|Mar-15-06|| ||Resignation Trap: If you are baffled about 22...Kf8, the explanation is simple! Torre was walking around the tournament room while Reti was thinking about his 22nd move. When he returned to the board, he discovered that Reti played 22...0-0-0! This is illegal, and at the time the rules stated that Black had to move the first piece that he touched. Reti couldn't remember whether he touched his rook or king first, so it was ruled that he had to make a king move.|
|Mar-15-06|| ||aw1988: Illegal?? It's perfectly legal...|
|Mar-15-06|| ||Ziggurat: <Illegal?? It's perfectly legal...> No. The a8 rook has moved.|
|Mar-15-06|| ||Eric Schiller: <ziggurat> I did that once in an international tournament, and I'm an International Arbiter! Once, as arbiter, I was asked by Viktor Korchnoi if he could borrow his opponent's scoresheet. His opponent, Waitzkin, was in time pressure so I said no, but Viktor didn't explain why he wanted it. It turned out that he couldn't read his own scoresheet (he shares that lack of skill with 99% of players and arbiters, Korchnoi's scoresheets are infamous for being illegible) and wanted to see if he had moved his rook. He wasn't sure that he could still castle.|
Returning a rook to its original square with the king still on the home square doesn't happen much, and it is very easy to forget about it during the game.
The programs like Chess Base can't even enter the game when it contains illegal castling, so examples are not easy to find in databases.
|Mar-15-06|| ||Jim Bartle: I once saw a game where Korchnoi castled after having moved his rook (or maybe his king), and neither player noticed it and the game went on, can't remember the result.|
|Mar-16-06|| ||keypusher: <Jim Bartle> The story I remember about Korchnoi is that he had to ask the arbiter whether he could castle when his rook was under attack:|
Korchnoi vs Karpov, 1974
Having been assured that he could, Korchnoi played 18 0-0!, avoiding 18 cd?? Nf3+
|Mar-16-06|| ||Jim Bartle: The game I remember is from the '90s.|
|Mar-16-06|| ||perfidious: <keypusher> back in 1984, I played a game in a then-popular line of the Benko with Patrick Wolff, in which he and Jim Rizzitano had prepared a line culminating with ....Nxf2, figuring I'd have to move the rook on h1, concluding that Black stood well...|
There was one small snag-I castled short, with a large advantage; only poor play by me later enabled Black to save a half-point.
|Jun-09-09|| ||WhiteRook48: 22...0-0-0 lol forgot the a8-rook moved|
|Apr-24-11|| ||Ozo Shatranj: uh, bad luck another one told the story about the reti king's move, i wanted to be the first! ;)|
|Feb-04-17|| ||zanzibar: According to Soltis, <Chess Lists-2e p26>|
<Postmortem analysis indicated [...] Black had only two good moves. Remarkably enough they were 22...O-O-O (if legal) and 22...Kf8!>
This isn't true, although 22...O-O-O would be the engine best move if allowed. 22...Kf8 is quite playable, but isn't a leading candidate move (it's #3 till about 22-ply or so in sf8).
But Black could play a variety of moves, e.g. 22...Qb6 is ok, etc.