Recap of the 2018 Candidates & Crazyhouse Motifs

This is a long overdue recap of the Candidates, completing the earlier half-time recap. All the highlights can be found in studies linked at the bottom of the blog, but one way of enjoying them might be to try solving the puzzles in part 2 of the Candidates Practice against Computer study and I share a hopefully instructive video with a selection from this study to illustrate 19 important crazyhouse motifs. The first part of this practice study was linked already in the half-time recap.

As expected, opperwezen closed out his campaign in style to secure his place as the Challenger, although conceding a draw to sexy_and_I_know_it to reach a final score of 9.5/10 with gsvc in 2nd on 8.5/10 and TwelveTeen on 8/10 who has sadly retired from the game at least temporarily. littleplotkin & JKtheBullfrog completed the top half, coming 4th equal on 5.5/10. IM opperwezen would indeed go on to become the 2018 Champion after a fantastic Final against JannLee spanning three weeks in November.

Here is the full index of Candidates Matches:

21st August 19:30 opperwezen - littleplotkin 8-2, first game, video (opperwezen)

23rd August 19:00 xuanet - JKtheBullfrog 5-5, first game, video (opper)

24th August 17:00 opperwezen - xuanet 7-3, first game, video (opper)

26th August 19:00 TwelveTeen - opperwezen 4-6, first game, video (opper) video (JannLee)

31 August 14:00 littleplotkin - gsvc 4-6, first game, video (JannLee & GM Yasser Seirawan)

31 August 19:00 gsvc - opperwezen 2-8, first game, video (opperwezen) video (JannLee & Yasser)

1st September 20:00 TwelveTeen - Kleerkast 6-4, first game, video (JannLee & Yasser)

2nd September 15:00 gsvc - Kleerkast 6-4, first game, video (okei)

2nd September 18:00 TwelveTeen - sexy_and_i_know_it 8-2, first game, video (JannLee & Yasser)

3rd September 19:00 gsvc - TwelveTeen 5-5, first game, video (okei)

5th September 03:00 TwelveTeen - littleplotkin 6-4, first game, video (okei)

7th September 19:00 opperwezen - Kleerkast 7-3, first game, video (opper) video (JannLee & Yasser)

8th September 14:00 gsvc - sexy_and_i_know_it 6-4, first game, video (JannLee)

8th September 19:00 blitzbullet - TwelveTeen 1-9, first game, video (okei)

9th September 13:00 sexy_and_i_know_it - Kleerkast 7-3: first game, video (okei)

9th September 15:00 JKtheBullfrog - gsvc 2-8, first game, video (okei)

13th September 17:00 JKtheBullfrog - opperwezen 4-6, first game, video (okei)

14th September 16:00 xuanet - gsvc 4-6, first game, video (okei)

15th September 12:00 mastertan - sexy_and_i_know_it 4-6, first game, video (JannLee)

15th September 22:00 xuanet - TwelveTeen 2-8, first game, video (zhbruhs)

16th September 00:00 gsvc - mastertan 6-4, first game, video (zhbruhs)

16th September 14:00 blitzbullet - gsvc, first game, video (crptone)

23rd September 12:00 opperwezen - mastertan 8-2, first game, video (okei) video (opper)

23rd September 22:00 TwelveTeen - mastertan 7-3, first game, video deleted by Twitch (zhbruhs)

28th September 09:00 mastertan - xuanet 4-6, first game, video (okei) analysis (okei)

30th September 10:00 Kleerkast - mastertan 6-4, first game, video (JannLee)

4th October 04:30 TwelveTeen - JKtheBullfrog 5-5, first game, video (okei)

5th October 16:30 xuanet - sexy_and_i_know_it 6-4, first game, video (crosky)

7th October 14:00 blitzbullet - Kleerkast 3-7, first game, video (okei)

9th October 12:00 littleplotkin - sexy_and_i_know_it 7-3, first game, video (okei)

13th October 13:00 sexy_and_i_know_it - blitzbullet 4-6, first game, video (JannLee)

13th October 15:00 xuanet - blitzbullet 5-5, first game, video (okei)

13th October 19:00 Fumitoks - opperwezen 4-6, first game, video (JannLee)

14th October 12:00 littleplotkin - Kleerkast 7.5-2.5, first game, video (okei)

14th October 16:00 Fumitoks - xuanet 7-3, first game, video (JannLee)

18th October 14:00 sexy_and_I_know_it - opperwezen 5-5, first game, video (JannLee)

20th October 13:00 Fumitoks - mastertan 5-5, first game, video (okei)

