A modest proposal....

You can drop the Elephant or the Hawk when the:

    a)  King, Queen, or Rook moves (but not the Bishop or Knight anymore)

    b)  Drop one when you castle on either vacated square (same)

    c) Allow for the units to be "daisy-chain dropped", for example:

        12. Qd3(Ed1)   ...some time later...   15. Ee3(Hd1)

Why the change...

     As it is now the fianchetto is busted in Seirawan Chess and this would give it a chance to become viable again.  With Seirawan Chess in its current state you get the "Bishop on a stick" problem of a fianchetto being busted too easily.  Now admittedly a fianchettoed bishop can be threatened with exchange in 3 moves in standard chess.  This is not about keeping it from happening at all.  But just restoring things back to standard chess.  The person would have more time to try and scuttle the trade attempt or do something else as it would take 3 moves again.  The bishops might not get traded quite as quickly in other circumstances as well.

Other ideas, less likely perhaps...

Also liberating the King more by letting him hop over check when castling to balance new defensive resources with offensive ones.

Rename the game "Dropablanca", because you are DROPping the CapABLANCA units onto the board.

Perhaps as a further variant, allowing for the extra pawns to be dropped too.  Capablanca chess has 10 pawns, they could be dropped like the Elephant and Hawk are.  This last idea though may lead to too much complexity and tedium. Dropablanca X  (X for 10 pawns) could be that variant's name.

Post Script:

I have tried Capablanc Chess variants and found that the sum is less than the parts.  You have trouble immediately on the flanks with pawns being attacked by the opposing players bishops before they even move.  Fianchettoing (Spelling?) is not much better in some cases.  Undefended pawns in the opening is another issue.  Ten pawns look great until you realize they are impossible to defend well.  The fact the the knight and king become useless on such a large board.  10 x 8 boards are not common either, although a printer and a spreadsheet program can overcome that fairly easily....

Hi Dropablanca and welcome. 

I read over your modest proposals, but since I'm still trying to get my head around the operation of the elephant and hawk, I'd rather wait a while before entertaining any modifications. 

As far as the fianchetto is concerned, it seems to me that if the player executing the fianchetto can fortify it so that it is even more impregnable than in regular chess. For example, in a game as white against Roy Gates (Gonnasuke) I tried 1. e4 b6 2. e4 e6 3. Bd3(H) to break into the white squares on the queen's side at some point, whereupon Roy calmly replied 3... Bb7(H) and I had to choose another course. White has an extra piece to break down a fianchetto but black, it seems to me has and extra two pieces to fortify it (i.e. in the case of black's queen side fianchetto, he has Bb7(Hc8) and alternatively, b6, Bb7, d6, and Nbd7 (Eb8)) I don't see the problem. 

Panzerschiff, who participates in this forum, has had considerable experience with chess variants. I'm sure he will be interested in your ideas.

Robert said:  1. e4 b6 2. e4 e6 3. Bd3(H)

Robert meant:  1. e4 b6 2. d4 e6 3. Bd3(H)...

Well, I see your point.  I also saw this although it may not be as good:

1.e4 b6 2. d4 a6  and now the Bishop and Rook without any help already deter anything on a6.  What had me concerned though was the fact that king-side fianchettos are obsolete.  For example...

1.d4 Nf6 2.c4 g6 3.Nc3 (Bh6 here loses the bishop to BxB(H)) However, what is black to do?  3....Bg7(H) prevents castling, dropping the Elephant when the Knight moves out does too, and 3....Bg7 4. Bh6(H) is not nice to look at.  4....O-O(He8) 5. BxB HxB 6. He3 to be followed by Qd2 and a second round of bunker-busting.  Perhaps the queen-side can be saved but a lot of good defenses go down the drain too quickly on the king-side.  That was the basis for my proposal.

Sero-Indian Defense
(It would appear that maybe there is a way to salvage a king-side fianchetto.)
1. d4 Nf6 2. c4 g6 3. Nc3 Bg7 4. Bh6 (Hc1) O-O(He8) 5.Bxg7 Hxg7 6. He3 Kh8 7. Qc1 Ng8 (Yet this may be far to passive for Black to try in a game.) 
Or 7....Rg8 maybe 8. Hh6 (threatens f7#) Qf8(Ed8) 9. HxH QxH .....

A funny thing just occurred to me.  Why not use the Queen to plug the hole in a fianchetto if hypothetically the bishop and then hawk would be traded?  The Elephant dropped at d8 should be a big enough boy to handle the d-file chores that the Queen does?

