Vladimir Kramnik has proposed on chess.com that castling be banned for the good of the game.

What are the kinds of objective points that might be made for & against this idea? Based on these, do you think that Kramnik's idea can succeed?

I am keen to avoid this becoming merely opinion-based. But it's such an interesting idea that it must be worth discussing. I do have a perspective, but am keen to hear others' ideas first.

  • 17
    It would also be interesting to hear why castling was introduced in the first place (probably to speed the game up). It may have been so early that there are no records though. Dec 1, 2019 at 12:12
  • 3
    I'd like to hear arguments for keeping castling that don't boil down to "If it was good enough for my grandfather, it's good enough for me"... I imagine the game without, and you trying to convince me to adopt this new rule: It has always seemed to me like a magical move with arbitrary restrictions (If the king can jump over a rook, why can't he jump over a rook that's moved one spot closer? Because he can jump shorter towards H-rook than towards A-rook, if A-rook moves to B why prevent castling?). It just doesn't make sense to me, it's an arbitrary rule+conditions. Dec 3, 2019 at 13:46

8 Answers 8


It would certainly allow for more attacks due to kings being stuck in the center, but fundamentally changing the game, which in a way dumbs it down, is not good. It would be less complex.

I also do not want to think that I spent 40 years of my life studying something only to have it changed. I do not want the rug pulled out from under me like that.

  • 25
    I’m generally not in favor of “arguments from tradition,” but I think in this case it’s valid. Chess is pretty old, and I like the idea that the core of the game hasn’t changes for a few hundred years. Nov 30, 2019 at 22:47
  • 8
    @NoahCaplinger The high draw rate at the highest level of play says something does need to change.
    – corsiKa
    Dec 1, 2019 at 2:18
  • 20
    Making a change could create a dangerous precedent. Why not allow taking your own pieces? That surely will allow for more attacking possibilities, and so on..., and this could lead to a mess. It is fine to add variants, but the main version of chess is perfectly fine.
    – Akavall
    Dec 1, 2019 at 2:33
  • 18
    @corsiKa, "draw rate at the highest level", I don't see it as big problem, people are getting better and better, so more draws are expected. And for me personally, a fighting draw is just as interesting as decisive result.
    – Akavall
    Dec 1, 2019 at 2:37
  • 22
    I actually view the high draw rate at the high levels as being proof that chess is perfect as it is. If there were high draw rates at the lower levels then I might agree that there is a problem.
    – dotancohen
    Dec 1, 2019 at 7:42

I have no real preference either way, but I'll point out that Kramnik based his argument on AlphaZero playing against itself. Because of AlphaZero's style, it might not be the case that these no-castling games end up in complicated positions more often.

Here're two more games to add using the no-castling rule, played in the TCEC viewer-submitted opening bonus games. Komodo (the second-strongest traditional engine) plays against Scorpio, one of the newer NN engines (objectively, it's likely weaker than Leela, Allie, Fat Fritz, and Stoofvlees). To this patzer's eye, they do look pretty complicated, adding more evidence for Kramnik's position. Caveat: because Scorpio is also an NN engine, it might play more similarly to AlphaZero.

[FEN ""]
[White "Scorpio"]
[Black "Komodo"]
[Result "1/2-1/2"]

