[Retros] shortest proofgames of castling switchback

[Noam Elkies]

"Mario Richter" <mri_two at t-online.de> writes:


> Joost wrote:



>> C+ for the first 17 moves. But white just has too many free moves,

>> I don't think it can be fully computertested.



> It can be done (takes btw less than 3 seconds computer time - but

> in the approach used, some minutes have to be added for

> preparing the computer input ), but not with the current

> versions of "Natch" and Co.



> And yes - the verdict of my program was "Found one solution".


Whew.  Thanks!  Which program did you use?

So this row of H.Onoda's table is no longer "vacant".  Still, perhaps
I gave up too late on the 19.0 move version: with 7 Bf4 no longer used
for Bh2,g3 etc.  we can try to put this Bishop on g5 to stop ...Kf6
(and while I'm at it also to help restrict the bK's wife):

  1 a4 g6  2 Ra3 Bg7  3 Rh3 Be5  4 d3 Nf6  5 Qd2 0-0  6 Qh6 Kh8
  7 Bg5 Rg8  8 Qf8 b6  9 Rh6 Ba6  10 h4 Qc8  11 Rh3 Qb7  12 Re3 Qf3
  13 Nh3 Qh5  14 g4 Bb5  15 Bg2 Na6  16 Qb8 Kg7  17 Bb7 Rh8  18 Bc8 Kf8
  19 b3 Ke8

  r Q B . k . . r  hg-19.0
  p . p p p p . p
  n p . . . n p R
  . b . . b . B q
  P . . . . . P P
  . P . P R . . N
  . . P . P P . .
  . N . . K . . .  16+16

  NDE 18.vii.2007

I also retain the option of changing 19 b3 to 19 Kf1.

--Noam D. Elkies