jlt wrote:Bill Spight wrote:But have the pro give a reverse komi of 150, which is about right, and see what happens.
In that case, the pro will play a very aggressive opening because he knows he has to catch up 150 points. If the pro plays a "normal" opening, he will gain fewer points at the opening, maybe 15? That's twice komi, so that's like one stone of strength difference. Is it negligible or not? It depends how far you are from your maximum level.
I think talking about "beating up on" sent the wrong message. What I had in mind was mainly a question of judgement. Some years ago, while the top bots played at amateur dan level, Yoda Norimoto gave 4 stones to one of them. Despite the 4 stone handicap, Yoda built up a large moyo. The bot, as I recall, managed to reduce the moyo enough to win the game. But suppose that Yoda had pursued a large moyo strategy in an even game. Without the aid of the handicap stones how would the bot have fared? There are many possible ways to deal with it, and the bot would have been on unfamiliar territory. If you look at historical pro games going back to at least Dosaku, you see pros dealing with large, thin moyos in various ways, not always successfully. Even today, bots find many pro errors in dealing with moyos, thin or not. How much more difficult would they be for SDKs? And it is not just SDKs who may have such difficulty. Once, in a tournament, as a 4 dan playing even against a 2 dan, I won by some 70 pts.!
I wasn't trying to bamboozle my opponent, I was just playing go. But in the opening I built up a large central moyo. At some point, probably around move 50, he played an incursion, I sacrificed a few stones to build a solid wall, and that was that. Now maybe he had a better play, like invading somewhere around tengen. Then I might have won by only 40 pts.
The thing is, he played a bad opening, misjudging my outside strength.
Here is a question that most experienced players have faced. You have a two space extension on the side, with opposing stones on either side, and you play a one space jump into the center. A bit later your opponent peeps at the jump from the inside. Do you connect the jump, or do you protect on the third line? A mistake could cost more than a couple of points. What do you do?
A mistake that we see at all levels of play is misjudging sente and gote. An extreme form is what I call
followitis, responding to almost all of the opponent's plays. Followitis is a quick way to lose many points. OTOH, if the opponent is much stronger, not responding to their plays may be a quick way to disaster. Anyway, in the opening if a player plays gote, gaining 5 pts. instead of taking sente and playing elsewhere to gain 15 pts. that is a loss of 10 pts. in only one play. That doesn't happen that often, but losses of around 5 pts. in one play instead of 10 pts. are not uncommon at the SDK level.
Average players do not just lack the knowledge of good plays, the actually prefer bad plays in many instances. To reach dan level most people have to unlearn their bad plays. To paraphrase the great bridge writer, Victor Mollo, faced with a 50-50 choice, average players will make the wrong choice 75% of the time. That's a joke, OC, but it illustrates Mollo's point that average players have poor judgement. In Mollo's opinion, which I share, amateurs would benefit greatly from improving their judgement. In the case of go, you can't have good judgement if your openings are bad. Sakata tells about his teacher, Masubuchi, coming by their home -- Sakata's family was poor, so he lived at home -- and setting up a whole board problem and asking him where he would play. He would make the play and she would pat him on the head. At the age of 6 Sakata had developed good whole board judgement. Sakata was known for his reading ability, but he pointed out that reading is not just a question of the calculation of variations but of the application of judgement, both in judging the result of reading, but also in the choice of candidate plays.
Given their poor judgement, 2-3 kyus will typically lose 20+ pts. to par by move 30, IMO, when left to their own devices. How do you get them to give up more points? One way is to give them more choices to make. The more choices they make, the more mistakes they are likely to make. Yoda's large moyo strategy is one way to give them more choices. (Dosaku also used that strategy in handicap games.) Aggressive play is another.
, 
,..., 
) then his level won't decrease by more than 2 stones, so that a high SDK trying to improve his opening can only expect to gain a fraction of a stone. If you consider that worthless, then almost every high SDK's opening is "good".
