Playing for as Narrow a Victory as Possible

(Moved from HardHeartedness)

Hans: ... In my opinion not playing the best move that you see (which may be a hard blow for your opponent) also shows a sort of disrespect for the game. Who wants a victory because the opponent did not play his best go?

SnotNose: ...unless you're playing for as narrow a victory as possible. Then the ideal is a 0.5 point win. That's even harder than winning. Sometimes I play for balance. That is, if the players switched sides (Black plays White and vice versa), I should be equally happy, right up until the end of the endgame. This is very hard. Does this disrespect the game?

Hans: I (and almost all amateurs) am not strong enough to play this way. I guess only a professional player can estimate the score accurately enough to play a balanced game from the beginning. So playing as you proposed will only be possible in a teaching game. And in this case this does not disrespect the game. But in an even game you always should try to play the best move. If your opponent is of the same strength the game automatically will be a close one.

This does not mean that you always should play the biggest move. If you have a clear lead and see some chance to gain some extra points but with a small risk of losing points, then it is better not to go for the highest win but the sure win. The "best move" does not only depend on the local situation but also on the actual score (due to the imperfect human understanding of the game).

SnotNose: Even if one is not strong enough to play this way and pull it off exactly (and few are, as Hans points out), one can do it within the range of error implied by one's ability. That is, if you count and see that you'd be 20 points ahead (or would be with a move at A, say), then you can look for a less profitable move that allows your opponent to bring the game closer to even. If you feel your counting ability is accurate to within +/- X then you'd be successful if you can keep the game to within +/- X by the end.

(Sebastian, under the assumption that this was countering Hans' statement: "you have to give up feelings like pity":) This certainly would make for a more interesting game. It would be a white lie, though. Some people (such as me as a child) would really be offended if they found out that you weren't really playing the way you "should". SnotNose, what would you say if your opponent asked you after the game: "At this move, why did you not play here?" Would you stay in character and say (or imply) "I didn't see this!" or would you fess up and explain your philosophy?

SnotNose: Firstly, I do not always play this way (deliberately ignoring a clearly supperior move to keep the game balanced). In fact, it is more the exception than the rule. When I do play this way, it is typically with opponents who already know that sometimes I do this and I know that they are comfortable with it. I might or might not tell them that I'm doing this for the particular game we're playing. So, it is up to them to figure out whether I'm missing the big move or just trying to keep the game balanced. Either way, I know they're not going to get upset.

The only major exception to the above is in teaching games. In this I might go so far as to point out that A is the big move and, were I to take it, the game becomes very easy for me. But, suppose I pretend I don't see it and play B. Then the student may continue having learned the big move and still with the excitement that the game is somewhat even still.

Having said all this, I do feel it is reasonable to play for a win and disregard the margin. That is, suppose I see that I can win by 5 points following sequence A and I can also see that I can win by 25 points following B. I don't see a problem taking path A over B.

I'm glossing over the major issue of uncertainty. That is, typically one is less certain of the accuracy of a prediction of a win by 5 than a win by 25. That is, the latter has a bigger margin for error and, so, is more likely to result in a win (maybe only by 15 if the error was 10). But, the former could end up a loss because one's counting and "look ahead" was off by as little as 6 points.

But, this is a rather simplistic view of uncertainty. I find it not uncommon to go for the bigger win, only to find that what I've really done is inspired my opponent to look more urgently for a comeback sequence (some call these "meltdown moves"). He sees he cannot win without taking a big risk, makes some very aggressive plays, starts a fight, and throws the game into chaos. Sometimes I lose these because I cannot cope with the chaos and in subsequent analysis my opponent and I agree that the other, more balanced, path would have more surely led to victory.

So, there can be real value in the "slow and steady" balance approach. That is, looking for a move that maintains the balance can be a valuable concept even when you do want to win by a large margin. I guess the lesson is that one has to combine one's view of the value of the move with a measure of the degree of uncertainty. I'm more likely to go for a balance move when I am uncertain.

Overall, my philosophy is to preserve the opportunity to win as long as possible. If I try to press for a win too early, I lose balance and fall over. If I patiently maintain the opportunity to win while I wait for that "golden opportunity" then I find my winning percentage increases. Judging that the opportunity is really a golden one and not just a mirage is the hard part. With each move I ask, "Did he make an error I can exploit?" Usually the answer "no"--I cannot see a way to gain advantage. So, I look for a balance move. If the anwer is "yes" I then have to figure out how much an advantage I should expect to get. It would be folly to think that every small error by the opponent (e.g., deviation from joseki) can be converted into a win. Some errors are only worth a point or two. Some deviations from joseki are warranted and are actually more optimal.

Well, I've gone on long enough...I hope I've clarified things. I'm happy to add more here or elsewhere another time.

(Sebastian:) Your description of playing safely is correct, but I think it misses the point. Of course, in any competitive game, someone who's way ahead should play safely in his/her own best interest. But the way I understood Hans' question is: What if you play a move that's not exactly in your best interest? Imagine, you play against 10 year old SebastianBoy. He's very eager to win, and you like him. When it becomes clear he has a snowball's chance in hell, you start playing for as narrow a victory as possible out of pity. "As possible" meaning: More so than the safest play.^[0311040] This inevitably increases the chance for him to win. Now what do you tell him after the game, if he questions your charitable move?

^[0311040] There is no sharp border between possible and impossible. "As much as possible" assumes a sufficiently clearly defined range of possible good plays. Since the value of plays is not a rectangle function (of a variable such as safety or size), but more something like a bell curve it doesn't have a clear edge. Therefore all we can say is: If there is anything we can call a "border" then it is not at the optimum.

