Luck and aji

Tamsin: Go is not a game of chance, but luck can be a factor, in the sense that one is lucky when unforeseen circumstances turn out to be favourable.

Now, it occurs to me that this view of "luck" in go might help one to understand aji better.

As you know, aji covers all those possibilities that do not work at the moment, but which may come to work later in the game as circumstances change. If you play these opportunities too soon, you gain little profit and encounter difficulties in creating favourable tactics later because you have made his position stronger. But if you preserve your aji, tactical openings seem to appear out of thin air as the game progresses. For example, while you are chasing a weak group, you generate threats to revive a nearby dead group of one's own. Eventually, he saves his weak group, but not without allowing you to bring your own supposedly dead group back to life. He mutters to himself something about you're being a "lucky so-and-so".

What's really happened is this: you do not know how your aji is going to bear fruit, if at all. However, if you preserve your aji and keep a careful eye on it, then you increase the chances of there being favourable unforeseen circumstances in that part of the board. In other words, when you treat aji properly, you store up "good luck".

In one of my own recent games, I played a kikashi in an early fight, against a weak group. My opponent's group lived, and I had to give up on the attack, but at least I had created aji. Somewhat later, a decisive fight broke out in another point of the board and everything rested on a ladder. "Luckily" enough, my kikashi stone was in the perfect place to break the ladder. There was no way that I could have foreseen this, but by preserving my aji instead of pursuing an attack in vain I made my own game-winning luck.

HolIgor: True, there are proverbs though. If you want to play on the left, play on the right first.

Charles Correct. The direction of play in go is very far from random, in good play. The proverb converse to stay away from thickness is head the fighting in the direction of the opponent's bad aji. This is especially true in handicap games as White.

I think one way to explain the centrality of the aji concept in strategy is that

the middle game is so complex that having bad aji can be an impossible constraint.

If one accepts that, it may come as a revelation. You might have thought that pros, with their deep reading, could 'get away' with having bad aji somewhere; and that it would be amateurs who needed to play honte to save calculations.

Actually it seems to work almost the other way round. The 'combinatorial explosion' of middlegame fighting is so serious that pros are reluctant to accept bad aji of any kind, as a rule. It is amateurs who stretch for gain at the cost of a little thinness, in the hope that just that weakness won't prove decisive in any later fight.

Tamsin: I fear that we may be moving away from my point, which was simply that one cannot foretell how aji will pan out, but that having bad aji will increase the possibility of "accidents" and that having good aji will increase one's number of "lucky finds". That said, I'm very interested in Charles's comments, anyway, about the professional attitude to aji. What, though, about players like Sakata, whose style was famously greedy, who grabbed territory at the expense of leaving "thin spots" (his own words)? Would he have been even stronger had he played honte?

Charles I was about to add a qualification, to the effect that the way pros play is more 'polarised': thickness here, thin area where I expect to have to make sabaki there. Rather than trying to hold everything together equally.

In the case of Sakata:

• exceptional players usually have exceptional styles (though there might be exceptions ...)

• he was more a shinogi player than a sabaki player.

In fact he got involved in some remarkable games, at his peak - notable particularly for having low scores (40-50) on each side. That is low territory, and probably a 'soft' approach would lose: he chose to win by making a group live, at the extreme end of possibility.

Actually I don't think this really strays from your topic, Tamsin. Under normal conditions, one cannot be sure from which side danger will come (might indeed be a ladder, ko fight elsewhere, as well as the impact of a drifting fight). The Sakata style was more like set up a specific problem to solve, to win the game. But that is abnormal, even for a 9 dan pro.

Talking about luck reminds me about my attitude towards tengen in the opening. I don't play it hastily unless I want to try it as the first move; but instead try to imagine where the real point of symmetry is - where the stone will exert equal influence on all the places where I expect it will really be useful. At the least, I try to see if opponent's sphere of influence is off-centre. And then I will often play a square or two away from tengen.

How is this related, you ask? There is a proverb (not generally applied to Go, that I'm aware of) the harder you work, the luckier you get. A stone anywhere near the centre is indeed a ladder breaker for all four corners (precise placement isn't really needed), but my hope in playing this way is twofold:

• Increase the overall probability that the stone will influence at least one situation on the sides - not as a ladder breaker necessarily, but a stone to connect to in some more general sort of semeai.
• Increase the probability of success, in the case that opponent closes all the walls on the sides and I am forced to try to live inside.

-- Karl Knechtel