Exeter Chess Club: Trawled from the 'Net

Newsgroups: rec.games.chess
Subject: Re: How to know when a sacrifice is lucrative?
Summary: 
Expires: 
References: <32npl2$i3p@senator-bedfellow.MIT.EDU>
Sender: 
Followup-To: 
Distribution: 
Organization: University of Exeter, UK.
Keywords: 

In article <32npl2$i3p@senator-bedfellow.MIT.EDU> 
richard@genome.wi.mit.edu (Richard Resnick) writes:
>Above and beyond being able to look very very deep, are there are
>guidelines that can be stated regarding the sacrifice of minor pieces?
>It seems to me that with all 32 pieces on the board, it becomes very
>difficult to be accurate when looking ahead, so I assume that the
>masters must have a set of guidelines that they use when determining
>whether to go down in material for the tactical advantage. Any help
>(either commentary or literature citations) would be greatly
>appreciated. Thanks very much.

:-)

In "How to Cheat at Chess" IM Hartston gives Hartston's Iconoclastic
Combinational Uncertainty Principle(?), which states (I am
reconstructing from memory, rather than quoting) that:

 for a given sacrificed piece value (S), the expected value (V) to be
gained from a sacrifice can be worked out, since we know,

 firstly, the hoped-for gain (H) and
 the number of moves deep the variations are (N).  

The motivation for the sacrifice is H-S, but this is tempered by the
increasing probability of miscalculation as we go deeper in analysis -
in fact, the probability increases as the square of the move depth. 

So, V =     H  -  S
           ___

             2
            N

Applying HICUP then, if you sacrifice a pawn (S=1) hoping to gain a
queen (H=9) in two moves (N=2), then V= 9/4 - 1 = 2.3 - 1 = 1.3; V is
positive, so the sacrifice is worth a go.  However, sacrificing a rook
hoping to gain a knight in three moves gives S=5, H=3, N=3 and so V =
3/9-5 = -3.7.  V is strongly negative, that is, you have probably
miscalculated and should give it a miss.

---

Back to Chess Coaching Page



This document (.html) was last modified on 
by

Dr. Dave