A new learning technique
By John Foley
King and pawn endgames are supposed to be simple but in practice they present the most tricky challenges. The improving chess player will want to revert to king and pawn endgames to study their subtleties. Mistakes are common and we can learn from other people’s mistakes as well as our own.
I present two chess positions which are so basic that at first glance the club player would probably deliver an instant opinion on the result. It would of course be possible to put these positions into a chess analysis engine or Tablebase and show how they should be won. However, that is not how to learn chess. We need to understand how people actually play chess and the nature of their errors.
Fortunately Tim Mifsud has created within Chessbites an excellent new facility to find any games in which a particular position has arisen. I checked out his Position Finder on a couple of basic king and pawn endgames.
A simple kings ending
| 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| 0 | 0 | 0 | 0 | 0 | 0 | 0 | |
| 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| 0 | 0 | 0 | 0 | 0 | 0 | 0 | |
| 0 | 0 | 0 | 0 | 0 | 0 | 0 | |
| 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
King and pawn v king
The database finds 14 games in which this position arose. The outcome is determined by who is to move, for that person has the King Opposition. The results of the games reflect the theory: with White to move the result is a draw but Black to move, White wins. It seems no rated competitive player has had any trouble with this position in practical play.
The position to be examined
| 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| 0 | 0 | 0 | 0 | 0 | 0 | ||
| 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| 0 | 0 | 0 | 0 | 0 | |||
| 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
King and two pawns v king and pawn on the wing
This position is more instructive. The database returned 13 games in which this position occurred. In theory, White to move should be able to win by manoeuvering round to capture the h6 pawn, benefitting from the extra tempo from the h-pawn if required.
Black to move first should be able to obtain a draw by opposing the White king and blocking the g-pawn.
However, some results were contrary to theory as the players made mistakes.
White’s mistake: overreaching
| 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| 0 | 0 | 0 | 0 | 0 | 0 | 0 | |
| 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| 0 | 0 | 0 | 0 | 0 | |||
| 0 | 0 | 0 | 0 | 0 | 0 | 0 | |
| 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
White to move
White can overreach and miss the win. In Moreira v Cleto (Portuguese U-18 Championships, 1995), the critical mistake arose after 1. Ke5 Kf7 2. Kf5 Kg7 3. Ke6 Kg6 4. Ke7 h5 5. g5 Kg7 6. Ke6 Kg6 7. Ke5 Kf7 8. Kf5 Kg7 this position arose.
What did White play and why was it a mistake?
White advanced 9.g6? which seems an active move, but actually throws away the win. The trouble is that after 9…Kh6!, White cannot protect the g-pawn except by 10. Kf6 which is stalemate.
White’s correct plan should have been to force the Black king backwards to where it cannot reach the security of h6 and only then to advance the pawn. For that purpose, White first needs to lose a tempo and 9.Ke4! would have done the trick. This demonstrates the need for constant vigilance even in a “won” king and pawn endgame.
Black failing to secure direct opposition
| 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| 0 | 0 | 0 | 0 | 0 | 0 | ||
| 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| 0 | 0 | 0 | 0 | 0 | |||
| 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
Black to move
Returning to the second position, Black did not always secure the draw. In Ventura Monfort v Iranzo Paricio (Castellon Championships, Spain, 2003) after the moves 1… Kf6 2. Ke4 the following position presented Black with a dilemma: whether to stay in front of the half-passed g-pawn or whether to maintain the Opposition by 2…Ke6.
What did Black play and why was it a mistake?
Black played 2…Kg6??
This allowed the White king to penetrate with 3.Ke5. Black soon lost after 3… h5 4. g5 Kg7 5. Kf5 Kf7 58 g6+ 1-0
(Note that the g6 advance occurs when Black king cannot reach h6.)
Instead Black should have played 2…Ke6 to prevent the White king from penetrating
Black failing to secure distant opposition
| 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| 0 | 0 | 0 | 0 | 0 | 0 | ||
| 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| 0 | 0 | 0 | 0 | 0 | 0 | ||
| 0 | 0 | 0 | 0 | 0 | 0 | 0 | |
| 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
Black to move
There was a second game in which Black failed to secure a draw. In Moritz v Iosif (Hessen Championships (Women), Germany, 1993) after 1…Kf6 White gave Black something to think about with 2.Ke3!?
What did Black play and why was it a mistake?
Black played the natural move 2…Ke5?? taking up close opposition.
Unfortunately this loses to 3.Kf3 after…Ke6 4.Ke4 Kf6 5. Kd5! and the white king breaks in.
The problem for Black is that to make effective use of the opposition, it is necessary to move sideways to keep the other king out. Unfortunately, Black cannot do this because the white pawn intervenes. Black missed the key concept: Distant Opposition. The correct move was 2…Ke7! and then if 3.Ke4 Ke6 keeps the White king out.
Although Black did err, she had been enticed to do so by White who had played a clever psychological move. Finding the right response required knowledge of the Distant Opposition concept or else the calculation of a subtle variation.
Thus using the position finder, we can learn a great deal about the types of errors of thinking that humans make.
Finding the position finder
The creator of Chessbites, Tim Mifsud has not played competitive chess for many years. Although a strong junior (he was the British u-13 champion in 1991) and champion of his native Malta, it was impracticable to develop a chess career from the Mediterranean island. Yet the chess muse still danced within his soul and he has been developing an innovative chess database which identifies whether any position has been played previously. His website provides a useful addition to the analytical tools for the chess improver.
Now relocated to Kingston, Tim has just implemented an elegant solution to database position search. He figured out an algorithm which makes the search very fast. Every position is encoded into a number and so the search is reduced to finding duplicates on a long list. His database contains over four million games which is updated weekly.