Lectures on the Latest Strategies, Chapter 1
From Shogipedia
Lecture 1: One-turn Loss Bishop Exchange
I will be discussing the latest strategies played by professional players. However, I won't be focusing on very difficult matters such as what the best move at a certain position is; what I'd like to convey to the readers are rather rough ideas, for example, why a strategy is so popular in the professional world nowadays and what kind of background theory it has behind it. If you come to understand these theories, then you'll be able to imagine what the professional players are thinking and subsequently be able to enjoy watching title matches and commentaries better than before.
1. Why lose a turn?
In Lecture 1, let's talk about the one-turn loss bishop exchange. What is a one-turn loss bishop exchange in the first place? Why should someone like a professional player even want to lose a turn? I believe that's what most of you are curious about, of course.
But if we take a look at the title matches in the latest couple of years, this strategy is often played in deciding games. Therefore, so to speak, the one-turn loss bishop exchange is a very important strategy which professional players have been putting their lives on. But it must be hard for you to think of this strategy positively as it forfeits a turn in the early opening.
So the theme of this lecture is to unveil the new world that is explored by this strategy and to discuss whether the turn loss is a real loss or not.
Now starting from the initial position, P-7f P-3d P-2f G-3b G-7h P-8d P-2e Bx8h+ Sx8h (Diag. 1) leads to the one-turn loss bishop exchange. The difference from the ordinary bishop exchange is that white's rook pawn is still waiting on 8d.
If they continue to develop their formations normally, eventually it comes to Diag. 2. If black were to attack from here, it would be P-4e, and if he were to wait, it would be K-8h. But whichever he does, at this position, white has the advantage of having lost a turn in the opening.
As an example where black starts the attack, there was a game that proceeded with P-4e Px4e P-3e P-4f after Diag. 2.
When I first saw this, I was surprised at the P-4f by white. There are many moves that are "not to be played" in bishop exchange games. There is a proverb which you might well know, "Don't push the 5th-file pawn after bishop exchange." Yes, it is inevitable that we have many restrictions on piece formation as both players each have a bishop in hand. And this move of P-4f was one of the most obvious "not to be played" moves.
With the normal bishop exchange (P-8e formation), this P-4f will be met by P-2d Px2d P-7e Px7e Px3d Sx3d P*7d (Diag. A), and it is bad for white. Once white gives black a pawn, there is always the tactic of P-7e Px7e P*7d, so white should not have played P-4f after black's P-3e. After P-3e, white should escape with S-4d and threaten to play P*3f if black plays Px3d. That was common sense for the bishop exchange as are the cases with Kimura Jouseki and Masuda Jouseki.
However, out comes the 8d pawn issue in the one-turn loss bishop exchange. After black's P*7d, white can play N-8e! This move is not so much of an active attack as it is a very good and strong counterattack when the knight is harassed. Therefore, here emerges a possibility that P-8e from the old times could have actually been a bad move!
How about waiting after Diag. 2? In other words, from Diag. 2, K-8h G5-4c R-4h K-2b. Actually, these moves of white's G5-4c and K-2b were also "not to be played" moves in the normal bishop exchange. Let me explain it more in detail. In the normal bishop exchange, after white plays N-7c, he cannot play G5-4c and K-2b. Likewise, he cannot play N-7c after he defends with G5-4c and K-2b. White has to choose either defending or counterattacking and give up the other.
In the normal bishop exchange, if white castles with G5-4c and K-2b when the knight is already on 7c, then black can play B*4a, and white's in trouble. Even if he replies with B*8c, he loses after Bx3b+ Kx3b G*8d.
However, in the one-turn loss bishop exchange, B*4a is met by B*8c (Diag. B), and white is fine. Even if black plays B*5a instead, then the knight can get away with N-8e. So if black attacks the knight's head, then white can defend there, and if black attacks the knight itself, the knight can run away. This is the mysterious phenomenon that takes place.
 |  |
As explained above, by making a turn loss, white can now play the moves that were not supposed to be played, and he or she has more options. Today, the move order before Diag. 2 has been even further improved, but I'll talk about that later.
