15 May 2013

Misleading Review

Flaws in a Review of Chess Quest

Chess Quest has been among my principal sources for tactics training through the past three months. I have solved nearly half of its 1200 problems, including all of levels 1 and 2, and slightly more than half of level 3. The problems have ranged from simple thematic checkmates that build pattern recognition to clever defensive resources that maintain drawing possibilities in slightly worse positions. The problem selection includes a reasonable number of essential endgame positions, while maintaining a focus on tactical ideas in middlegame positions.

Not everyone who has tested Chess Quest shares my enthusiasm for the iPad/iPhone app. C. K. Sample III, for example, identifies "flaws" in some of the problems. Of the twenty-four problems he solved in the basic level before writing his review, he found two where "no checkmate was achieved and [he] can see multiple moves that can be taken by the other player to effectively recoup the game" ("Review: Chess Quest," Sample the Web, 12 September 2009). He offers screenshots of problem 24 to illustrate this "flaw". Problem 17 appears to be the other that he has in mind (see screenshot). Sample concludes that he cannot recommend the app because of these two problems with incorrect solutions, and notes: "If it keeps that 1 in 12 ratio, that means that rather than 1200 puzzles for $2.99,* you’re only really getting 1100 puzzles and 100 duds."
Problem 17, Basic Level

Had he continued solving in levels above basic, Sample would have found a far higher ratio of problems that do not end in checkmate. This ratio should not come as a surprise for Sample's "flaw" is the norm in tactics training. A significant percentage of the problems in Fred Reinfeld's classic 1001 Winning Chess Sacrifices and Combinations (1955) do not end in checkmate. That proportion is even greater in the modern Anthology of Chess Combinations, 3rd edition (2005). Sample reveals not so much problems with the Chess Quest app, as his own qualifications for assessing tactics training resources. His review resembles in one aspect the many five star reviews of the Social Chess app by users whose only point of comparison is Chess with Friends (see "Social Chess iPad/iPhone App: Review"). Such reviews are misleading and offer disservice to readers.

Beginning Tactics: Instruction

C. K. Sample's criticism reveals that there is a market for developers of iPad/iPhone apps aimed at beginning chess players. While collections of carefully selected instructive tactics positions serve those serious about tactics training, elementary instruction is needed, too. Chessimo is one such app (see "Chess Tactics Training on the iPad"). Most Chessimo problems end in checkmate, at least among the 3000+ that I have solved. Nonetheless, that feature alone does not satisfy the need of explaining what constitutes a decisive advantage.

As chess players grow in strength, they learn an abundance of patterns and techniques. They begin to recognize an advantage. They develop the skills to convert an advantage into victory. It was clear to me that winning the bishop was a sufficient solution in problem 24 in the basic level of Chess Quest. But, Sample needs more explanation. This need prompted him to give Chess Quest a misleading negative review.

Sample correctly identifies one element that could be incorporated into tactics training apps: a chess engine. In order to demonstrate that problem 24 in Chess Quest's basic level is not flawed, I played the position against HIARCS on my iPad.

Checkmating the engine took me 42 moves from the problem position. Along the way, there were several positions where learned themes from prior training rendered my moves nearly automatic. The entire battle required less than ten minutes. A stronger player could have concluded the battle more rapidly, and in fewer moves.

White to move

James Stripes -- HIARCS


Chess Quest ends the problem after this move has been found. Pinning the bishop wins it.


Black avoids exchanging queens, and makes two threats. The vulnerable a-pawn is attacked, and the queen threatens to fork White's bishop and king with 2...Qe1+.

2.Qxb7+ Kf6 3.Qc6 Qxa3 4.Qc3+ Kf7 5.Qc7+

Black to move


It seems to me that Black could protract the struggle with 5...Kf6. Nonetheless, the engine considers its move best unless it thinks longer that was permitted in a five-minute game.

6.Qxe7+ Kxe7 7.Kg1

Black will have a passed pawn on the a-file that must be stopped in addition to the isolated and passed e-pawn. The bishop cannot stop both. As the most powerful piece left on the board, White's king must become active.

7...Kd6 8.Kf2

8.b4 is better, fixing the a-pawn.

8...a5 9.Ke3 Kc5 10.Kd3 a4 11.bxa4 bxa4

White to move

At the heart of this position is an elementary idea that Jeremy Silman calls "the fox in the chicken coop."


12.Bg6 would have reduced Black's counterplay.

12...g5 13.Kb2 Kd4 14.Bf3 e5 15.Ka3 Kc3 16.Be4 Kd4 17.Bd3 e4

After a few suboptimal moves, I found myself in a position where only one move maintains the win. Fortunately, this one move had been central to my plan leading up to this position. I meant to sacrifice my bishop to eliminate Black's last viable threat.

White to move

18.Bxe4! 18...Kxe4 19.Kxa4

Material equality had been restored for one-half move, but the pawn ending is an elementary win for White. If Black goes after the c-pawn, the White king will get Black's remaining pawns. The alternative, chosen by HIARCS, is to temporarily go a pawn ahead but let the c-pawn promote.

19...Ke5 20.Kb4

20.c4! is better.

20...h5 21.Kc4 Kf4 22.Kd4 Kg3

White to move

Once again, White has only one winning move, but it is the move that has been prepared. This pawn race has been carefully calculated.

23.c4 Kxg2 24.c5 Kxh3 25.c6 h4 26.c7 Kg3 27.c8Q h3 28.Ke3 h2

White to move


29.Qf5 is objectively superior, but White opts for a standard technique to prevent promotion of the h-pawn. It is worth noting that Chess Quest or Chessimo (perhaps both) offer positions solved with the technique evident in the continuation 29.Qf5 h1Q 30.Qxg5+ Kh3 31.Qh5+ Kg2 32.Qg4+ Kh2 33.Kf2!

29...Kg2 30.Qd2+ Kh3 31.Qe1 Kg4

White to move


32.Ke4 is much better, leading to checkmate in three fewer moves.

32...Kf5 33.Qxh2 Kf6 34.Ke4 Ke6 35.Qh5 g4 36.Qxg4+ Kd6

White to move

One hopes that Sample would not consider a problem flawed if it ended here. And yet, beginners need problems that begin here.


37.Kd4 is the other optimal move. I teach young players to place the queen a knight's throw from the defending king and deliver no check until checkmate. Hence, I exercise that technique here.

37...Ke7 38.Ke5 Kf7 39.Kf5 Kg7 40.Qe8 Kh7 41.Kf6 Kh6 42.Qh8# 1–0

*The current price is $4.99 "on sale for a limited time."

No comments:

Post a Comment