Intermediate Level: Reading and Writing: Application example
A few days ago, the following puzzle–like position was posted in the forums:
Below it, the following text:
@pfren: "The positions are never evaluated properly taking into account just the material differences, and in many cases engine evaluations are not terribly relevant, too.
How would you evaluate this one, and why? White is two pawns down, and Black has the move."
As explained, we use a looping process of evaluation and analysis, to read and write on the board.
Now, keep in mind the following idea: We evaluate a position not to take a picture, but to modify its balance between both sides. Ideally, to our advantage, but when this isn't possible, to not to fall in disadvantage.
To do either, we make plans: From the position's demands (inner logic) we estimate (deduction: logic) where our pieces should be (piece activity), and calculate (precise calculation) the tempos (time), threats (piece activity), squares (space) and mobility (space and time) required to achieve our goals, and its possibilities of success (evaluation based on logic, precise calculation, knowledge and intuition).
A piece of knowledge before you attempt to solve the position above. A good plan seeks to modify the balance of your material, time, space and piece activity, reducing the opponent's piece activity, space, time and material. Practice shows that (generally), when a player is deprived from activity, he will later be unable to hold his space, will lose mobility and, eventually, material as well. Besides, tactics (a sudden increase in piece activity) can make these elements interchangeable with each other.
Now, take your time trying to solve the position above. Ah, don't use an engine, engines aren't good in this kind of positions, you can't use one in OTB games, and won't learn as much from this exercise.
The following is the solution I posted:
The position in #14 is troublesome for Black (the pawns deficit has no real meaning in this position, because they can't be used to oppose White's plans).
The light squares weaknesses around Black's castle, e6 and c5, plus the lack of control of c4, e5 and g6 are the main source of problems for Black. Between White's possible plans, Nf3–e5, Bd3–c4 and Qc2–g6 draws the attention, as Ne5–g4 (hitting e6 and h6) may be hard to meet.
Such an active plan should be opposed by Black's piece activity, but with a caged Bb7 and heavy pieces blocked by the Nd5, Black has no activity whatsoever.
Let's examine passive defense means. Retreating Black's Knight exposes the e6 pawn to additional pressure from Bd3–c4, and doesn't really solve a thing. With three pieces dancing around Black's King (White's Q, N and B), Black needs two defenders plus the King to maintain the balance over any square, and that's problematic in this position. Furthermore, when White's R can join the ensemble. Then, passive defense doesn't seem to have any bright future –if any–, as both Black's B and R are limited in their mobility and activity.
(The following board may make it easier to follow the lines, as this is the check–up part using –mainly– precise calculation:)
In the R's case, its mobility is reduced from the need to protect the 8th rank (Qc2–g6–e8 may be a problem), but in the B's case, it's a matter of pawn structure, and the B is doing nothing at the moment. Even in a blitz game, I'd play 1...c4! 2.Bxc4 c5 immediately (modifying the pawn structure to increase the B's mobility and activity), as for the price of a pawn the B returns into play hitting everything in a8–h1, and by getting control of a6 the move Bb7–a6 becomes available, indirectly helping in the defense of Black's castle as well as e6.
If White attempts to continue with his original plan, based on 3.Ne5, then 3...Nf6 4.Qg6 Rf8, and Black's ready for Bb7–d5 with good play. The direct attempt against Black's lines comes with the tactical 3.Qg6, but with the Bb7 in play, it doesn't work well: 3...Ne7! (hitting the Q and threatening Ne7–c6–d4), when 4.Rxe6 Qe1+ 5.Re1+ Kh8 6.Qb6 (otherwise Qd1xf3 wins for Black) 6...Bxf3! 7.Rxd1 Rxd1+ 8.Bf1 Be2! 9.Qe6 (double attack) 9...Rxf1+ 10.Kg2 (10.Kh2, to avoid the light squares, meets 10...Bf3 11.g4 Rxf2 12.Kg3 Re2 saving the piece, and winning for Black) 10...Re1! 11.Kh2! (this wins a piece) 11...Ng8 12.Qe3, when Black can't defend against Qe3–d2, but can pick the b3 pawn with 12...Rb1 13.Qxe2 Rxb3. In the resulting ending, I don't see how Black can avoid White's ideas h3–h4 and g3–g4–g5, opening the "h" file and securing the draw.
Of course, White may do better not risking a thing and picking up the second pawn with 4.Qxe6+ Qxe6 5.Bxe6+ Kh7 6.Ne5, trying to make play against Black's c5. But with the checkmate threat against White's King (Rd8–d1–h1) I don't see White with enough tempos to both defend against it (to free his R) and coordinate against Black's c5. I'd continue 6...Nc6, either forcing more trades (with a draw) or getting Nc6–d4 (with good play for Black).
And this was @pfren's final post on the subject:
On #14: White's initiative on the light squares is extremely annoying, and Black must be on the alert. Engines have no fear, and evaluate the position as completely level by various ways.
I think that Black wasted his saving resources at this very point by not playing 30...c4! 31.Bxc4 c5, when that fat piece of wood at b7 is starting to look like a piece. White has no trouble picking back the second pawn, but it's hard to avoid simplifications. I guess that Black should draw this with some mild trouble.
Black decided to trust the engine evaluations (it was an official correspondence game where engine usage is allowed) and keep the material by 30...Nf6. He never got the chance to play ...c4 under decent conditions, and was completely outplayed. I love this game, as I had to dismiss the engine several times, and plan my moves according to pure human understanding. The resulting play is very subtle- a lot of tempo winning/losing resources had to be used to achieve a winning position. The final position where Black resigned is picturesque: The queens are off the board, Black is still a pawn up, but he is totally paralyzed. Complete game analysis is still pending, because there are a lot of lines and ideas to consider, but I hope I find enough time to complete it. Here is the game, uncommented:
(By the way, the game alone, even without annotations, justifies the article; excellent game @pfren.)