Anyways, behind the popularity of the one-turn loss bishop exchange lies the 8d pawn issue. Once white holds back the P-8e move, he can not only activate the knight onto 7c but also castle the king with G5-4c and K-2b! The moves you wanted to make are now no longer not supposed to be played, and you can play them all! Suddenly, the variations of the strategies have multiplied.
2. Resurrecting Variations
Bishop exchange strategies can be classified roughly into climbing silver, reclining silver, and rapid advancing silver. And it is well known that these three were in a relationship of "rock, paper, and scissors;" that is, climbing silver is strong against reclining silver, and reclining silver is strong against rapid advancing silver, and rapid advancing silver is strong against climbing silver. Especially the relationship between reclining silver, and climbing silver or rapid advancing silver is important. Needless to say, there are very complicated lines in these jousekis, but roughly speaking they were in that kind of relationship. However, this relationship became unbalanced after black's new strategy of holding the rook pawn (P-2f formation), and black became able to deal with any white's strategy with only reclining silver. And it is not too much to say that this discovery was Kouji Tanigawa's achievement. These explanations might be a little too complicated, but please just know that's how it was.
However, there came this new strategy, one-turn loss bishop exchange. With the small difference of pawn on 8d or 8e, white now has many ways of dealing with black's attack. And as a result, many variations that bishop exchange originally had are now resurrecting all of a sudden, and most professional players are trying them. Now, let's see some of the examples that are worth watching.
Diag. 3 shows a game of climbing silver versus reclining silver, and white has just played S-2b backward. If white's rook pawn was on 8e, this silver move wouldn't have worked because black could have then played B*6f P-6e L*8d.
Diag. 4 is a formation where white defends against climbing silver with 4th-file rook. This strategy is also made possible by P-8d formation, since it would be uncomfortable to castle in right-hand king if it was P-8e formation.
 |  |
3. Black's Right-hand King and Double Reclining Silver
The position we see an especially interesting thing is Diag. 5. Black has just played P-1f. If white replies to that with P-1d, black will go straight for rapid advancing silver. Then it will lead to Diag. 6, but in this case, it means black could omit the move of K-6h avoiding a king-rook fork. So summing it all with the facts that black plays always faster than white, and that white has even lost one turn in the early opening, it's like they have a difference of almost 3 moves. This seems to be too many of move losses for white, and I don't feel like playing white from this position.
Then what if white doesn't reply to the edge pawn push? The position about 20 moves after that is Diag. 7. Black is using a right-hand king! This is unusual. But when using a right-hand king, occupying the 5th rank at the edge is a great advantage. Not only that the king's mobility is wide, but also there's no worry of being attacked from the edge. This is a new strategy invented by Manabu Senzaki.
I might mention that if black plays a right-hand king, white can only go for repetition. There's no way white can break the cold war unless black opens a battle from his side.
The game proceeded to be Diag. 8 later. From here, black made the best use of his king's wide space with K-2g B*6i S-3h, and won.
Black's right-hand king is usually not an option, but it's effective against one-turn loss bishop exchange, and black has a high win percentage.
 |  |  |  |
Also, as a countermeasure against black's rapid advancing silver, 4th-file rook formation has been started to be used such as in Diag. 9. They are applying to one-turn loss bishop exchange the technique that they developed in black's P-2f formation. In this case, the turn-loss is not a real loss if we think of it as a procedure required for right-hand king.
 |
In one-turn loss bishop exchange, too, reclining silver is the most popular strategy. In Diag. 2, white has already played N-7c, but in the latest jouseki, it is replaced with G5-4c, and then R-4h N-7c K-8h K-2b leads to Diag. 10. So, by playing G5-4c, white is saying, "You don't have any effective move anymore, do you? You'll be able to play just small moves like K-8h or R-4h, right?" And it's safer for white to play N-7c after that.
