i think it would be worth it to sac (although i'd be inclined to sac the exchange rather than a whole piece) and take away your opponents castling rights if you get a more long lasting advantage(eg.useless bad bishop), or a really quick attack.
4...Bxf2?? wasn't a great sac, if white developed calmly, eg. Kg1, Nc3, d3, Be3 instead of grabbing a pawn 7.Nxe5?? Qd4+ -+, then you'd be struggling.
about a pawn, generally, to deprive the opponent of the privilege to castle.
If this idea were anything close to correct, you'd see masters playing that sac, and you don't. In the example game given, if instead of the 7.Nxe5 blunder, White played 7.Kg1... White has castled "by hand," and Black is down a B for a pawn with negligible positional compensation.
Gonnosuke:
"Once castling has been prevented, it requires 2-3 tempi to reposition the King and Rook vs the single tempo it would have cost to castle. A tempo is worth about 1/3 of a pawn so I think a good estimated value of the right to castle would be somewhere between 2/3's and 1 full pawn"
Your analysis can't be right (with all due respect) as why would anybody bother castling by hand if its value was equal to the number of moves it took to do it. That would mean there was no net gain to castling by hand (once you lost the ability to do it normally.)
There is no way castling is as important as a minor piece. If you lose castling rights and your opponent doesn't, your king isn't necessarily endangered. Besides, I'd take an extra bishop over castling rights in almost any situation - if all else failed, you could block checks with the bishop. Nonetheless, trading queens or removing your opponent's queen removes much of the need to castle, so if this is possible after castling rights have been lost, I don't feel too disadvantaged in a game.
#8 -
If using the Game Explorer there were a way to compare the winning percentage of players who do not castle and players losing a minor piece that would say something.
interesting. Seeing as i rarely am on either side of this minor-piece-for-castling exchange, i can't really comment as to its effectiveness. i believe that, fundamentally, the sacrifice is unsound unless there is a positional/initiative gain.
Gonnosuke (#8):
I see I was wrong now, as your simple point was that whatever position can be achieved by castling can be achieved by a set number of additional moves if the ability to castle is lost, and there is an accepted point value for a move.
However I think I would take exception to that value of 1/3 point per move. There are roughly 30 moves expended in a win on average. The total point value of all pieces (excluding king) is 41. However, not all pieces are actually utilized in a win, and the total point value of pieces actually utilized has to be closer to 30 it would seem. So if you have roughly 30 moves per game and roughly 30 points of piece value utilized per game it would be tempting to say the value of a move is closer to 1 point. It seems you cannot divorce the value of a piece from its ability to move.
However, the point value of moves doesn't seem to tell the whole story either. For example, if your opponent were to get a three move head start in a game, the value of that would seem to be far in excess of 1 single pawn. Just the winning percentage of white with a one move head start (roughly 39.5 for white 32.5 for black) would seem to indicate the value of a move is much higher than 1/3 of a pawn.
Just thinking out loud here.
Gonnosuke:
Now I am going to go back partially to my original position. IF the opponent is able to subsequently castle by hand, then the value of the sacrifice is determined by the number of additional moves to castle imposed on him. OTOH, he may never gain the ability to castle again. In that case the number of additional moves has no bearing - the question is merely how advantageous is it on average to have your king on the outside and your rook positioned on the center of the board. (And I realize my analysis is still somewhat subjective however.)
-----------------
edit:
So the question is - how often is a player able to castle by hand later on after losing the ability to castle, and my guess would be not that often. It seems he may flounder around with several additional moves to only achieve some rough approximation of a castled position.
Eberulf: Interestingly enough, I think Bronstein once said 'a tempo was worth a pawn'. But I think it may have been implied: A tempo is worth a pawn To BRONSTEIN~! 
For the rest of us, i have to agree with Gonnosuke as what he says about a tempo worth about 1/3 pawn is pretty much the universally accepted rate. Note he said "about 1/3" as well, not exactly. Which is pretty much my own experience when you consider a Gambit Opening for instance. Therein, I've found that to gambit a pawn for two tempo alone is usually not enough, but you also get a half open file when you do that and if it is a very useful one, it can be worth another "half-tempo" 
and fully compensate for the pawn. But if the gambiter gets THREE tempi for the pawn, he should have advantage. In general, imo.
