Understanding Defensive Resources

Problem 147 in level 3 of Chess Quest gave me some trouble this morning. Spotting the sacrifice to open the h-file took me a few seconds, but after that the problem became more complicated. In retrospect, the ability of the knight on e5 to cover h8 from g6 should have been obvious.

White to move

The main solution in the Chess Quest app is 1.Nxh5 gxh5 2.Nd5

White's second move accomplishes several things: discovered attack on the troublesome knight, interference along the a8-h1 diagonal, and attack of the Black queen.

Chess Quest has two responses for Black.

a) 2...Ng4 is the main line

b) 2...exd5 is the alternate line

Neither of these moves are the top choices of the engines I employed in my effort to comprehend the problem after multiple tries were necessary to solve it on the iPad.

Houdini 1.5 played 2...Ne3.

White to move

I first tried 3.Nxc7 and lost my way after 3...Nxd1 4.Bd4 Ng6 5.Nxe8 Rc1 6.Bf1 e5 7.Bxb6 Rc3

White to move

Taking the queen maintains an advantage, but Black is able to complicate matters enough that my humanity reveals itself through errors in battle with the silicon beast.

I fared better against the machine with 3.Rxh5 Ng6 4.Rh6 e5 5.Nxc7 Rxc7 6.Rxg6+ fxg6 7.Qxe3

Black to move

With Black's pesky knights gone and the queen removed at the opportune time, White has a clear advantage and Black has limited counterplay.

Chess Quest's main line continues 2...Ng4 and White must stop the threatened fork. 3.Rf1 exd5.

Black has other moves that delay checkmate more moves. These are spite threats that give away major pieces. In terms of the problem themes--h-file penetration, and Black's efforts to control h8--3...exd5 is almost as good as anything else. Perhaps, however, 3...Ne5 might have been entered into the sequence by the creator of the problem.* White must then play 4.Bxe5 before proceeding along the h-file.

The main line concludes 4.Rxh5 Bg7 5.Rh7 and White is winning.

The alternate line tests the solver's understanding of the importantance of the knight on e5.

2...exd4 3.Bxe5 (3.Rxh5 Ng6 leads to a winning position for Black) 3...Rxe5 4.Rxh5 h-file penetration is now or never 4...Bg7 5.Rh7 Kf8

White to move

After 5...Kf8

 5...Kf8 is a subpar move. Black puts up more resistance with 5...Rxg5.

Unfortunately, Chess Quest rejects 6.Rh8+ in favor of 6.Rxg7, while both moves lead to checkmate on the following move. After 6.Rxg7 Ke8, Chess Quest accepts only 7.Qh8+. According to Stockfish 2.3.1, this move is the ninth best. 7.Rg8# is obviously best. 7.exd5 leads to the second fastest checkmate.

After 7.Qh8+ Chess Quest's line continues 7...Kd7 8.Rxf7 Ke6 and now White has three routes to checkmate in five moves. Chess Quest accepts only 9.Bh3+.

Chess Quest also has 8...Kc6 9.Rxc7+ (exd5+ is a better move, but not accepted by the app).

The problem is well-designed for highlighting defensive resources in a battle against an h-file battery, and for understanding that checkmating the king takes precedence over gaining material. Alas, the errors in the alternate solution confuse this message.



*Chess Quest "About" claims the GM Leonid Yudasin created the problems.

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
14.05.2013

1.Qb8

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

1...Qa5

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

5...Qe7

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.Kc3 

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.Qc1

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.Qh1

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.Qc8

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."

Training Log: April 2013

Fighting Addiction

My progress towards my training goals in April ran aground of my need to play. In particular, I went on blitz binge, decided to temper it with new discipline, and then played 320 bullet games over a twenty-seven hour period. I posted my training goals as New Year's Resolutions on the last day of 2012.

1. In 2013, I will solve correctly 300 tactics problems each month.

I met this goal, completing 329 problems in the month. In April, my time for tactics training was spent playing chess online. However, I spent enough time on training that I met my tactics target. I am working through Level 3 on Chess Quest, and have done a few problems in Level 4.

I ignored Chessimo until near the end of the month. Although I have solved 3269 problems on Chessimo, I count towards the month's total only those problems solved six times. In April, I completed 60 problems for the sixth time each.

My web-based training (Chess Tempo and Chess.com) remains light with one Chess.com session and a few short Chess Tempo sessions. I resumed use of the Anthology of Chess Combinations late in April after neglecting it in March.

2. In 2013, I will study whole games and whole books.

I looked through a few of the games of Wilhelm Steinitz for my scholastic lessons, and started going through his annotations of the games of others. Nonetheless, my study of whole games was swamped by my play of blitz, bullet, and a handful of fifteen minute games.