1. Nf3 Nf6 2. Nc3 Nc6 3. Rg1 Rg8 4. Rb1 Rb8 5. Rh1 Ra8 6. Ra1 Rh8 7. Ng1 Ng8 8. Nb1 Nb8 9. c4 c5 10. Nf3 Nc6 11. d4 cxd4 12. Nxd4 g6 13. e4 e5 14. Nc2 Bc5 15. Nc3 Kf8 16. Rb1 a5 17. h4 d6 18. Bg5 Qe8 19. h5 h6 20. hxg6 fxg6 21. Be3 Be6 22. f3 Kg7 23. Qd2 h5 24. Na3 Nd4 25. Nab5 Nxb5 26. Nxb5 Qe7 27. Bg5 Nf6 28. a3 a4 29. Kd1 Qf7 30. Rc1 Rac8 31. Rc2 Rc6 32. Kc1 h4 33. Be2 h3 34. gxh3 Nh5 35. h4 Ng3 36. Rd1 Kh7 37. Qe1 Nxe2+ 38. Qxe2 Rf8 39. Rc3 Bb4 40. axb4 Bxc4 41. Rxc4 Qxc4+ 42. Qxc4 Rxc4+ 43. Kb1 Rxb4 44. Nxd6 Rxf3 45. Rd2 a3 46. Ka2 b5 47. bxa3 Rbb3 48. Ne8 Rxa3+ 49. Kb1 Rfb3+ 50. Kc1 Rc3+ 51. Rc2 Kg8 52. Nd6 Rxc2+ 53. Kxc2 b4 54. Bf6 Kf8 55. Bxe5 Ke7 56. Nb7 Rh3 57. Bd6+ Kf6 58. Bxb4 Rxh4 59. Bc3+ Ke6 60. Kd3 Rh3+ 61. Kd4 g5 62. Nc5+ Kf7 63. Nd3 Ke6 64. Bd2 Kf6 65. Ne5 Rb3 66. Ba5 Ke6 67. Nd3 Rb1 68. Bc3 Kf6 69. Kd5+ Kg6 70. Bd4 g4 71. e5 Kf7 72. Bf2 Rd1 73. Ke4 Rf1 74. Bh4 Rh1 75. Bg5 Rd1 76. Nf4 Re1+ 77. Kf5 g3 78. Bd8 Rh1 79. Bb6 Rf1 80. Kg4 Rb1 81. Ba5 g2 82. Nxg2 1/2-1/2

Reverse game:

[FEN ""]
[White "Komodo"]
[Black "Scorpio"]
[Result "1-0"]

1. Nf3 Nf6 2. Nc3 Nc6 3. Rg1 Rg8 4. Rb1 Rb8 5. Rh1 Ra8 6. Ra1 Rh8 7. Ng1 Ng8 8. Nb1 Nb8 9. d4 d5 10. c4 dxc4 11. e3 e6 12. Nc3 a6 13. a4 b6 14. Bxc4 Nf6 15. e4 Nc6 16. Nge2 Nb4 17. f3 c5 18. Be3 cxd4 19. Nxd4 Bc5 20. Qb3 Qc7 21. Rc1 Ke7 22. a5 Bd7 23. axb6 Qxb6 24. e5 Nfd5 25. Nxd5+ exd5 26. Bb5 Ke8 27. Nf5 Qxb5 28. Nxg7+ Kf8 29. Bxc5+ Kxg7 30. Qxb4 Qxb4+ 31. Bxb4 a5 32. Bc3 Be6 33. Kd2 Kg6 34. Ra1 a4 35. Ra3 Rhc8 36. h3 Kg5 37. Ke3 Rc4 38. g4 h5 39. f4+ Kh6 40. f5 Bd7 41. Bd4 Kg5 42. h4+ Kxg4 43. e6 Be8 44. Rf1 Rb4 45. Rf4+ Kg3 46. Ra1 Rb3+ 47. Bc3 a3 48. Rg1+ Kh2 49. Rgf1 d4+ 50. Rxd4 axb2 51. Rf2+ Kg3 52. Rf3+ Kh2 53. Rd2+ Kg1 54. Rg3+ Kh1 55. Kf2 Kh2 56. Rdd3 b1=Q 57. Rh3# 1-0

Update: Chessbase reports that a tournament between fairly-strong players (about 2300-2400) resulted in 89% decisive games. Caveat: because the format is new, one would expect more decisive games. If no castling chess became the standard, theory will develop, openings will become more refined, and we should expect draws to increase in frequency.

  • 4
    Sorry to be a little off topic, but what's the point of the little "dance" at the beginning of these two games where the knights and rooks all move then move back?
    – JPhi1618
    Dec 2, 2019 at 21:07
  • 15
    @JPhi1618 Presumably it's a predefined set of opening moves that render castling illegal (as the rooks have been moved) and then manually reset all pieces to their original starting positions, thus simulating games in which castling is banned from that point forward.
    – Abion47
    Dec 2, 2019 at 22:53
  • 6
    @JPhi1618 it's so that any standard engine can be used for analysis without having to customize it with a "no castling" rule, so after this 8 turn fixed opening we're at the starting position but castling is not possible anymore.
    – Peteris
    Dec 3, 2019 at 9:37

This answer will focus on chess as it is played, rather than subdomains like composition or engine development.