21st October 14:00 Fumitoks - Kleerkast 6-4, first game, video (JannLee)

21st October 16:00 Fumitoks - sexy_and_i_know_it 7-3, first game, video (okei)

22nd October 16:00 xuanet - littleplotkin 4-6, first game, video (okei) 

23rd October 02:00 littleplotkin - mastertan 5-5, first game, video (okei)

23rd October 23:00 Fumitoks - littleplotkin 8-2, first game, video (crosky)

25th October 13:00 Kleerkast - JKtheBullfrog 2-8, first game, video (JannLee)

25th October 15:00 JKtheBullfrog - littleplotkin 7-3, first game, video (JannLee & Yasser Seirawan)

25th October 17:30 Kleerkast - xuanet 4-6, first game, video (JannLee)

27th October 12:00 littleplotkin - blitzbullet 5.5-4.5, first game, video (JannLee)

28th October 13:00 opperwezen - blitzbullet 7-3, first game, video (JannLee) 

28th October 16:00 Fumitoks - gsvc 2-8, first game, video (okei)

1st November 23:15 Fumitoks - JKtheBullfrog 4-6, first game, video (okei)

4th November 14:30 sexy_and_I_know_it - JKtheBullfrog 3-7, first game

4th November 18:00 TwelveTeen - Fumitoks, first game, video (Jann)

9th November 02:00 mastertan - JKtheBullfrog, first game

12th November 00:30 Fumitoks - blitzbullet, first game

12th November 00:30 blitzbullet - mastertan, first game, match cut short at 1-6 & forfeited

12th November 02:00 JKtheBullfrog - blitzbullet, match never happened

Final

10th, 17th & 24th November 18:00 opperwezen - Jann Lee

Final Study (60 games)

JannLee live-stream of the Final without commentary

Day 1 

Day 2

Day 3

opperwezen live-stream of the Final without commentary

Day 1

Day 2

Day 3 

Commentary of the Final

Day 1 Flourish & Antic

Day 2 Flourish & Antic

Day 3 okei, MMichael, Flourish, Fumitoks, and others from the House Discord

Final Highlights video (okei feat. opper)

Full 2017 CWC Video Playlist

Full 2018 CWC Video Playlist

I have chosen a selection of puzzles from the second half of the 2018 Candidates to present in the following instructional video to highlight 19 important crazyhouse motifs. I do not move the pieces in these puzzles to try to encourage viewers to pause and try to calculate the lines. For advanced crazyhouse players who just want to see the puzzles without commentary, see the last 5 minutes and pause for as long as you need to solve each one.

The motifs illustrated are:

1) Back Rank

2) Queen Invasion

3) Drawing the King Out 

4) The Yoyo Mate (drawing out with pawns & pushing back)

5) Deflecting Defenders

6) The Magnet Mate (displacing the king to better mate it)

7) Discovered Attacks

8) Exploiting a Pin

9) Interference

10) Minor Piece Mates

11) Double Mate Threats

12) Assessing Whose Mate is Faster

13) Queen Factory

14) Corner Rook Drops

15) Smothered Mates (with knights)

16) Smothered Mates (with pawn or queen)

17) Clearance

18) Turning Defence into Attack

19) Drawing Motif: Perpetual Knight Checks

We hope to delve deeper into each of these and many more crazyhouse motifs in future instructive videos in the coming year. You can practice these puzzles and many others against the computer, or you can see the full lines and variations in the following studies:

Candidates Study Part 1

Candidates Study Part 2

Candidates Study Part 3

Candidates Study Part 4

Candidates Practice against Computer Study Part 1
Candidates Practice against Computer Study Part 2

Wishing everyone a good New Year!

—okei

CrazyAra plays JannLee for Christmas

We had an equinoctial matchup at 18:00GMT on 21st December (midwinter in the Northern hemisphere and midsummer in Australia) between the neural-network based CrazyAra and the 2017 Crazyhouse World Champion JannLee. CrazyAra was created as part of a university project at the Technische Universität Darmstadt by Johannes Czech (IQ_QI), Moritz Willig (RiseOfLight) & Alena Beyer  inspired by the methods of AlphaZero which they were being taught as part of a masters’ course "Deep Learning: Architectures & Methods" held by Prof. Kristian Kersting and Prof. Johannes Fürnkranz in summer 2018. 

