da]
)

)

[Chew Terr]

For instance, suppose that the last two plays are a gote that gains 1 point and a reverses sente that gains 1 point.

I'm sorry, I still don't understand. If black can choose a 1 point gote or a 1 point reverse sente, I do not see how gote could be better.

Case 1: If black takes the gote, he gets a point, white gets a point for the move that is his sente, then black blocks. It's 1-1 and white's turn.

Case 2: If black takes the reverse sente, and white takes the gote, it's 1-1 and black's turn.

The only difference that I see is that in case one, black agrees to let white keep sente when white plays the sente move. Would you please explain this a bit more? The only situation in which I can see case 1 be better is in rules like AGA rules, and then only if this would cause white to have to pass first.

[padic]

why gote is preferred to reverse sente?

In general, to get the last play at the level of the plays,

For instance, suppose that the last two plays are a gote that gains 1 point and a reverses sente that gains 1 point. If you take the reverse sente, you gain 1 point, but then the opponent takes the gote to gain 1 point, as well. Net result: 0. But if you take the gote, you gain 1 point, and then the opponent takes his sente, for no gain. Net result: 1 point for you. Note that in the second line of play you get the last play (the reply to the sente).

I'm trying to understand this.

Let's say we modify the situation a bit and say that there are two gote plays left and two reverse sente plays left, all equally valuable.

It seems to me that if you play gote in this situation, your opponent will play the two sente plays and then the last gote play, but if you play reverse sente, your opponent will play sente, then gote, leaving you to get the last gote play. So, we should play reverse sente to get 2/4 of the available points instead of 1/4. Is my thinking wrong somewhere here, or is this a situation where the guideline is inverted? And in that case, what's so exceptional about the above situation that the guideline is still a good guideline?

Edit: here's an illustration of how I imagine this situation, possibly clarifying where I'm confused.

(;GM[1]FF[4]AP[Drago:4.01]SZ[19]CA[UTF-8]C[Black to play. Reverse sente along the top, gote along the bottom. Ignore the center.]AB[aa][bb][bc][ac][cb][db][sa][rb][rc][sc][qb][pb][ar][br][cr][dr]
[bs][ds][rr][qr][sr][rs][pr][ps]AW[ad][bd][cd][cc][dc][eb][ec][ea][oa][oc][ob][pc][qd][qc][rd][sd]
[aq][bq][dq][cq][eq][fq][gq][hq][iq][ir][is][gr][gs][hs][fs][jq]
[lq][kq][mq][nq][oq][pq][qq][rq][sq][kr][ks][js][ns][ms][ls][jr]
[mr]
(;B[er]C[Black prefers to play gote first.];W[da];B[ca];W[pa];B[qa]
;W[or]C[Black has two points in each corner; 8 points. White has 3 points (including the eyes along the bottom).])
(;B[da]C[Black prefers to play reverse sente first.];W[pa];B[qa];W[er]
;B[or]C[Black has 9 points, white has 3 points. Black is better off by one point.]))

[topazg]

I also don't get this, and it wasn't my understanding at all. Saying this, I'm not the 5 dan, so I'm probably wrong, but Chew's analysis is exactly how I saw it:

Assume there's a 2 point gote move, a 2 point reverse sente move, and a 1 point gote move left. Player takes the 2 point gote, opponent gets 3 points from the reverse sente and the remaining gote. Player takes the 2 point reverse sente, opponent takes 2 point gote, player takes 1 point gote. In the latter example, player B is 1 point better off, but has lost the initiative. The assumption in the initiative will be worth the next biggest move, which presumably is smaller than the ones taken. It seems like reverse sente is worth half a move more than gote for the same point value plays?

As a result, I compare biggest reverse sente + second biggest gote against biggest gote (as if I take the biggest gote my opponent takes the reverse sente, any other sente plays, and then the next biggest gote, whereas if I take reverse sente, my opponent takes remaining sente plays and then the biggest gote, letting me get the second biggest gote. Of course, that means he gets the 3rd biggest gote instead of me, but then I get the 4th and so on, and via diminishing returns the net value difference seems to be that half move?)

PS As an aside, Ogawa Tomoko and James Davies give reverse sente as double the value of gote plays as a rule of thumb in The Endgame book, which doesn't seem consistent with taking gote first.