SIX PIECE HANDICAP (white removes rook, bishop, both lances, and both knights) is an enormous handicap, roughly comparable to Queen odds in chess. Although officially the Shogi Renmei considers it only one step above four piece handicap, in reality the difference is immense.
According to the late Nada Rensho 9 Dan, once challenger for the Meijin title (grand champion) and the leading authority on high handicap shogi, it is nearly twice as large a handicap as four piece. This is only a slight exaggeration in my opinion, yet in actual practice strong amateur kyu players often lose to pros at six piece.
There are two reasons for this. First of all, many amateurs don't know or don't use the "Josekis", the recommended openings for the handicap. They play a normal game, in which white's weaknesses on the edge play little role. The game starts to resemble a two piece game (rook and bishop handicap only), and since the players are not of the strength (about 2 dan) needed to win with that handicap, they end up losing.
The second reason is that most of the published lines for six handicap are really five-piece handicap openings, in that they attack only one edge or the other, so players who use these lines are still not getting the full value of the handicap.
Let's start by looking at some of the lines often recommended, and seeing how white can make the game difficult. One often recommended line is to attack white's left flank by P7f, P-2f-2e (to prevent ... P2d and ... G2c, defending the edge), P-1f-1e, L1g, and R1h. However, there are three objections to this plan. First of all, it takes several moves, giving white enough time to achieve a decent development of his own forces.
Moreover, there is a defense (not mentioned in the books) which I like that complicates things a bit and buys more time: 1... S2b 2 P7f S8b 3 P2f G3b 4 P2e P1d! 5 P1f S1c! 6 L1g K4b (defending 3c to free up the gold) 7 R1h G2b!. Now if 8 P1e? Px1e 9 Lx1e? P*1d 10 Lx1d Sx1d 11 Rx1d L*1c and the rook is trapped.
Instead black can break in by climbing his silver to 2f (6 S3h, 7 S2g, 8 S2f) and then playing P1e, but this is slow and usually means exchanging silvers, giving white some ammunition for counterplay.
Yet another (non-book) defense for white runs: 2... G6b 3 P2f P5d 4 P2e G5c 5 P1f P4d 6 P1e G4c 7 L1g G3b 8 R1h G3d 9 P1d Px1d 10 Lx1d G2e 11 L1b+ P*1f. Black wins silver for lance, but he cannot easily promote his rook.
Another popular plan is to attack white's right edge from the start. This often requires the sacrifice of bishop for gold and pawn. It is the quickest way to break in, but the sacrifice of material means that black must play well afterwards or face a dangerous counterattack.
For example: 1... K4b (the best defense to this plan) 2 P7f G7b 3 B6f S8b 4 P9f P7d 5 P9e P6d (if white plays 5... G7c directly a later ... G8d can be met by B5e) 6 P5f G7c 7 P9d (better is 7 B5g and switching to attacking white's left, as recommended by the late Oyama, 15th lifetime Meijin) ... Px9d 8 Lx9d G8d 9 R9h P*9e 10 Bx8d !? (strong but risky; however 10 L1b+ S7c is not very effective; the promotion of the lance is not enough reward for all the time black has invested in his attack) Px8d 11 Rx9e P3d (giving the king some place to run) 12 L9c+ (12 L9b+ though slower is probably better) Sx9c 13 Rx9c+.
The exchange of bishop and lance for gold, silver, and pawn is about even, and black has promoted his rook. Black can follow up by P7e, answering ... Px7e by P*7d and P7c+ next, though the promoted pawn will be a bit far from the enemy king. However the white king can run to 3c and after G3b or S3b he has a decent castle and can complicate the game with his two pieces in hand. In my opinion a weak player cannot beat a pro from this position.