Castling is a move that usually accomplishes at least one of two things:

  1. It gets the king to safety.
  2. It develops the rook.

Or both, in just one move. As castling stands now, it requires one move to reach a position that would otherwise take three moves. This accelerates the game considerably, where a player can get king safety and an attack quickly.

[As an aside, historically, the current version of castling is not the first. There have been several others during the 16th to 18th century, like swapping the king and rook directly, or "free castling" (put the king and rook somewhere on the squares between them, inclusive.) They all had a slightly different flavour from modern castling and are interesting in their own right, but the effect is the same, returning to the two points above.]

The effect of banning castling would be to remove this acceleration of the game. However, it does not automatically make the game slow or boring. I draw from my experience of shogi for this part regarding openings.

  • In the opening, the choice between king safety and rapid attack would be even more pronounced than before. This would of course radically change opening theory (move sequences), but also opening strategy (what to prioritise).
  • Players could consider spending several moves adopting structures that allow the king to walk from the centre into safety in the wings. This multi-move strategy is termed "castling" in shogi (unfortunate name), and is a large and varied part of opening strategy.
  • There would be options to forgo this "castling" in favour of immediate attack, leaving the king in the centre. This is clearly double-edged.
  • If executing "castling", players would have to be cautious - the king remains fairly vulnerable for a while, and players have to maintain flexibility to avoid being caught off guard during this.

Of course, shogi is its own completely different game (drops make it more akin to crazyhouse chess than regular chess), but the general ideas behind its opening strategy would seem applicable to no-castling regular chess, hence why I refer to it in this answer.

What is certain though is that this change would have the desired effect: opening theory (move sequences) would need to be completely revised as many lines hinge on castling quickly. Opening strategy would also change to an extent, where the balance between king safety and rapid development adds another dimension to the game.

On the other hand, endgame theory would not change (barring a tiny handful of composed studies). Middlegame skill and general "chess skill" would be also largely untouched, perhaps even promoted in importance as the game would for a while rely more heavily on piece coordination and chess sense than on opening lines.

This no-castling change seems interesting and has the advantages of (a) not changing endgame theory, unlike the stalemate = win suggestion often brought up, and (b) not changing pattern recognition, unlike chess960 which yields unfamiliar positions from the get-go.


Pros: Tal will rise from the dead in excitement at castling not being allowed. Cons: Fischer will also rise from the dead in excitement. Undecided: Petrosian will have many of his masterpieces, where he castled late as was his style, left to not be worth as much. How does he feel about this? Who knows?

In all seriousness, this would make the advantage of playing White much bigger. Opening theory would develop a lot more gambits for White. I expect that after a year there would be an irresistible call to modify the rules over checkmate; one way would be to make checkmate only possible after 20 moves and any checkmate before then is a draw.

I agree with previous posts; if people want a chess variant there are many already. Just for laughs, if I had to ask myself "what would be the fairest way of having castling not allowed?" maybe if you give a check then from then on castling is allowed. It could be more interesting still if there was a move limit ; after the check you get 10 moves of allowable castling. Or for a crazy idea, allow castling anywhere for the whole game but the king and rook have to be vertically spaced for squares. That would be weird lol.

Good question for laughs at least.

  • Castling is already not allowed before the fourth move!
    – Michael
    Dec 2, 2019 at 1:33
  • 5
    Not allowed or not possible? I cant think of how it could possibly happen before the fourth move.
    – Joe
    Dec 2, 2019 at 20:24
  • 1
    @Joe There's no difference in this case; what's not possible within the rules is obviously against the rules.
    – TylerH
    Dec 3, 2019 at 22:09

My 2 cents - it's a very stupid idea, on a par with Nigel Short's suggestion to change stalemate into a win.