3. In 2013, I will finish my Pawn Endgame Flash Card project.

I am continuing to use these flash cards in elementary classrooms while teaching beginners chess. Nonetheless, progress studying Dvoretsky's Endgame Manual and mastering the blue diagram positions must be rated lackluster.

4. In 2013, I will lose fifteen pounds.

I remain at my winter weight due to inconsistent exercise and too many fast food meals.

A Twist on Smother Mate

This twist on the smother checkmate idea popped up in Level 3 on Chess Quest, an iPad tactic training resource.

Training Log: February 2013

I posted my chess improvement resolutions for 2013 on the last day of 2012. In February, I progressed well in my tactics objective, and made some progress on the other resolutions as well.

1. In 2013, I will solve correctly 300 tactics problems each month.

I met this goal in two weeks. Solving tactics every day made 300 correct an easy goal. I solved over 250 on Chess Quest, which I started using the second week of the month. See my review "Chess Tactics Training on the iPad." Beginning use of this app has slowed my use of Chessimo, where my average per problem is under ten seconds. Although I have solved over 2300 problems on Chessimo, I count towards the month's total only those probems solved six times.

Lev Alburt came off the desk, briefly. I solved six problems.

I finished my second run through the puzzles in Shredder's iPad in January, but my pace through the third cycle is too slow to meet: "the third cycle will be completed in 2013 with the goal of 80%+ score." I finished the 300 exercises in Chess-wise Pro, and reset the program for the second pass.

My web-based training (Chess Tempo and Chess.com) remains light. My pace through the Anthology of Chess Combinations is not a source of pride.

2. In 2013, I will study whole games and whole books.

I continue with a disappointing slow pace through Logical Chess: Move by Move, but it is not my sole source for whole games. After struggling with one combination in the Anthology, I spent some time studying the whole game: Polugaevsky -- Nezhmetdinov (1958). The combination was also featured in a YouTube video that I watched: "Nezhmetdinov Chess Biography, part 3."

I count my February progress on this second resolution adequate.

3. In 2013, I will finish my Pawn Endgame Flash Card project.

I am continuing to use these flash cards in elementary classrooms while teaching beginners chess, and some of the positions came up in the first endgame lesson that I completed in Chessimo. Nonetheless, progress studying Dvoretsky's Endgame Manual and mastering the blue diagram positions must be rated substandard.

4. In 2013, I will lose fifteen pounds.

Not at this rate, I will not. I have added Wii boxing back into my exercise regime. As the whether improves with winter becoming spring, and my wife's broken leg heals, my time on the Spokane River Centennial Trail with my puppies will increase and the weight will come off.

When I walk one puppy, the other sits at home and howls.

Chess Tactics Training on the iPad

Comparing Applications

Chess apps for the iPad and other iOS devices (iPhone, iPod, iPad Mini) proliferate. When I wrote my initial review of iPad chess apps, there were slightly more than a dozen. Now, it would take some effort to count all of the available programs in the Apple App Store. There might be dozens of applications specifically designed for chess tactics training. Other multipurpose apps offer training resources. Currently, I am using five apps in my tactics training: Chessimo, Chess Quest, Chess-wise Pro, Shredder, and Tactic Trainer. What are the principal features of each? What are the strengths and weaknesses of each? Read on.

Chessimo

Chessimo offers training in tactics, endgames, strategy, and openings. It also has a rating estimator offering play against Crafty, and can transform the iPad into a chess set with clocks. There are over 800 classic games included for playing through, and these are referenced in the opening training.

The rating estimator should not be taken seriously. I found that Crafty played typically absurd "weak" computer moves until my rating was near 1400, then became far tougher than players at that level. Nevertheless, I discovered a weakness, and by playing the same game repeatedly was able to push my rating over 2300.

For players just beginning tactics training, Chessimo is worth considering. It may also be useful for improving board vision for stronger players (I hope so). The app is free, and lets the user play through the first four units of the first training module. Each unit introduces thirty problems, and then repeats them, or repeats exercises from previous units. An in-app purchase ($2.99) opens the rest of the tactics modules (five total, consisting of 51 units each). For $7.99, the user may purchase all modules in tactics, endgames, commented endgames, strategies, and openings.

Chessimo's method is repetition. Each exercise is repeated six times. The tactics begin with one-move checkmates, then two-move. Eventually, checkmate threats are present, but such tactical motifs as interference must be employed for the solution. The exercises are timed. If the answer is not found before time runs out, the app executes the solution. The time per exercise can be modified by the user, and the solution can be turned off.

Chessimo has vastly more exercises than most other training apps, but the difficulty level of these may be of little benefit to advanced players. Even so, the large quantity of problems that gradually increase in difficulty drills essential patterns into the memory. At least some of the checkmate in two problems come from the same games as the checkmate in one.

