No of moves before castling is a misnomer. Here

Kasparov vs Ivanchuk, 1995.

I was lured by 26. Qh6+, <Once>, but without 27. Qh4+ - there's a limit to how much one can resist the 27. Bxd4+ temptation: Such a tasty Knight, IN the center, WITH check...

Chigorin and donkey knights as so often with a dilly it faced in black

spent too long dithering in the opening although 23...nxd4 settles it

mate inch 7/6. In forget he cash up queen it hope in thinkin it abouts

my bibelot shake drat tell a nd4 rolled again rxh7 off manage in low

go down it was in would kingg8 hold? In slug it rookxh7+ alive in dead

jab vocal slow into fun creeps qh1+ ah good black stale in performance

a. 26...Kh7/8 27. Bg5+ Kg7 28. Qh6#

b. 26...Kf6 27. Qh4+

b.1 27...g5 28. Qxg5#

b.2 27...Kf5 28. Qf4#

White to play (24.?) "Easy"

Black must have been expecting 24. Bxd4+ Rf6 25. Bxf6+ Qxf6, with an even game. Of course, White has much better, which I'm sure he didn't miss!

24. Rxh7+! Kxh7 25. Qh1+ Kg7

Tossing in the bishop with 25...Bh3 doesn't change the outcome.

26. Bh6+!

Better than 26. Qh6+.

26...Kf6

Or 26...Kh8(h7) 27. Bxf8+ Bh3 28. Rxh3#.

27. Qh4+ Kd4

Alternatives are 27...g5 28. Qxg5# and 27...Kf5 28. Qf5#.

28. Qxd4+ Kf5 29. Qf4#.

24.Rxh7+ Kxh7 25.Qh1+ Kg7 26.Bh6+

A) 26...Kh7/Kh8 27.Bxf8+ Bh3 28.Qxh3#

B) 26...Kf6 27.Qh4+

B.1) 27...Ke5 28.Qxd4+ Kf5 29.Qf4#

B.2) 27...Kf5 28.Qf4#

A famous mating combination from Steinitz, the great World Champion.

<24. Rxh7+! Kxh7 25. Qh1+ Kg7>

(25...Bh3 26. Qxh3+ and the combination continues as per the main line)

<26. Bh6+!!>

The most accurate - Mate is now thoroughly forced.

<26...Kf6>

(26...Kh8/...Kh7 27. Bxf8+ Bh3 28. Qxh3#)

<27. Qh4+ Ke5>

(27...g5 28. Qxg5#; 27...Kf5 28. Qf4#)

<28. Qxd4+ Kf5 29. Qf4#>

Réti annotates this game in 'Modern Ideas in Chess.'

On White's 4th move (d3), he writes:

'One sees here at once the difference between Morphy and Steinitz. The former was always anxious to press on at the earliest possible moment with d2-d4. Steinitz on the other hand does not want to break through the center, but is more concerned with building up for himself a strong position, to enable him subequently to prepare an attack on the kingside.'

On 5. c3, Réti comments:

'The position of the pawns on c3 and e4, which makes the forcing of the center by the black pieces impossible, runs with regularity through the Steinitz games whereever they are opened with e4."

And on 6. Nbd2:

'With the intention of moving the knight (by way of d2 and f1) to e3 or g3 to carry out the attack. This maneuver, so much in favor today, originates from Steinitz. As a fact we find very often in Steinitz's games these extended knight maneuvers.

With Morphy, who always brought about an open game, that kind of maneuver was impossible, as he dared not permit himself in open positions to lose so much time. Noteworthy and typical of Steinitz is the delay in castling: so that the possibility of castling on the queenside remained open to him.'

On 8. Ba4:

'In order to have this bishop ready for the attack. These are all far-reaching and preparatory maneuvers for which in open positions after d2-d4 there would be no time.'

On 8. ... Nd7:

'With the idea of making the game an open one if possible by means of ... Nc5 and ... d6-d5.'

On 11. h4:

'Now at this early stage the attack on the king's wing commences and indeed, clearly contrary to Morphy's principles, from an undeveloped position. But the essential point is that Black's counterplay against White's center does not lend itself to a successful result on account of the latter's assured position.

Equally remarkable is that the move h4 is not to be found in analogous games of Morphy, the reason being that Morphy unlike Steinitz always castled early in the game.'

On 13. ... fxg6:

'Perhaps 13. ... hxg6 was better. Steinitz would have continued with 14. Qe2 in order to avoid the exchange of queens, as one will find happens in similar positions with other players; at the same time the strong pawn structure formed by the pawns at e4 and c3 would have been maintained and Black would have gained little by the opening of the d-file, as no points of attack are to be found thereon. After the weakening of the diagonal a2-g8 through ... fxg6, Steinitz opens the diagonal completely by the exchange on d5.'

On 20. Qf1:

'Apparently a defensive move to provide against ... Nd4. In reality preparation for the decision of the contest.'

On 21. d4:

'This ultimately brings the other bishop on the right diagonal a1-h8 for the decisive mating attack.'

On 24. Rxh7+:

'After the deep and quiet preparation the end is brought about magnificently, inasmuch as the whole of the pent-up energy becomes active.'

Steinitz vs Blackburne, 1876

After 6. Nbd2
<Perhaps this move shows masters have lost as well as gained in the last 60 years. We can recall few master games since 1945 in which white has castled QR in the Ruy Lopez. The one outstanding example is in the Steinitz defence Deferred: 1. e4 e5 2. Nf3 Nc6 3. Bb5 a6 4. Ba4 d6 5. Bxc6 bxc6 6. d4 f6 7. Be3 Ne7 8. Nc3 Ng6 9. Qd2 Be7 10. o-o-o> "A Guide to Chess Openings" by Leonard Barden (Routledge and Kegan Paul, 1957.)

Long castles in the Lopez is thus very variation specific. Another example is in the Schliemann (3...f5 where the 4. Nc3 fxe4 5. Nxe4 d5 line often sees both colours castle q-side).