A game(time: 25/5) played on FICS a few minutes ago. During the game, I thought he was playing some radom moves in the opening to confuse me. I checked with OE after the game and he knew what he was doing until move 7.

1.e4 d6 2.d4 Nf6 3.Nc3 g6 4.Nf3 Bg7 5.Be2 0-0 6.Bg5 c6 7.Qd2[diagram in the link above] Nbd7 8.Bh6 e5 9.Bxg7 Kxg7 10.h4 exd4 11.Qxd4 Nc5 12.h5 Ne6

click for larger view

Here, I thought of playing 13.h6+ and was hoping I would gain access to the g7 square in the future. But I remembered a puzzle posted by <sneaky> in his forum where he sac'd the h-pawn for a powerful kingside attack. So, I analyzed a similar sac here and it looked good!

13.Qd2 Nxh5 14.g4 Nhf4 15.Nd4(attacking the knight on f4 with my queen) 15...d5 16.Nxe6+ Nxe6 17.exd5 cxd5

click for larger view

Here 18.Qh6+ wouldn't work as the black king can escape via f6, e7, e8. So, I decided to restrict king's escape squares first.

18.Nxd5 Qg5 19.f4 Qd8 20.f5 Ng5 21.f6+ and the knight is gone. But the game went on...

I have heard that in general, the B on e7 is more solid, yet passive, and the bishop on g7 is riskier, but gives better attacking chances. Also, I've heard that the pawn on c4 gives white better central control, but may prove to be a weakness in the endgame. These generalizations have lead me to 2 conclusions, and I'm wondering how accurate they are: 1) The Pirc is better than the Philidor. This is because without white's pawn on c4, black's attack/counterattack has a better chance of success, and black shouldn't sit and wait for an endgame. 2) The OI is better than the KID. This is because with white's pawn on c4, black's attack/counterattack doesn't have as good a chance of succeeding, and black will be fine holding out for a favorable endgame.

<Also, I've heard that the pawn on c4 gives white better central control, but may prove to be a weakness in the endgame.> This seems groundless to me. The pawn seems just as likely (or unlikely) to wind up an endgame weakness at c2 or c3. It all depends on how the rest of the game goes. In many lines of the KID the pawn goes to c5 and then is exchanged for black's pawn on d6, thus not surviving to the endgame at all.

As for the relative strength of these defenses, I think it's impossible to say (besides you are getting advice from Ray Keene on his page, which is of much greater value than my opinion anyway). The KID and the Pirc are more popular than the Old Indian and Philidor's today. Partly that might just reflect changes in chess fashion, which can be as irrational as changes in any other kinds of fashion. But partly it is that the KID and Pirc are generally more dynamic that their Old Indian and Philidorean antecedents. In the OI and the Philidor (overgeneralizing again) Black places his center pawns first, and tends to his other pawns and pieces later; in the KID and Pirc, Black develops his flank first and figures out what to do with his center pawns later. Since it is the placement of the center pawns that does so much to determine the strategic contours of the game, Black has a little more freedom in the KID and Pirc. (But not too much -- Black can't take all day to decide what to do about a strong pawn center!)

<Also, sometimes white delays development of his QN and the c-pawn, so you don't know for a while how that's going to go.> Black can cut down on White's freedom depending on how he plays. 1. e4 d6 2. d4 Nf6. Now white has to make a decision about how to defend the e-pawn. Usually he plays 3. Nc3 g6, and you've reached the main line of the Pirc. But 1. e4 g6 2. d4 Bg7 gives White more choices. (It gives Black more choices too, which is why a lot of people like it.)

The King's Indian usually starts 1. d4 Nf6 2. c4 g6 3. Nc3 Bg7. But the King's Indian setup can be used against pretty much anything White plays after 1. d4 Nf6. (Or for that matter, pretty much anything White plays after 1. c4 or 1. Nf3.) The Pirc and KID are complimentary: the first can be used against 1. e4 and the second against almost anything else.

<1) The Pirc is better than the Philidor. ... 2) The OI is better than the KID.>

I don't think these kinds of generalizations are worth making. I have no idea if either one is true or false, and it doesn't matter anyway. As a GM said, below the master level all openings are playable. The key is to find something you like.

1. e4 Nf6 2. Nc3 d6 (<If I can't get a normal Alekhine, I go for the Pirc.>) 3. d4 g6 4. Nf3 Bg7 5. Be3 O-O 6. Bc4 c6 7. O-O (<Maybe 7...Nxe4 8. Nxe4 d5 is good for me, but I didn't notice it until now.>) b5 8.Bb3 a5 9. a3 Qc7 10. e5 (<Hey, I wanted to put my pawn there!>) dxe5 11. dxe5 Ne8 (<I didn't want to block in my bishop with Nfd7. Perhaps I should have.>) 12. Ne4 (<Yay, I win a pawn!, Boo, I lose my dark square protector.>) Bxe5 13. Nxe5 Qxe5 14.Bh6 (<Uh, what are you trying to do?>) Qxe4 15. Re1 Qg4 (<Please trade queens.>) 16. Qd8 Qd7 (<Pretty please?>) 17. Qb6 Ng7 18. Rad1 Qb7 (<With sugar on top?>) 19. Qd4

click for larger view

(<Okay, just take a look at this. First, you see that white has his rooks and queen on the central open files, and the bishops and queen aimed diagonally at black's kingside. White's king is happy and safe. Black apparently forgot how the pieces move -- fianchettoed knight, nearly imprisoned queen, and 3 untouched queenside pieces. Then you see that black is up a knight. Next you see that white will win back the knight with mate on g7. Finally, you notice that you don't need to protect the knight, you need to protect the square the knight is on, and so...>)

19...Nf5 20.Qh4 Nxh6 21. Qxh6 e6 (<Gotta block white's bishop, right?>) 22. h4 (<Hmmm, can he win this with the immediate 22. Re4?>) Na6 (<And 22...Nc6 loses to 23. Rxe6, right?>) 23. h5 Nc5 24. hxg6 fxg6 25. Re3 Nxb3 26.cxb3 Qg7 27. Qh4 Qf6 28. Qg3 (<Still won't trade queens...>)