There are already dozens of professional games played from Diag. 10. White has a hard castle of gold yagura, he has no idle piece, his knight is all right if it's attacked. So, I think it's that professional players can make the best effort from this position playing white.
 |
4. The Mystery of 40% Winning Percentage
One day, I was playing a game in Ranking League. Looking around me for other games, I noticed that as many as 3 games were one-turn loss bishop exchange games! I couldn't concentrate on my game thinking "Hey, what are we gonna do if we proceed exactly in the same way?" You can see how popular it is from the fact that it was played in 3 games in the same day, at the same place. And another thing I'd like to emphasize is that although the opening moves are the same, they all branched off into completely different games. My game was black's climbing silver, but the others were different. One was rapid advancing silver, and the other was black's right-hand king. Countermeasures for both black and white are not yet developed, so they play it as if they're enjoying to read the moves without having to think about anything else.
By the way, if we take a look at the winning percentage of white in one-turn loss bishop exchange, actually it's only around 40%! Isn't it an awful strategy in terms of that? I've never seen any other strategy that is popular when its winning percentage is only 40%. The reason why professional players play it nevertheless, is that they want to play various new moves.
In one period, I heard many fans say "Professional games these days are boring, because they always keep playing side pawn picker." But we, the professional players, too, want to play many different strategies. It was just that we couldn't do that because jousekis are too well developed these days.
However, in one-turn loss bishop exchange, we can play different moves from before. That's why it's so popular despite the low winning percentage.
I also have to discuss why white's winning percentage is so low as 40%.
It is inevitable that white's attack tends to be slower because of P-8d formation, so there are often the cases where black's forcible attack actually works. Thus, although white can have more option in the opening and middlegame because of the turn loss, it is pretty often that the turn loss actually makes himself suffer at the very end. If we take a look at the games of one-turn loss bishop exchange, the player who win as white are always the same players.
For example, look at Diag. 11, a game between Gouda and Habu. White has just played B*2c after black moved the bishop from 7c to 5a and promoted. White won this game thanks to this good move, but it is not so easy to find this kind of bishop drop. If he couldn't play B*2c, he would have lost. So one can't win as white unless he has a good ability of finding the critical move at first glance.
It is also the feature of this strategy that the things to be discussed in the post-game discussion is so many, and there are many different lines. Speaking of that B*2c, there was a pawn sacrifice of P-2d Px2d before that. The discussion will never end if they start talking about like, "what if, then, there wasn't that pawn sacrifice?"
5. Punishing the Turn Loss
Let me show you an example of a game in Meijin title match, in which white couldn't win. Diag. 12 shows the 1st game of Meijin title match in 2005, played between Habu and Moriuchi. White has just played L*7g. Responding it with Nx7g is met by Px7g+ Sx7g Rx4i+ and black will lose. This is how the endgame of double static rook game is like. S*8i from white will be a powerful attack in the next move, and even if black plays G*7i responding to Rx4i+, white is better with N*7f.
The move Habu made was G-6g! while there's even a proverb of "Force your opponent's gold to move diagonally forward". He moved the gold diagonally forward voluntarily, moving away from his king, and leaving the lance's attack there. Normally, this kind of move is not understandable.
However, this game is a one-turn loss bishop exchange! And moreover, Moriuchi has made a unique strategy of even putting the rook pawn still on 8c. So the situation is very different from usual cases.
If white were to approach the king after black's G-6g, it would be Rx6g+ Sx6g Rx4i+ Gx4i S*7i K-9g S*8h K-8f (Diag. C), and here counts the turn loss in the opening. If only a move of P-8d was made before, it would have been a win for white.
After all, Moriuchi pulled his rook (R-6e), because there was no way he can keep approaching. Can you imagine how Moriuchi must have felt when playing this move? He used 18 minutes for this move, so it was obvious that he had to change his plan after G-6g.
 |  |
So, this is the difficulty of playing one-turn loss bishop exchange. Incidentally, Moriuchi actually played P-8d later in this game. If I were him, I couldn't have played it because it would've been too mortifying. But he quietly did it. A lot of professional players say they were struck with awe seeing it.
This Meijin title match ended up both the 1st and the 2nd game being one-turn loss bishop exchange. Although white got better in the middle game in both games, the player who won was both black. This result proves the difficulty of one-turn loss bishop exchange; That's what I was thinking at the time.