SnotNose: I wasn't talking about a safety play when ahead. I was talking about a "balanced" play whether ahead, behind, or even. By a "balanced" play, I mean one that maintains the the chances of victory at the current level. My argument was that it may be the best play when one considers uncertainty. Perhaps we all (or many) have this concept in mind with the first several plays of the game (despite what DirectionOfPlay says, we don't expect to win or even have a lead with the second move). To maintain this point of view throughout the game is a different thing, and quite interesting.

I don't play any move "out of pity." I play for my own enjoyment. I enjoy a close game. Against players who are already comfortable with this idea I might play a known sub-optimal move for the purposes of keeping the game close, for my own enjoyment. Opponents who don't like this have the chance to say so in advance. I won't play this way against them. I don't enjoy pissing off my opponents. I'd rather not play someone than piss them off. I'm not that hard-hearted :)

Thus, I would not be in the situation you propose: having to explain to a kid that I played a bad move because I took pity on him. For the sake of argument, suppose I found myself in this position. I might say, "I am not fit to play with you. Let's try to find you a different opponent next time." And that would be honest. I don't find playing out of pity a fitting state of mind so I would try to avoid it in the future.

(Sebastian:) Well, then I misunderstood your first reply. I thought you were countering Hans' statement: "you have to give up feelings like pity". If this was not what you meant, then this discussion, beginning with your reply, had nothing to do with hard-heartedness, and we should move it to a different page. (done now)

SnotNose: I was responding to the statement that "not playing the best move that you see (which may be a hard blow for your opponent) also shows a sort of disrespect for the game" and re-emphasizing my earlier point that there is more than one answer to "what is Go?" even when one restricts oneself to a single rule-set. The determination of the degree to which Go is "hard-hearted" might need to take account of different answers to this question.

That is, the rules do not define what is meant by the "best" move. Yet, many would say something like "the best move is the one that most likely leads to victory." But, this is just one point of view. Depending on how one extracts joy from a game, the "best" move may be something else, like "the one that maintains equal chances of victory." For lots of folks it seems like the "best" move is "the one that most likely leads to a big, messy fight, regardless of whether it increases my chances to win." :) :)

There is a sentiment among those who hike the Appalachian Trail (a 2000+ mile trail that runs from Georgia to Maine in the US) that each should "hike his/her own hike." This means that each is entitled to set his/her own rules for "completion" (i.e., decide if it is okay to leave the trail, to sleep in a bed, to return home, etc.). Some people are very adamant about how the trail should be hiked in order to consider it "completed." But if one is honest, one cannot avoid admitting that this is arbitrary. Thus, fighting over it makes no sense. Each should just "hike his/her own hike" and leave others alone. Swapping philosophies is okay, but insisting that someone else's philosophy is not valid, or disrespects the trail, or some such is silly.

I think the same goes for Go. Each should "play his/her own game." I have a strong sense of what my game should be and I don't understand the game some others play. But, I don't think it makes sense to claim that one style or another "disrespects the game." It might be inconsistent with another style or culture, but "the game" is too broad (methinks).

(Sebastian:) SnotNose, I have a related question for you: My sensei was 1d, and I was ~17k, and I always lost, since we never played more than 9 handicap stones. So I decided for myself to be content if I lost with less than 50 points. He, however, said it was meaningless and refused to give me 50 komi or to play for as big a victory as possible, because that would keep him from playing honte. Would you agree or do you think it is possible to play honorably for a narrow victory against 50 komi?

SnotNose: Well, either the komi is reasonable or it is not. Were you losing by 50 points on average? If so, then the 1d need not change his style at all and he should still win half the time if he gives 50 points in komi. But if the margin was generally smaller then taking that komi would force the 1d to change his style, perhaps playing more unreasonably to trick you. That is likely not the kind of game he would want (nor would I), as it is equivalent to an incorrect handicap. It would be the same as him being way behind in a game where not taking risks would surely lead to defeat.

Anyway, my reaction to the proposal would have been as follows. As a ~1d, I wouldn't care what kind of game my ~17k friend wished to play so long as it had pedagogical value. To me, any non-tournament game with 9 stones should conclude with a lesson for the weaker player. Fundamentally, it should be a game that helps the weaker player learn. I would want to play moves that I could point to later and say "this move is a good move by White, here is why" and I would not want to have to say over and over "this move is an overplay...this move is bad...you could have punished this move" and so on. There will be some of those moves, probably, but the fewer the better. Better to illustrate Black's errors and have him learn from them than to teach trick plays.

Because it is fundamentally a teaching game, the score is irrelevant so komi is irrelevant. Perhaps this is what the 1d meant by "meaningless" ^[1].

As for what can be done "honorably," I do not think the question has any meaning. Define honor as you like. (Again, I fear I have misunderstood.)

[1] (First time I ever did a footnote--maybe not getting it right--SnotNose.) By the way, for me winning is not near the top of my list of priorities for a game of go, except in tournaments. I play to learn about the game, to be faced with puzzles, to be delighted by good moves (by either side), and to practice the discipline required to make good move after good move. I feel this way because winning does not, in and of itself, imply increased knowledge. Winning may follow from increased knowledge, but it is not an ideal measure. One may need to lose to gain deeper understanding.

So, if winning is very important to my opponent, I am happy to guarantee it for him by giving a big komi, provided that doesn't change his style (though it is hard to see how it could not if the komi has meaning to him).

Or, more commonly, I am typically happy to let my opponent believe the excuses he makes for losing after a game. That makes him feel better and does me no harm. Usually those who make excuses are the ones who value winning. I know this because the excuse begins with "I would have won if not for..." or "I was ahead until..." If one values other things, one might say, "Very good game. I'm curious about this play I made..." or "I'd like to find where the game turned in your favor."