There are two very good reasons not to do it:

  1. It would completely change opening theory
  2. There are already 959 alternatives if you want to have a form of chess which is similar to the standard format but with different start positions which gives a lot more scope for innovation than just banning castling. There is now a world champion in this who is not called Carlsen.

In Short (pardon the pun), it is a solution to a problem which doesn't exist / already has a solution (959 solutions).

  • 3
    Doesn't (2) invalidate (1)?
    – Quintec
    Dec 1, 2019 at 1:35
  • 37
    Completely changing opening theory was the entire point of Kramnik's suggestion. You can argue that is unnecessary or a bad idea, but you cannot refute the idea by pointing out that it would accomplish what it is meant to do. Dec 1, 2019 at 2:04
  • 4
    Carlsen did take runner up in that tournament, for what it's worth.
    – corsiKa
    Dec 1, 2019 at 2:20
  • @AlexanderGarden: A variation where stalemate into a win would increase White's advantage, but might not be a bad idea if it included measures to equalize the game. One measure I'd like to see explored (even for otherwise-standard chess) would be specifying that after one player announces the first move for White, the other player gets to decide whether to play black, or take over the white pieces but be required to open with that move. So White's first move would be the one that most nearly balances the game going forward.
    – supercat
    Dec 3, 2019 at 16:28
  • I’d like to change checkmate into stalemate. That way can focus on the inherent beauty of the game without being distracted by this winning/losing stuff :-)
    – Laska
    Nov 19, 2020 at 4:23

Who cares? Let 1,000 flowers bloom.

There are already a bunch of popular variants so arguing from tradition is pointless at best. Does anyone play Chess 960 exclusively? King of the Hill? If they do, then those are valid creations of clearly high interest to them. If not exclusively, then still clearly fun for them. If not quite so, just eventually of passing interest, then they shall pass.

This too. It would either make it or it wouldn't. If some site programs it, and doesn't shoot for exclusivity, exposure might find lots of fans. I'm actually curious at the moment whether average players would be happier the "magic" move isn't there to upset their plans or if they'd be more dismayed that it isn't easy to hide the king and get the rook out.

Personal bet? It would unbalance the game enough to have the practical effect that folks cut the middle game rather short, choosing to trade off the moment anything goes wrong since attacks might be much more "blitzkrieg" than currently (excepting the good Mr. Fisher's legacy (which includes Chess 960, I believe)) and the risk tolerance might be fairly low when it's not just "now I have a sucky game and have to fight hard for a draw" but rather "geez, nowadays an attack like that might be two moves away from slaughtering me where I used to have five, six moves to wish I could trade down."

But who knows? Wonder what was said when letting the pawn move two spaces on its first move or when castling was first introduced?

Baseball used to go four strikes and fouls counted the whole way, not just to one strike short of striking out. Anyone not build a kitty of money for someone who lands on Free Parking? Not truly identical, but tweaks that worked out.

Let it exist, see if it takes over. If so, maybe it should have. If not, what got hurt?


As a chess composer myself, it would be extremely upsetting. Over two centuries, give or take, worth of wonderful art would be slammed into dust, now entirely useless somewhat. What's to do with it all now if castling is banned? Banning castling would be like burning down the Library of Alexandria-hudrends of years of work would be lost for many people.

