benefit of doubt in removing aji

[entropi]

It is (in my opinion rightfully) adviced that if you are not sure that you can kill a group, don't attack for killing but try to gaining some profit or outside power from it.
So, if I have doubt about the life status of an opponent group, I give the benefit of doubt to outside moves, or to tenuki (of course unless I desperately need to kill that group).

Here comes my question: What about aji keshi. If you have doubts whether a move removes opponents bad aji or not, do you in general try it anyway, or leave it alone for the moment?
Or does it depend too much on the situation that no generalization can be made?

Thanks

[flOvermind]

I think this approach is the wrong way round.

You don't play a move just because it doesn't remove aji. You play a move because you gain something from it. When you're in doubt whether you gain something from it, you shouldn't play it.

Or, rephrased in terms of aji: When you don't see a clear gain from a move, you shouldn't really consider playing it. When you do, you can think about aji. Does it gain more profit than is lost in aji? When you're not sure about the aji, it depends on the gain. When you think you absolutely need to play the move, then play it, don't even think about the aji. When it gains only a few points, it would perhaps be better to preserve the aji, even if you're not sure whether it will be useful later.

Of course, there may be the additional issue that a move that is forcing now might not be forcing later. In that case it might be good to play it anyway, even though you don't see any immediate profit. But in these situations, you should be sure that it doesn't remove too much aji

To summarize: It depends
But in practice it helps a lot to first look at the profit you get from the move, and then ask the question again. Often the answer is clear after you realize that the move doesn't really gain that much

[entropi]

Does it gain more profit than is lost in aji?

Exactly that is the question that confuses me all the time. It happens very often that I am not sure whether to remove opponents aji by playing a forcing move that gains something small, or not.

But I guess a generalization would indeed be too difficult and it depends too much on the specific situation. I was just wondering whether there is a general principle to follow about aji-keshi. Probably there is not.

[EdLee]

Or does it depend too much on the situation that no generalization can be made?

I think so.
I often make moves that are aji keshi (e.g. bad sente moves)
or miss good sente moves that I incorrectly thought were aji keshi,
only to find out later either the hard way or from a good teacher.

[flOvermind]

As far as generalizations go, I usually try to play "small" sente moves only when I'm sure they are not aji keshi. So when in doubt, I don't play it. But I don't know myself if that's a good heuristic

The one big exception are moves that are sente now and I have a future development in mind that would make them not sente. That's along the lines of the proverb "play kikashi before living", or the analogous situation when forcing the opponent to live locally...

[Shaddy]

When I'm indecisive, I choose randomly and see if I end up regretting it later.

[John Fairbairn]

I often make moves that are aji keshi (e.g. bad sente moves)
or miss good sente moves that I incorrectly thought were aji keshi,

Focusing on the latter half of the quote, my experience is that while you may have indeed missed a sente move, you were probably right anyway in thinking they were ajikeshi. One thing that repeatedly astonishes me in pro commentaries is when they refer to aji where it never occurred to me there was anywhere. In discussing this once with T Mark, about Go Seigen games, he put it nicely when he said there is go aji and there is Go aji.

Related to that is a weakness in kyu play, even some amateur dan play, I have noticed. A player knows about aji but associates it only with stones inside enemy groups. But groups without any enemy stones in them can have bad aji, too. Not recognising this, many players think they are subtly adding aji to a position by adding some stones, whereas they are clumsily removing the original, pristine aji. That said, it has also become apparent to me that one of the great skills of a top pro is being able to add aji to an aji-free zone. This is often where Go aji comes in.

The bottom line, though, is if aji bothers you, play thickly.

[Bill Spight]

It seems to me that the question here is when to play kikashi. There is no hard and fast answer. Of the stars of the 20th century, Sakata was known for playing kikashi early, while Takagawa was known for playing it late. Since they were often matched, their games are interesting from that point of view.

In theory, if you have only one threat in a region, then the only thing that you lose by playing it early is a ko threat. However, as John points out, there may be threats that are not obvious to amateurs. And if you have more than one threat, then you should wait as long as possible to choose one. Often the winning play is one that makes two threats, both of which the opponent must answer. If you go around removing your own threats, how can you make such a play?

IMO, amateurs err on the side of aji keshi, not on the side of waiting too late to play kikashi.

[entropi]

Thanks for all the replies.

It seems that the answer to my question is blowing in the wind. Apparently it not only depends on the precise situation, but also on the players style, thus there is no clear guideline in that respect.

It will be fun to look at Sakata-Takagawa games from that perspective.

[John Fairbairn]

It seems that the answer to my question is blowing in the wind. Apparently it not only depends on the precise situation, but also on the players style, thus there is no clear guideline in that respect.

It will be fun to look at Sakata-Takagawa games from that perspective.
Thanks for all the replies.

Takagawa said the timing of forcing moves is perhaps the area in which pros differ most. A fortiori for amateurs, I suppose.

[TMark]

One thing I noticed about Sakata's games (perhaps I have looked at a lot) was that, where other pros would play some forcing moves or probes around the board and then play a sequence, Sakata had the habit of playing the first one or two moves of the sequence, breaking off to play a probe, then back to the sequence and then another forcing move/probe and back to the sequence again. As I am receording lots of games, I get used to how pros play sequences and Sakata's style of doing this became very annoying, especially as he also would play out the game to the bitter end and the game record was usually on one diagram.

No kifu, aji, kikashi or keshi has been harmed in the compiling of this post.

Best wishes.

[Bill Spight]

No kifu, aji, kikashi or keshi has been harmed in the compiling of this post.

[daniel_the_smith]

No kifu, aji, kikashi or keshi has been harmed in the compiling of this post.

I'm afraid I can't say the same about any of my games...

[EdLee]

John & Bill, thank you.