I did this on the fly so there is probably errors, comments anyone?

Sero-Indian Defense (Continued)

Two more lines to play with...

1. d4 g6 2. c4 e5 3. dxe5 Bg7(Hf8) 4. Nf3 ...

Branch A (4. ... Nc6 5. Bf4(Hc1) Nh6 6. e4 Ng4 7. Hd3)
Branch B (4. ... He6 5. Bg5 Ne7 6. Bh6 O-O(Ee8) 7. Bxg7 Hxg7 8. e4  Nbc6 9. Hd3)

It all still seems rather uneasy for black in my mind, but I have not spent much time on it.

You obviously have a better grasp of openings for this than I do.  I am going to digest what you posted and come back with some finer tuned stuff.

This is an interesting discussion.  I think Seirawan's own idea of slowing down play might make the king-side fianchetto more viable in S-Chess.  Somewhere he mentioned the possibility of separating the entry move from the normal move and you instead would just spend a move making a drop for the new pieces.  I personally like the faster pace, but if the advantage of the first move is too big with the combined move/drop combined with the two extra powerful pieces this would be an option, perhaps better than restricting entry to the movements of the big pieces and the king.

Also in the 1. d4 Nf6 2. c4 g6 3. Nc3 Bg7 4. Bh6(H) 0-0(He8) perhaps a direct attack on the king with a few sacrifices like what Seirawan, himself demonstrated in a Youtube piece might be an option here and can appear to be really scary for the fianchetto, but black again has his resources.  For instance 5. h4 d6 6. h5 Nxh5 7. Rxh5/Eh1 gh5 8.Exh5.  White can now offer up a further "Gothic Exchange" sacrifice if black is a bit greedy say with 8...Hf6 9 e4!? (Less accurate seems to be 9.Exg7+ Hxg7 10.Bxg7 Kxg7 11. e4 f6 12 Qh5 Qe8/Ed8 13.Hh6+ Kh8 when black seems to be able to defend himself.) 9...Hxh5 10.Qxh5 f6 11.Bxg7 Kxg7 12.Hh6+ Kh8 13.Qf7 A little showy, but black can again hold on with simply 13...Rg8 14. Bd3 Qe8/Ed8. Black also can probably avoid a lot of the "drama" and play simply 8...f5.

Thanks, although looking at my analysis with a more alert "morning" brain, I wonder if white's attack could be further strengthened by playing 9.Bxg7 since Hxh5 is impossible because of 10 Hh6 mate. 9...Hxg7 then is forced when white's attack looks dangerous 10 Hh6+ Hxh6 11.Exh6+ Kg7 12. Eh5+ Kh8 13 Qd2.  That probably makes 8....f5 9.Bxg7 Hxg7 10. Hh6+ Hxh6 11. Exh6+ Kg7 12. Eh5+ Kh8 13. Qd2 Qe8/Ed8 necessary after all for the defense and not the case of avoiding a "little drama".

Also one might wonder if 3. Nf3 is more accurate if white intends the h4 bayonet like attacks.  That Knight then would be more available for an attack on the king-side.  Black might also change his play and reinforce the fianchetto with 3...Bg7/Hf8 at least preventing a quick Bh6.  Unfortunately outside preventing the bishop exchange, he  otherwise isn't that well placed at f8 this early in the game.  Probably a more waiting game with 3...d6 to see what white wants to do with his Hawk is better as there is always the temptation to revert to the natural developing move 4.Nc3/Hb1 in the Queen-side openings.

Sharper Chess does have the advantage of describing the game plus honoring both inventors and we can still use the abbreviation S-Chess in both cases!

Are there any computer programs that play S-Chess, I can't find any?

If there is none is one at least being worked on?

Dropcapa--No, and there are no databases either. And that's a good thing!

Panzerschiff. I spent quite a while looking at the ramifications of 1. d4 Nf6 2. c4 g6 3. Nc3 Bg7 4. Bh6(H) 0-0 (Hh8) This placement of the Hawk, a supercharged fianchetto, makes a difference in some lines, makes no difference to the immediate exchange, of course, and I find it aesthetically pleasing.5. h4 and now instead of 5.... d6, I tried 5.... Ng4 and discovered that this move gives black good counter play in all lines if white retreats the bishop. Unfortunately, the straightforward 6. Bg7 is a killer. For example:.7..... Hg7 8.h5 gh ( 8.... e5 9. hg fg 10. Rh7(E) Nf2 11. Rg7+ Kg7 12. Hh6+ Kh7 13. Hg5++ Kg8 14. Hd8 ) 9.Rh5(E) h6 (9....Nf6 10. Rg5) 10. f3 and black is toast.