6. Yasumitsu Satou as a modern Kouzou Masuda
However, after watching the '15 consecutive title match games' between Habu and Satou in 2005, my thought was changed. The last game of Kisei title match and the last game of Ou-i title match, they were both one-turn loss bishop exchange, and in both games white won. As a result, Satou defended the Kisei title, and Habu defended the Ou-i title. Although the overall situation of "40% win for white" hasn't been changed, wouldn't there be possibilities that white can actually do better?
The first game of the Kisei title match I mentioned went through the most popular position of Diag. 10. I think there was an implicit agreement between them that they wanted to solve this position together, putting win or lose aside.
We've seen an example of black's right-hand king earlier, but when a different player plays it, the strategy becomes different, too. It is not too much to say that all shogi fans must check out Satou's right-hand king!
Diag. 13 shows the 2nd game of Kisei title match. The moves from here (from black) were awesome. S-5f! P-5d P-6f G5-4c G-4g! After that he attacked with P-4e.
Diag. 14 is the 5th game of Kisei title match. Satou's eager to attack never stops, and he played even more aggressively (as white), S-6d Sx9a= P-7e Px7e N-8e S-7f S*6f L*6h Sd-7e(Diag. 15).
What a strategy! After seeing this, we can't any longer say "Play a right-hand king as what it should be". Don't you feel Satou's mind saying, "Studying the openings are only for avoiding stupid losses. More important in a game is just the strength." And I'd like to emphasize that this was the last game of the title match! The most important game of "Hold the title if a win, lose the title if a loss". However, I would say to the readers, "Amateur players shouldn't copy Satou's style.", while in the old days they said, "shouldn't copy Masuda's style."
 |  |  |
7. Turn Loss and Strategies
After all, in one-turn loss bishop exchange, no matter what kind of formation it goes into, the moves you made often fail to be the best ones later on. And you should decide your move after seeing your opponent's move. That's the concept behind one-turn loss bishop exchange, which makes it a popular strategy.
Something I recall here, is what Habu said like a prophecy at a symposium. He said, "Moves in shogi are mostly bad moves. Most of the moves had better not be played."
Many shogi software programmers participated in the symposium, and they were all laughing at Habu's word. (lol)
But anyway, now we're in a situation where white's P-8e could have been a bad move in the first place, and have been a move that should not be played. Thinking in this way, I can't help thinking over "What does 'the best move' mean in shogi?"
After the symposium, I also spoke to Habu about other things such as chess. And I even thought over after that and realized; Isn't shogi the only chess-like game that has as many as three pieces - knight, lance, and pawn - that can't move backward again?
Especially, a shogi knight can't move in 8 directions like a chess knight, so it is often not a good idea to advance it too early. However, a knight is in a way the strongest attacking piece. Let's suppose a 1-on-1 confrontation where a king is checked by a knight. The knight won't be captured by the king, nor blocked by any other piece. Even if the knight was captured by another piece, it's not a good piece to drop for defense. Moreover, since a knight moves very fast and has a pretty good piece value, it's formation is very important from the early opening. Recent new strategies that became suddenly popular, such as Fujii system and R-8e strategy, tend to use knights very effectively. Looking from a long perspective, perhaps modern shogi is on the track of rediscovering the value of a knight.
In western chess, there is a strategy term of "zugzwang". It's meaning is like "forcing the opponent to play an undesirable move". In other words, it's an advanced technique to create a better situation by making a move that is almost equal to a pass. Not only western chess but also other games such as contract bridge (worldwide popular card game) have similar ideas.
In shogi, it can be referred to as an "advanced waiting move". But the turn loss in one-turn loss bishop exchange is done in the very early opening, so it doesn't seem like a "waiting move". More deeply we analyze this strategy of one-turn loss bishop exchange, more I feel that we don't have enough vocabulary to talk about modern shogi.