So it seems that it's not so easy for black to win at this huge handicap. One book by Naito Kunio 9 Dan even went so far as to advocate a slow buildup with (in the preceding line) 7 N7g G5b 8 P4f P5d 9 S3h G5c 10 S4g, which he calls the "missile silo" system, the bishop being the missile and the silver on 4g plus the center pawns forming a "silo" for the bishop to rest safely on 5g.
The point is that unless you plan to attack quickly, the head of the bishop or the knight (the square in front of it) is a weak point and should be defended by a general if possible. Then black has plenty of time to play for an attack on either or both wings, depending on white's play. However it does give white time to make an ideal formation, and it does not seem to me to give black as large an advantage as he deserves at six piece. I do think this system is pretty reasonable at five piece (right) handicap, where white may have enough time to bother the bishop.
In my opinion, the best system for black at six piece handicap is one recommended by former professional player Hiroyuki Iida, who has published two books of his handicap games. The idea is to attack first white's left edge, then his right, getting full value out of the handicap.
Let's see: 1... S2b 2 P1f G3b 3 P1e S8b 4 L1g P2d (else 5 R1h and 6 P1d is quick and easy) 5 R1h (Kazuharu Shoshi 6 Dan recommends instead 5 P7f K4b 6 P2f, followed by S3h-2g-3f and then P2e, which is an excellent alternative to the text) G2c. Now 6 P1d? Px1d 7 Lx1d? P*1c just traps the lance. So black's attack appears to be a failure.
Considering only this part of the board, that is true. However, black's plan is to switch to the other flank, in which case white will suddenly wish that his left generals were back home where they might defend his king against attack from his right.
So, 6 P7f K4b (both to guard 3c better and to run away from attack on his right) 7 P9f G7b 8 P9e (8 B6f first looks better, but white can transpose to the text lines or even sacrifice a pawn by 8... P8d, so it really doesn't matter). Now white must choose between ... P8d or ... P7d (intending eventually ... G-7c-8d).
First let's consider 8... P7d: 9 B6f P6d (if 9... G7c 10 P9d Px9d 11 Lx9d G8d 12 B5e! S7a 13 R9h P*9e 14 L9c+ S6b 15 B8b+) 10 P5f (so the bishop can remain pointing at 9c after ... P6e) G7c (probably white should just play 10... K3b, allowing 11 P9d Px9d 12 Lx9d, though 13 R-9h next will lead to promoting lance, rook, and perhaps also bishop) 11 P9d Px9d 12 Lx9d G8d 13 R9h P*9e (or 13... S7c 14 L9c+ P*9e 15 Bx8d! Sx8d 16 +Lx8c, which is similar) 14 Bx8d! Px8d 15 Rx9e P3d 16 L9c+ (or 16 L9b+) Sx9c 17 Rx9c+.
In contrast to the similar position analyzed above with the left generals at home, white cannot build a castle for his king nor can he interpose generals between his king and the black dragon (promoted rook). This position is an easy win.
So probably better is 8... P8d: Then 9 B6f G8c 10 P5f (to give the bishop an escape) and now 10... P7d or 10... G7d.
If 10... P7d 11 L9g white is temporarily safe on both edges but has no piece free to wander but his king. Black can develop leisurely or attack either edge, for example: 11... P6d 12 B5g P3d 13 P2f P5d 14 P2e Px2e 15 P1d Px1d 16 Lx1d P*1c 17 Lx1c+ Sx1c 18 Bx1c+ Gx1c 19 Rx1c+.
Just as in the 8... P7d line the material exchange is about even, but again the lack of any generals in position to help the king against the dragon means that white's game is hopeless.
So best is 10... G7d: Then 11 N7g (to prevent ... G6e) S8c 12 L9g, followed by R9h and P9d will lead to a break-in with no sacrifice, while white still has the awful shape of two generals on what will be the wrong side of his king. This is the variation quoted by Iida.
Although it probably represents best play, white's position is critical, and I think a player in the 5-7 kyu range could defeat the Meijin from here. If you play it right, six piece handicap is really decisive!
Lessons from this handicap :