  Dropcapa  You might check to see if there is a file available on the Chessvariants web site.  The Zillions of Games Engine can play hundreds of different chess variants and while they might not have an exact match for S-Chess the limited drop idea has been used in other variants and something similar might already be available.  The Zillions engine is not particularly strong and might play at a low Class A level.  That is its penalty for being able to play about any type of chess variant out there.

More on the SeiraIndian, Panzerschiff Variation. After pushing the pieces around for quite some time, it seems that Panzerschiff's defense 1. d4 Nf6 2. c4 g6  3. Nc3 Bg7 4. Bh6(H) 0-0 (He8) 5. h4 d6 6. h5 Nh5 7. Rh5(E) gh5 8. Eh5 8. f5 9.Bxg7 Hxg7 10. Hh6+ Hxh6 11. Exh6+ Kg7 12. Eh5+ Kh8 13. Qd2 Qe8/Ed8 not only staves off the assault to the fianchettoed bunker, but allows black to turn the tables. At that point, I thought, why should white be in such a hurry to tear apart the king's side? After all things are balanced there. To that end why not leave the h-pawn on the fourth rank and play 6.e4 as per the King's Indian. It turns out that black has to take some care with his next move. Let's see what happens if he doesn't: 6 ... e5 as per the KID 7. de! whereupon black realizes that recapturing straight away is out because white will play Qd8(E) in reply, so he plays, 7 ... Ng4 when 8. Bg7 (8. Bg5!? f6 9. ef Bf6 10. f3 Ne5 is good too.)Hg7 9. Hg5 f6 10. ef Hf6 ( 10 .... Nf6? 11. h5!) 11. Qd2 and white has a pawn and a good position. 6 ... Nbd7 provokes 7. e5while after  6 ... Nc6 (or c6 or a6) white crashes through on the kings wing as in Panzerschiff's analysis with 8. h5 Nh5 9. Rh5(E) gh5 9. Eh5 f5 10. Bg7 Hg7 11. Hh6+ Hh6 12. Eh6+ Kg7 12. Qh5! (+/-) This leaves 6 ... c5 when white can capture and defend his booty with 7. dc Qa5(E) 8.Hb3 Qc7 9. Bg7 Hg7 10. cd Ed6 11. Qc2(E) or play 7. d5 creating a Benoni structure where the h-bomb hangs over black's head and it's hard to see how he will get counter play.

 A bit late now and I'll have to look at the analysis tomorrow.  I did have a chance to look at the chessvariants site and see if they had anything new on Seirawan Chess. They do have a Seirawan Chess Courier board set up, which can be used for correspondence games.  It is a little clumsy and you would have to type in say after 1. c2-c4 e7-e5 2. N b1-c1; A-b1 (for a Hawk entry on b1. Black would use a lower case "a". They use the universal Capablanca designation A for Archbishop and  M for Marshall for the R+N piece) There is not a Seirawan Chess Zillions file although something called "IAGO standard Fantasy Pocket Chess" can be downloaded in the Zillions engine. It is the slower form where the pieces enter on any square in the back row, but as an individual and separate move. 

     Responding to various points of this thread in no particular order:

     I went to Zillions and found Azerbaijan [Spelling?] Chess, which is something like my Dropablanca X proposal.  The Courier board Panzerschiff found is something, but what we really need is a full-on GUI chess client that lets us play each other over the internet, or a server to host it.  That was what I wanted to say when I mentioned a computer that could play S-Chess earlier.  Good Lord, I couldn't beat either of you let alone a computer! 

     I had made screen shots of S-Chess position mockups to post for images of critical positions of this thread, but I decided not too.  Reason one, chess.com's uploader is not as friendly as I had hoped.  Reason two, all of the work so far can be summed up rather easily. 

If you are fianchettoing as a means to castling, DON'T DO IT!  It will trigger a "Rook Three Attack" that will leave the fianchetto largely gutted and your King walking a tightrope.  If you are fianchettoing and not castling behind it you should be fine, and even get great benefit from the fianchetto. 

A "Rook Three Attack" is a name I picked out because no matter which fianchetto you do, of the four availible for both sides, the respective rook three square can be unloaded on by the elephant attack plus bishop-hawk battery too easily.