  • 16
    Isn't your argument like saying that us having replaced "thou art" with "you are" makes all classical English literature entirely useless? We can still enjoy Shakespeare; we just need to remember that he used a slightly different language than today's standard one. Dec 1, 2019 at 8:44
  • 6
    While I appreciate the previous comment (it’s such an imaginative comparison), I still have to question it’s validity. A chess composition often involves casting in a critical way (the aha moment), whereas “you are” vs. “thou art” does not invalidate crucial plot twists in Shakespeare’s writing. Dec 1, 2019 at 15:24
  • 13
    @RebeccaJ.Stones There are some famous compositions where the aha moment relies on abusing some loophole that is patched by now (e.g. castling with a freshly promoted rook on the opposite rank or promoting into a piece of your opponent so their king cannot take it). We can still enjoy the clever ideas even though we are not allowed to replicate them anymore in a real game.
    – Annatar
    Dec 2, 2019 at 7:37
  • 1
    I love compositions too, but I think that this is primarily a matter for “Spikes” (serious players of the game, in useful Magic the Gathering terminology) rather than we “Johnnies” (creators of interesting artistic effects, in MtG lingo). The compositional world has long neglected to keep a historical timeline of occasional changes to rules or more frequently conventions, which would give insurance against the impact of such changes. There also needs to be imo a principle of maximal soundness, which means we can judge a composition by whatever time period best highlights its intention
    – Laska
    Dec 2, 2019 at 11:08
  • 5
    I cannot see how anything would be "lost". Castling chess and non-castling chess would simply change places (one being the mainstream variant and the other one being one -- one of the more popular -- of apparently 985 rule variations). The existing body of theory would not change at all. Dec 2, 2019 at 19:50

Thanks to everyone for their replies. As the original questioner, I've deliberately come late to this party.

I think it's instructive (and fun) to look at GM Kramnik's arguments from the perspective of the collectable card game Magic the Gathering. Think what you may about Magic, its developers are the absolute experts in rule changes, since they've been doing it continually since 1993. How might an imaginary Magic Developer respond to Kramnik's (paraphrased) arguments?

Kramnik: (1) There are two related problems: too many draws and opening theory carries on too long into the game.

Imaginary Magic developer: "YES absolutely. We try in Magic to avoid draws and drawish positions, which are anathema to player fun. The too lengthy opening theory in chess shows clearly that the format is too nearly "solved" and needs refreshing. You poor guys!"

Kramnik: (2) Chess960 won't interest spectators, won't interest some players, and is not "aesthetic".

Imaginary Magic developer: "NO WAY. Fischer's genius invention of Chess960 is the nearest that chess has to the concept of "printing new Magic cards". It's a great way to expand the game and make it fresh for players and spectators. The recent World Championships in this format was a resounding success, with commentators on the edge of their seats right from the beginning of the game. It must be very difficult for professional chess players, who have invested their lives and bet their economic future on a particular form, to contemplate their main asset being devalued by Chess960, when there are always newcomers and computers to adapt to the new formats quicker. This is really hard, but I don't think you can use this point anecdotally to argue against Chess960. And I think you confuse "aesthetic" with "familiar".

"We Magic developers are sensitive to the views of the pro player community, because we understand that where they lead the rest of the community can follow. But we can't let a key format decay indefinitely."

Kramnik: (3) Chess would benefit at top levels from banning castling .

Imaginary Magic developer: "This is a very interesting proposal. We are alert to the playability of formats, and will sometimes step in to ban powerful cards if the format is not healthy. We particularly focus on the balance between attack and defence, and here arguably the classic chess format is suffering, due the capability of computers and players to find better defences against (sacrificial) attacks. Castling is a powerful card, a kind of "Black Lotus", offering "three tempi for one". Castling also resembles the pattern of recent conditional Magic cards, in that there are some preconditions that you need to satisfy, but those are the kind of things that you would want to do anyway (develop minor pieces, avoid checks). So the preconditions aren't really a downside. Only a player of Carlsen's calibre has been able to reliably eke wins out of drawish positions: which is what needs to be done if romantic-style attacks cannot be successful. So here we do need to shift the balance from control to aggro."

"Removing castling at the top levels of chess is an elegant proposal, and may help fix draws. But we don't want to separate the game the Pros play from the game that our paying public plays. And banning one kind of move can only refresh chess opening theory as a short-term side effect. Ultimately chess need to "print new cards" with Chess960. This is the real future of chess."

"Castling by the way sits uncomfortably in Chess960 anyway (even top GMs don't know how touch move works with it) and is arguably unnecessary. We Magic developers are alert to fussy rules and would try to get rid of castling in Chess960. Since Chess960 is new, we don't lose any history by so doing. Of course, castling breaks left-right symmetry, so there would be only 480 effectively distinct positions.”

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Not the answer you're looking for? Browse other questions tagged or ask your own question.