8. It Evolved Expandingly
Mentioned above were the basic lectures. But those were our understanding at the time of around November, 2005. The evolution of jousekis are becoming surprisingly faster these days, and that mostly is the case with one-turn loss bishop exchange. If we overwrote them with the latest theory, we wouldn't be able to see the process of evolution. So in this book, I adopt the style of adding the latest trend to the basic lectures. (I even left my embarrassing mistake as it was)
First of all, let's talk about reclining silver. The important theme positions of Diag. 2 and Diag. 10 were played in big matches like Meijin title match. Let's take a look at what kind of game they go into.
 |
|  |
Black attacks on the 4th file concentrating his attackers. If white responds to P-4e with Px4e, there's a violent attack of Sx4e Sx4e Nx4e S-4d Nx5c+ Sx5c B*7a from black, so white makes a counterattack on the king's temple with N-8e. Diag. 16 appeared several times in title matches, which is the standard sequence now. Isn't the difference so large between one-turn loss and identical formation, although their both bishop exchanged double reclining silver games? After this position, there continues a very subtle and difficult battle, with both players trying to attack the opponent's right knight, and with black's horse and white's rook finding their way to occupy better positions. These lines are still under study.
The game that were elected as the 1st of Year's Best Games (in magazine "Shogi Sekai", March 2007) by professional players was also a game of the popular reclining silver, which was played between Habu and Tanigawa as a playoff in A-class. From Diag. 2, it proceeded P-4e Px4e P-3e S-4d P-2d Px2d Rx2d P*2c R-2h N-8e S-8f B*7c(Diag. 17). The battle starts similarly to that in identical formation in reclining silver, but once again the white's counterattack of N-8e and B*7c is fast, and it becomes different from identical formation. Diag. 17 is also played in several professional games, and is an important theme position.
In this game, 'mate or no mate' battle in the very endgame was extraordinary, and became one of the most historic professional games. From Diag. 18, Habu went for mating white's king, but it was barely no mate. As a result, Tanigawa became the challenger to Meijin, which led to the heat battle of Diag. 16 later. In Diag. 18, it was discovered later that if black had played G-5h, it would have been a threatmate-removing threatmate, and black would have won. But it doesn't seem like a move that anyone can find in an extreme situation of having only a few minutes of thinking time left.
 |  |
In climbing silver games, Satou's right-hand king (see Diag. 4) became a standard. And we saw a very interesting tactic in it. From Diag. 19, Manabu Senzaki revealed his deliberate tactic of P-2d Px2d P*2b! If white takes it there's B*8b. Therefore, now it is said that white might have to play K-7b before playing N-7c. This is a pretty uncommon tactic indeed. New strategy gives birth to new tactics.
 |
And the things that most made me shocked was the great change of interpretation to the lines after P-1f in Diag. 5.
"If white responds to P-1f with P-1d, it would be bad for him if black goes for rapid advancing silver." That's how one must think seeing Diag. 6. But when we have tried it, we realized actually that's not necessarily the case. When I was pointed out, "White can still play pretty well after he just allows black to make the silver trade.", I even thought, "That's not true!"
But after I studied on it, I realized it's indeed still not so easy for black for several reasons such as, black's king is still unguarded being a sitting king, counterattacks of B*5e attacking the rook's temple is fierce. It's unbelievable that white can still play well after losing almost 3 turns! This line appeared in Meijin title match as well, and white actually won after a worth seeing battle.
However, today, when black goes for right-hand king with vanguard pawn on the 1st file, black doesn't even make the move trade of G-7h P-8d, as you can see in Diag. 20. It's idea is, let's omit even G-7h as white won't attack on the 8th file anyway. In this case, white is less likely to respond to P-1f with P-1d.
 |
And talking about right-hand king today, there are even some cases where white castles in anaguma. A modern proverb is "Castle in anaguma to lessen your disadvantage in strategic phase." We see one of them in a title match (Diag. 21).
 |
Recently, we're also starting to see an opening shown in Diag. 22. From the initial position, P-7f P-3d P-2f P-9d P-2e P-9e. If this was played 10 years ago, we would have said, "This is far from being a professional game!", but now they even play it in title matches with no hesitation. What white is trying to do here is, no matter what black's next move would be, such as G-7h, S-4h, or K-6h, he is going to go for 4th-file rook with R-4b, or alternatively, Bx8h, S-2b, S-3c, and then R-4b. After he does that, his castle will be mino castle with a vanguard pawn on the 9th file, and it's very reliable (even being good at dealing with static rook's anaguma). But needless to say, black wouldn't want to let that happen. He'll try to make a rapid attack trying to punish white's turn losses, and it will go into a full scale battle.