As a summary of what has gone so far:

I was wrong to say fianchettos are totally broken.  You can avoid being checkmated if you castle behind one, but I would not recommend fianchettoing to castle.  The fianchettoed castle will get badly mauled by a very easy combination your opponent can unload.  Going back to the beginning of the thread, I think the best fianchetto is one like Robert demonstrated.  1. e4 b6 2. d4 e6 3. Bd3(Hf1) Bb7(Hc8) .... and no castling to that side.

The only other things I can think of is 1. d4 d5 2. c4 c6 and if 3. Nc3(Hb1) then Black might be able to spring g6 as a surprise, or Nf6 then g6.  This does nothing to stop the Elephant attack.  What you end up with is a waiting game, zugszwang in the opeining of all things, where each side waits to see where the other one drops their units before committing to something.  You would be trying to run the other guy out of moves before you do and have to commit to a formation.  The last thing that occured to me was playing h5 to respond to h4.  Yet what is the point in castling if you are going to have to move your pawns so quickly?

I wrote to Yasser Seirawan suggesting, among other things, that an online GUI chess client, as mentioned by Dropcapablanca, was necessary for the popularization of the game. That letter and Yasser's response may be viewed here:

http://www.chess.com/groups/forumview/my-letter-to-yasser-seirawan-and-his-reply

I also wrote to the site administration of Chess.com inquiring as to the feasibility of a S-Chess client. The wrote back to the effect that that they were up to their armpits trying to make this site the best chess site on the web and that they had no plans for variants.  

The King's Indian set up is certainly a different breed of cat from the same structure in regular chess. And this goes for most of the other openings too, as I'm finding out a some cost. Yet if the King's Indian structure proves to be problematic for black (and that's by no means certain yet) he can take consolation in fact that in the black side of the Sicilian, he appears to be able to equalize at once. Thus: 1. e4 c5 2. Nf3 Nc6(H)! when if white continues with the rote 3. d4 he finds himself in the jam after 3 ... cd 4. Nd4 Nd4 5. Qd4 Hc6 when it turns out his best move is probably 6. Qd1 since 6. Qe3?? and 6. Qd3 lose at once to 6 .... Hb4+ and 6. Qd2 loses a pawn for squat. Notice that deploying the elephant or hawk on his fifth move doesn't help white since d1  clogs up the retreat square and his pieces just get in each other's way. Black's hawk stoops and snatches white's pawn at its peril now as white's pieces come flying out after 6 .... He4 ?! 7. Nc3 Hc6 ( 7 .... Hf6 8. Bg5(H) wins the bird.) 8. Bb5 (H) Hb8. But 6 ... Nf6 followed by 7 ... a6 gives black a fine game.

Dropcapa's suggestion of h5 has been successfully employed by Dragon specialists. If we can't revive the KID structure, that may well be an excellent place to look next.

  I would agree that the slower build up with e4 is better and probably better yet is holding back the h4 until development is a bit more complete, although as long as the Elephant has not been committed the potential edge attack threat is always a big worry as Dropablanca notes above.  With a slower white build-up, maybe something more inline with a Saemish-like structure keeps those attacking options in reserve while building up one's forces in a more controlled fashion. However, I would not quite give up on that King's Indian yet.  At least in that early bayonent charge, Black does not have to accept white's  exchange offer and in the line with 6...Nc6 black can still obtain a playable game against 7.h5 with 7...,e5!?  and if white is still in a hurry black should do O.K. after 8.Bxg7 Hxg7 9.hg6 fg6 10.d5 Nd4 11.Hh6+ Hxh6 12.Rxh6/Eh1 Qe7/Ed8 when ...Ng4 is even in the air. 

Black might also investigate "Rat-like" stuff: 1...g6 2...Bg7 3...d6 4...Nc6.  Holding the Knight back on g8 discourages a quick Bh6/Hc1.

Yes, I did see that Azerbaijan Chess variant when I revisited the Chess Variant Website.  Adding two extra pawns to the drop is an interesting twist.  A good GUI play system would be nice for Seirawan Chess.  I remember mentioning to Bruce Harper that they might contact the people at Brainking.com.  The Czech-run Brainking site already have various chess variants including many large board variants that use these pieces as well as the games Shogi and Chessgi that have drops.  It probably would not involve much programming to offer Seirawan Chess as well.  Perhaps just some kind of agreement and interest by people on that site to play the variant.

I sent a letter to a Danish programmer asking what is involved in assembling an S-Chess client. The fellow has developed a strong program called Glaurung. He's also developed a program for four handed chess. So far, I haven't received an answer. If you guys know anyone capable of accomplishing this task, let me know and I'll contact them. If you don't ask you don't get. 

Cheers 

Rob

btw wonderful Avatar Dropablanca!