The app tracks haw many problems I have solved, the average time for each set, and the total time.*

As with all automated chess training programs, alternate solutions are not always supported. Sometimes, the preferred solution is not the best (see "Endgame Training with Chessimo"). Chessimo works in landscape mode on the iPad and does not rotate.

GM Gilberto Milos selected the exercises.

Chess Quest

Chess Quest offers 1200 exercises in six levels (200 each). After the so-called "Basic Level" for beginners, the exercises in the Level 1 are comparable to the educational exercises in the Anthology of Chess Combinations. In fact, some of them are the same exercises. Exercises in level five are challenging for advanced players, perhaps even for masters.

An advantage of Chess Quest is that completed exercises can be returned to for further study. Another advantage is the option of playing through alternate responses by the defending player. A small tree symbol appears in the game annotation where alternate responses are possible. Some exercises have complex trees.

The app has a simple design. Pressing a plus advances through the exercises in each level, or through the several levels. The user may configure whether the objective (White to checkmate, White to draw) is announced at the start of the problem, or remains hidden. I find this feature useful, and have turned off the notification. Hence, I must assess whether I am playing to win or to draw. Sometimes. the objective is merely to gain an advantage, or to avoid losing equality.

The quality of the problems selected is a significant strength of Chess Quest. They are well chosen.

The app works in portrait mode and does not rotate. I paid $4.99 for the app, but this price is a sale price "for a limited time."

GM Leonid Yudasin selected the exercises.

Chess-wise Pro

Chess-wise Pro ($4.99) is a comprehensive chess program offering play against the computer, six-piece Nalimov tablebase access, online playing at FICS, database features, and tactics training. There are 300 training exercises. Through the first 100, a hint can be seen on request.

The exercises are moderate in difficulty (comparable to levels one and two in Chess Quest). Some classic exercises appear in both apps, and can be found in many tactics books. One drawback to the program is that it asks only for the first move. If a user sees the correct move, but miscalculates a sequence, he or she gets the problem correct. In Tactic Trainer below, one error in the sequence causes a fail on the problem as a whole.

The app tracks which exercises have been completed. All exercises are accessible from the entry screen, making it possible to solve them in any order, and to return to any problem at any time. After I completed all 300 exercises, I was able to reset the list so that all are marked as unsolved. I like the possibility of going through the whole set a second time.

I use Chess-wise Pro as an integral part of my training, but do not rely upon it. I recommend against using it to the exclusion of other resources.

Several sets of chess pieces are available to suit different preferences. The app rotates fully.

Shredder

Shredder ($7.99) is a terrific training tool as a playing program, and it offers 1000 tactics problems to solve with the clock ticking. Speed and accuracy are necessary for full score.

Tactics training with Shredder resembles training with Chess Tactics Server or Chess Tempo in blitz mode in that problems are scored and time counts. Wrong answers and too much time both reduce the possible score.

Most of the problems are slightly less challenging than the norm in Chess-wise Pro, but the timer balances the  effort needed. Shredder is an integral part of my chess training. Recently, I completed my second pass through all 1000 problems. In 2013, I aim to complete my third, but with a higher scoring percentage than achieved through the first two passes.

There are several board colors and piece sets from which the user may choose to render the appearance aesthetically pleasing. The app rotates fully.

See my more in-depth review at "Tactics Training: Shredder iPad App."

Tactic Trainer

Tactic Trainer ($2.99) offers exercises to be solved that increase in difficulty after success, and decrease in difficulty after failure. The developers claim that it contains more than 20,000 problems. Solving the problems produces a rating--Glicko rating system--that determines the level of problems presented.

Strengths of the app are that it is simple and offers problems of an appropriate degree of difficulty for chess enthusiasts of a wide range of skill levels.

Weaknesses are that access to problems previously solved does not permit trying again. An update in May 2013 added tracking for number of problems attempted, and how many were solved correctly. The app is like a Chess Tempo lite with minimal capabilities for review, and with limited performance data. However, Chess Tempo requires a web connection. Tactic Trainer can be used in the back country or on an airplane far from internet access.**

The app offers a range of chess sets and pieces from which to choose (slightly fewer than Shredder). It rotates fully.

See my in-depth review at "Tactic Trainer for iPad, iPod, iPhone, Android: Review."

I believe these five apps are among the best resources for tactics training using the iPad (or iPhone and iPod touch). But, it may be that there is a terrific app unknown to me. Please leave a comment if you know of one that should have been mentioned in this comparative review.

*In two and one-half weeks use, I have spend more than five and one-half hours on tactics (1707 exercises) and another two and one-half on endgames (307 exercises).

**This paragraph was altered 8 June 2013 to reflect changes included in an app update several weeks earlier.