An extraordinarily unique variation

Boxum - Wiersma Drachten 2023
Boxum - Wiersma Drachten 2023

On 7 December 2024, a spectacular game took place in the Dutch team competition between multiple world champion Harm Wiersma and the talented Matheo Boxum. Wiersma represented Damcombinatie Fryslân, while Boxum played for Hijken DTC. There was particularly much attention for this game on Toernooibase. This was mainly due to an excerpt from an email by Harm about the game, which was posted on Toernooibase the following day by Hanco Elenbaas:

"It is indeed very rare that I have had an 'onvoltooide hekstelling' position on the board. However, I may send you a number of variations from that game today that are extraordinarily unique. During my game, I saw the outlines of the variation, but due to a lack of time, I did not take the variation, which in hindsight seems to be a masterful winning plan. During the game analysis at the board immediately after the game, I did mention it, but it was dismissed by onlookers as unrealistic. I will send you the message with the request to post it on Toernooibase. I have never posted anything there (!)."

This post immediately created a lot of suspense, since it was a great mystery which variation Harm was talking about. There was much speculation about it on Toernooibase, and it took a whole week before Jan Pieter Drost finally managed to discover the variation in question. Ultimately, it was Harm Wiersma himself who provided additional clarification during the live analysis of the world championship match Anikeev - Groenendijk. Below we can see for ourselves that the qualification extraordinarily unique was not an exaggeration.

Matheo Boxum - Harm Wiersma 0-2 (7-12-2024)

1.32-28 18-23 2.33-29 23x32 3.37x28 20-25 4.41-37 17-21 5.37-32 15-20 6.39-33 19-24 7.44-39 21-26 8.50-44 26x37 9.32x41 16-21 10.41-37 14-19 11.38-32 10-14 12.43-38 5-10 13.49-43 21-26 14.36-31 12-18 15.46-41 8-12 16.31-27 3-8 17.41-36 11-16 18.37-31 26x37 19.42x31 7-11 20.47-42 11-17 21.27-22 18x27 22.31x11 6x17 23.42-37
diagram 1

This position has appeared before in the game Bezvershenko - Krajenbrink 1997 during an international match between the Netherlands and Ukraine. That game resulted in a draw after

17-21 24.36-31 21-26 25.31-27 13-18 26.35-30 24x35 27.29-23 18x29 28.33x15 8-13 29.38-33 13-18 30.43-38 19-23 31.28x19 14x23
diagram 2
This is the position where it all revolved around. The comments below are from Harm Wiersma, for which I used the transcription of Hanco Elenbaas as the source.

32.33-28

This seems logical. After 32.48-43 , for example, 2-7 33.33-29 (on 33.33-28 12-17 34.28x19 and 17-22 ) 7-11, and white is almost forced to play 34.27-22 18x27 35.29x7 1x12 36.32x21 16x27, after which black at least has the best play. On 32.33-29 2-7 33.39-33 (on 33.38-33 7-11 34.33-28, black gets almost winning play through 25-30 and continuous breakthrough threats on white's long side) and now again 7-11 34.27-22 18x27 35.29x7 1x12 36.32x21 16x27, etc., and I prefer the remaining position for black.

By the way, black can also take the exchange 25-30 34.34x25 23x34 35.40x29 18-22 36.27x18 12x34 37.39x30 35x24 with equal play.

The variation that troubled me during the game was, however, 12-17 33.28x19 17-22. Now 34.39-33 or 34.38-33 or 34.27-21 all provide good play for black. Only 34.19-14 22x33 35.14x5 seems good, and the consideration was 26-31! 36.39x28 18-23 37.28x19 9-14! 38.19x10 and 31-36.

diagram 3

As a winning plan, I didn't trust this, but upon analysis, it seems white can escape in a treacherous way. The only method for this is 39.32-28 36-41 40.28-23! If black now plays 41-47, then 41.23-19? loses to 2-8, but white can immediately play 41.48-42! If black captures 47x50, then white moves piece 23 to 14(!) with 42.23-19 35x44 43.19-14 and can subsequently free the trapped king and pieces with always 14-9 for a draw. If black captures 47x18, then 42.34-29 18x34 43.40x29 follows. After 2-8, white does not play 44.29-23? (which loses to 8-13), but 44.29-24, and due to the freeing threat 24-20, black must allow 25-30 45.24-19 30-34 46.19-14, and white can always 'move' piece 4 in the long run, the ultimate theme.

