Tewari is an important tool for analysing local situations, allegedly invented by Hon’inbō Dōsaku. It is perhaps the field I am best at in Go, though I’m nowhere close to the world’s best tewari analyst whom I’m going to introduce further below.
Tewari 手割り, or shǒu gē 手割 in Chinese, is spelt with the characters of “hand/move” and “cut apart”. That is also what it implies:
The senseis page on tewari is incomprehensive, so here are the various tewari techniques:
- Simplification through removal of logical exchanges.
- Comparison with familiar shapes.
- Rearranging the order of moves. (Senseis calls this transposition.)
Also note that it’s useful to keep in mind the difference in number of moves both players spent. (Achieving a locally balanced result with a move count of 1 more than your opponent means you got bamboozled.)
Often the result of the analysis is contrary to the impression at first glance.
The man who has mastered tewari and is widely considered the world’s best tewarist is Yu Bin 9p, the trainer of the Chinese national team. He allegedly developed a super secret technique to take apart positions in an instant. This means that he sees faster than anyone else who is favourable in position so-and-so. When he plays, he can discard certain variations immediately and use his energy to search for new ones, like in this game:
Yu Bin (White) shoulder-hits the black 3-3, but instead of finishing the joseki with a move protecting the push-and-cut, White plays tenuki to go for a big move 7.
And thus Black pushes at 8. White plays another tenuki 9 and lets Black obliviate White’s shape completely with 12. What a horrible result for White! …that’s what tewari novices would think.
Appearances are deceiving. By dissecting the shape, we find out that Black is the one that got swindled. Remember the move count? At move 6, Black had +1 stone in the upper right, and by move 12, it became +3. So we have to regard the upper right as a formation where Black has 3 more stones. (Duh.)
Taking away an equal amount of stones away for both colours, we arrive to this familiar shape:
This is the position we should regard the corner as. Black adding a move to the shimari is undoubtedly a ridiculously slow move at the current stage of the game.
If this observation does not satisfy you conclusively, we will put back the exchanges in a logical order. (To arrive to the above diagramme, I took them away in a reverse process of what follows.)
White plays the peep at 1 and attachment at 31. These exchanges are more profitable than not. They are outside vs. inside exchanges which is generally regarded as favourable for the outside stones.
And then White even made a dent in Black’s territory, potentially decreasing it by the 2 marked points, making it even better for White.
Following from above analysis, I dare say that by move 13 (move 18 in the game), Black is already behind by close to one move. (Perhaps Black shouldn’t have played at 12. Then both players would have disregarded that area for a while.)
You can see the full game on Gokifu.
Tewari is a powerful tool and I’m hoping to master it one day.
NEXT UP: An entire joseki move destroyed by tewari.
1Do not argue that White would never play at 3 and if White did, Black wouldn’t answer at 4. That’s not how this works:P The exchange is already there and it’s a matter of who gains more. Sometimes (not always) the answer is obvious.
P.S.: Note that I chose the clearest way (for me) to rearrange the exchanges, by inserting one more step that dents away 2 points in Black’s territory, which is easily comparable. In contrast, I would find the following tewari illogical and this would legitimise the concerns voiced in above footnote:
“To White’s attachment 1, Black made a bamboo at 2 (?). And then White pushed at 3 (?). So 1-2 is a good exchange for White and 3-4 slightly bad. (???)” Though this works as well, nobody can quantify how much Black lost by not blocking at 3 immediately and how much White lost by self-bumping a liberty. With the extra step, we could say that we actually saved brainpower because these questions became irrelevant. Those two factors combined equals (potentially) 2 points.