This white's strategy is something like a fusion of one-turn loss bishop exchange and a free-style "turn loss ranging rook", and it's very complicated. So, one-turn loss bishop exchange has not only made the potential variations of bishop exchange resurrect, but also made the internal energy of shogi gush. It's an outrageous era in shogi history!
 |
Thanks to the outcome of the Meijin title match, the jouseki of rapid advancing silver with K-6h formation has been improved to a great extent. Diag. 23 is the very important position where they have just played (from black) P-3e Px3e Sx3e P-8e. White's P-8e is the move that has been thought as a good countermeasure. If black then plays S-7g, there's a joining pawn tactic of P-8f Px8f P*8e. However, they have found out that black can attack with P-2d immediately, not responding to P-8e.
Another white's frequently played countermeasure against rapid advancing silver is a formation in which he plays P-6e and B*6d. But black would be able to make it a favorable position by bringing it to a slow game.
So, as the 3rd countermeasure, white came up with this idea. "There's no point in holding the rook pawn against rapid advancing silver. Why not just play P-8e before getting attacked?" That's Diag. 24. Although doing this means he has made a pure turn loss, white can try it relying solely on the principle that reclining silver is strong against rapid advancing silver. Surprisingly defiant idea as it is, there once was a week in which as many as 3 games with that opening were played.
The reason the rapid advancing silver had lost its popularity in the ordinary bishop exchange, was because of its weakness against reclining silver. But it's now drawing attentions again with the difference of white's losing one turn, which I think is a very interesting phenomenon.
Fujii system has changed our idea on king's safety. R-8e strategy has freed the selection of rook's position. And, one-turn loss bishop exchange has destroyed the common sense that "turn-loss is a downside". This was the greatest achievement by one-turn loss bishop exchange. But nobody can tell how shogi is going to evolve as the result of these changes.
Game Data
- Diag. 3: Tamura vs Takano (Feb. 2004, C2-class, Ranking League) Kifu for Java | KiFLA
- Diag. 4: Habu vs Satou (Jul. 2005, the 5th game of Kisei title match) Kifu for Java | KiFLA | with comments
- Diag. 7, 8: Senzaki vs Shima (Sep. 2004, B1-class, Ranking League) Kifu for Java | KiFLA
- Diag. 9: Tanigawa vs Senzaki (Jun. 2005, Ou-i title league) Kifu for Java | KiFLA
- Diag. 11: Gouda vs Habu (Dec. 2004, Oushou title league) Kifu for Java | KiFLA
- Diag. 12: Habu vs Moriuchi (Apr. 2005, the 1st game of Meijin title match) Kifu for Java | KiFLA | with comments
- Diag. 13: Satou vs Habu (Jun. 2005, the 2nd game of Kisei title match) Kifu for Java | KiFLA | with comments
- Diag. 14, 15: Habu vs Satou (Jul. 2005, the 5th game of Kisei title match) Kifu for Java | KiFLA | with comments
- Diag. 16: Moriuchi vs Tanigawa (Jun. 2006, the 6th game of Meijin title match) Kifu for Java | KiFLA | with comments
- Diag. 17, 18: Habu vs Tanigawa (Mar. 2006, playoff in A-class, Ranking League) Kifu for Java | KiFLA | Hidetchi's commentary video
- Diag. 19: Senzaki vs Miura (Jul. 2006, NHK-cup tournament) Kifu for Java | KiFLA
- Diag. 21: Satou vs Habu (Aug. 2006, the 5th game of Ou-i title match) Kifu for Java | KiFLA | with comments
- Diag. 22: Moriuchi vs Satou (Feb. 2007, the 2nd game of Kio title match) Kifu for Java | KiFLA | with comments
- Diag. 23: Fukaura vs Watanabe (Jul. 2006, Ouza title league) Kifu for Java | KiFLA