About that Question on "Why hand castle if it costs a pawns worth of tempi to do it"? Remembering that a piece was sac'd to prevent castling!... if he can hand castle to safety at cost of a pawns worth of tempi, plus the pawn the piece took, he is still winning materially by at least a point. Granted "points" often mean nothing, and "position" is always key. So it's a generality. But I can tell you, in my last 10,000 blitz games, I really only remember losing about two vs such a piece sac vs the KBP, and winning dozens. On the otherhand, if you can sac a piece that forces the King out into the field at the same time, say his third rank or beyond, I'd certainly take a strong look at it. Example:
1.e4 Nf6 2.Bc4 Nxe4 3.Bxf7+ Kxf7 4.Qh5+ (threatening 5.Qd5+ to recoup the piece) if BL there tries to walk the K out [Ke6 or Kf6] to hold the piece with it, I believe he gets mated trying... (as I remember it working out ages ago) Or at least wins back his piece, regardless, with the added danger for BL.
{ I thought you might find this an interesting sac to try out, since you like this kind of play? 
But if so, I just add the warning that there IS a line where BL comes out better by ...g6 and giving up the N right away, tho material is even in it. But that is why I don't play it in anything but blitz. I think the good BL play is from MCO-12, but rather hard to find during a game otb.}
It seems you cannot divorce the value of a piece from its ability to move.
I hadn't thought of this before. This brings up an interesting idea: "A piece is not worth its full point value until it has been developed." For example, if you're playing white and your c1 bishop is blocked in by the b2 and d2 pawns, it's certainly not worth three points. I'd daresay that, AT THAT POINT IN THE GAME, your central pawn has comparable value. Once the bishop has been developed, it is, of course, worth more than that pawn. But a blocked-in bishop is essentially useless.
This is *not* a technical response, but I'd be happy to get the bishop and the knight. For low level pleasure players (myself included), removal of two of my opponent's power pieces is equivalent to a castle. Yes, it takes a bit of mucking around to protect the king, but piece advantage is a great tool. Also, although the movements of a bishop can be replicated by the queen, once a knight is gone it is irreplacable. (Unless you manage to promote a pawn.)
In situations where a trade results in white's inability to castle, I'd say go for it.
Personally, if I get a chance to trade even pieces and take away my opponent's ability to castle, I'l take almost every time.
Gonnosuke, you probably mean "GM" Kaufman. Winning that one tournament doesn't change the player that he is.
Gonnosuke, you probably mean "GM" Kaufman. Winning that one tournament doesn't change the player that he is.
*ouch*
Isn't that a little harsh? I think Chess Life said he had actually increased his performance rating by about 100 points since working with the Rybka project. I think that put him in GM territory as well as winning the World Senior CH.
A bit too harsh, perhaps you're right.
How many norms has he earned over his career?
Some interesting points. I'm not sure blocking me from castling would work well, as I sometimes play a whole game without castling, and for me, it is not a must. Against someone who uses castling as a sole part of his strategy, it is useful. I think it all depends on the other player, like for all.
There isn't a single instance when I have ever regretted taking away the opponent's ability to castle by sacrificing a minor piece in trade for the opponent's king's bishop pawn. Not just in the Fried-Liver proper, but in any conceivable circumstance, playing as white or black or whenever. It clearly is a highly beneficial sacrifice always. It would lead me to say that the ability to castle has to be worth at least slightly more than a minor piece. So, you're giving up two points on paper to take away his ability to castle, and I could never say in retrospect that doing so was merely a wash strategecally in the course of the game. It was always advantageous. I would say denying the ability to castle is worth more than one minor piece because two pieces are impacted - one a major piece (the rook) and the other probably the most prized piece of all. To deny effective mobility to a major piece for a substantial part of the game and to make the other crucial piece permentantly vulnerable I would say has to be worth at least 4 points. (Just my opinion.)
In your game, White only needed 2 moves to castle by hand. Normal castling requires 1 move. That means your piece sac only delayed him by 1 move -- how can that one single move be enough compensation? In order for that sac to sufficiently compensate, you'll have to assume that a single move is worth 2 points! (the amount you lost when you sacrificed a piece for a pawn).
I would happily give up a move if I can gain a piece for a pawn.
There isn't a single instance when I have ever regretted taking away the opponent's ability to castle by sacrificing a minor piece in trade for the opponent's king's bishop pawn. Not just in the Fried-Liver proper, but in any conceivable circumstance, playing as white or black or whenever. It clearly is a highly beneficial sacrifice always. It would lead me to say that the ability to castle has to be worth at least slightly more than a minor piece. So, you're giving up two points on paper to take away his ability to castle, and I could never say in retrospect that doing so was merely a wash strategecally in the course of the game. It was always advantageous. I would say denying the ability to castle is worth more than one minor piece because two pieces are impacted - one a major piece (the rook) and the other probably the most prized piece of all. To deny effective mobility to a major piece for a substantial part of the game and to make the other crucial piece permentantly vulnerable I would say has to be worth at least 4 points. (Just my opinion.)