Now at move 38 black could also play 16-21 to lure white into 39.48-42, to then have good winning chances after 2-8. On 40.32-28 21-26, the exchange possibility 31-37 42x31 26x37 could play a decisive role. And after 40.34-29 25-30 41.29-23? a nasty gambit follows with 21-27 42.32x21 and 8-13! Therefore, 41.32-28 is better, but then black wins after 21-26, in my opinion. By the way, that is still quite a task, because after 42.28-23 8-13 looks winning. White can then only miraculously escape with 43.42-38 31-37 44.38-33 37-42 45.23-19 13x24 46.29x20, etc. Thus, black must directly play 31-37 43.42x31 26x37 44.23-19 37-42, and now after 45.29-24 42-47! and after 45.19-14 subtly 8-13. (After 42-47, white can still play 46.44-39!! 35x24 47.14-9 4x13, and 48.10-4 for a drawn endgame.) To then play after 46.29-23 30-34! 47.40x29 42-47. The characteristic white confinement with the king on 5 is then maintained.

Finally, at move 38, black could also have played 2-8. Then the advance of 39.48-42 is still very dangerous for white. The best move is again 39.32-28 31-37 40.28-23 37-41. Now 41.48-42? is lost due to 41-47! (because piece 2 is already on 8, the maneuver 42.42-38, etc., does not work). 41.23-19? also loses due to 41-47! 42.19-14 8-13! and white cannot progress further. However, the white draw plan is then 41.34-29! 25-30 (otherwise, white can still force this move by playing 40-34 followed by 29-24 with the threat 24-20). 42.23-19 1-7 (on 16-21 43.48-42 can already be played). Now white can draw by playing 43.48-42 or 43.29-24. But the following is quite remarkable. After 43.48-42 41-47 44.29-24 47x50 white can give an extra piece with 40 which is a draw. But I have been experimenting a bit. If white plays 45.24-20 35x44 46.20-14 then the threat is 14-9 4x24 48.10-4 to make a second king. At that moment I thought hey, I can give two pieces with 8-13 47.19x8 7-12 48.8x17 and make a king with 44-49. The position is a draw now. But if white plays 49.14-9? then black can simply take 50x6 50.9-3 6-22 with a winning endgame. Much nicer is the variant after 49.17-12 then black plays 49-21, and this is quite unique, it goes on with 50.12-7 21-3 51.7-2 3x25 52.2x35

diagram 4

Normaly you would say this position can never win for black. But in this case it is winning, despite that black has only three pieces. This is caused by the extra formation 4/5/10/15. It is similar to what has been what has been written in 1989 by dr. Tjeerd de Reus from Heemstede. As Rein Halbersma pointed out, in 1989 De Reus wrote a book about draw restrictions. In this book he explains the idea of Mertens, namely that if you remove square 5 from the board then the three against one endgame is winning. In this case black can make a third king and build a capture position. Nobody knows about this.


9-14 33.28x19 14x23 34.38-33 1-7 35.48-43 10-14 36.33-29 14-19 37.39-33 12-17 38.33-28 17-22 39.28x17 25-30 40.34x25 23x34 41.40x29 18-23 42.29x18 7-12 43.17x8 2x42 44.32-27 42-47 45.44-39 47-42 46.39-34 42-33

The intriguing position of diagram 3 has occupied Wiersma for days. He has figured out all of the variations without the help of a computer. Even the computer struggles with an endgame like the one in diagram 4. Only after several minutes of calculation it is able to find out that it is winning for black.