Since, in crazyhouse, you can drop any piece in hand (up to 5 choices) on any unoccupied square (at least 32 choices), the number of possible moves and move sequences is several orders of magnitude higher than in regular chess. Nevertheless, it is still several orders of magnitude less than the square of the number of squares on a chessboard, so it was not too difficult to encode all these possibilities. Still, most of these possible moves are clearly not even worth considering. But how to build a program that has this intuition so that it does not waste precious time and electricity on obviously futile moves? This is where the CrazyAra neural network comes in which attempts to predict most-likely moves based on a period of training on over half a million crazyhouse games played on Lichess over a two-and-a-half year period (January 2016 – June 2018) between players both rated over 2000 in the variant. This training data featured over 25 000 games by mathace, as well as over 10 000 games played by ciw, retardedplatypus123, xuanet, dovijanic, KyleLegion, JannLee, crosky, mariorton, opperwezen, saturos & mastertan, all well-known strong regulars in the crazyhouse community. But as well as this store of learned intuition, the program also needed to be able to follow through concrete lines. For this, it uses a Monte-Carlo tree search (MCTS in the jargon) to search through all possible moves evaluating likely outcomes. While at depth 1 (one half-move deep) all possible moves are considered, the deeper it analyses, the more it draws on the neural network of intuition built up from half a million games of training data.

To delve deeper into the technicalities, check out the github and wiki pages, but we hope to explain the basic idea here. For every board position, CrazyAra compares to the neural net built up from its training on the games data to assess firstly the probabilities or p-values of every possible move in this position. If a move is more likely, then CrazyAra will explore that branch of the tree to a greater depth. Secondly, the neural net will try to fit the given position to the neural net produced from all the positions on which it has trained to give it a q-value indicating the likelihood of a win or loss (ranging from -1 for a loss to +1 for a win) from the given start position. This network q-value will change of course as the Monte-Carlo search tree is run depending on the q-values of the children nodes reached by following the most likely lines of continuation. 

The name CrazyAra is based on a genus of South-American parrot (ara), the etymology of which presumably derives from the sound they make “arara...”. Parrots are intelligent animals which are able to learn from and mimic human behaviour. In the case of CrazyAra, the neural network tries to replicate human play and predict the outcome of games. In fact, the name DeepLee was originally considered as a combination of DeepMind (the British company bought up by Google which created AlphaZero) and JannLee, the World Champion who was a major inspiration for starting the project and how the second author (IQ_QI) and his collaborators were introduced to the crazyhouse variant. Crazyhouse seemed like an interesting hybrid between chess and shogi (both of which AlphaZero had learned to play) which had not been tackled before by any neural network engine. But the name DeepLee was already taken, so CrazyAra it was. Nevertheless, it was a great privilege that JannLee was willing to play it.

CrazyAra first came online on Lichess three months ago to test its mettle against members of the crazyhouse community such as okei, Andrey2888, Flourish and gsvc. Version 0.1 impressed with its attacking flair and opening innovations, but it was still very beatable. 

The first author (okei) was disappointed to win only one game out of ten, letting slip many promising positions. In one surprising game with gsvc, CrazyAra was doing very well against the Paraguayan IM who is in the top 5 in the world, but then CrazyAra help-mated herself in one and gsvc duly gave the smothered mate. 

Version 0.3 came online for just 24 hours three weeks ago and was a much tougher proposition. TheFinnisher was beaten 10 times before finally being the first to beat it. The programmers had completely re-written the Monte-Carlo search tree algorithm in Version 0.2 as well as completely re-training the neural network using a new deeper architecture (19 layers instead of 7, as well as squeeze-and-excitation modules) which not only gave better predictions but also took less time to train. Originally, it hadn't been planned to continue the project further. But the developers were surprised by the interest shown as well as the supportive attitude from the main stockfish developers for crazyhouse, Fabian Fichter, Daniel Dugovic and well known chess-engine programmers such as gbtami. Version 0.3 included faster profiling, evaluating 250 nodes of the search tree per second (up from 70) and better time management thanks to feedback from gsvc, so that more obvious moves would be played more quickly.

Some further tweaks were made in Version 0.3.1 which was the version that played JannLee three days ago. For example, a position already analysed previously in the Monte-Carlo tree search would not be analysed afresh if it occurred again, and the speed was upped to 300 nodes per second.

Here is the live stream of the five games that CrazyAra played with JannLee. Following the video are annotated studies of each of the five games, looking from CrazyAra’s perspective.

JannLee came into game 1 thinking, “If I'm going to have any chance against the engine, it's going to be a strategic motif, not open tactical warfare.” And so he began to probe the engine’s positional understanding but CrazyAra proved herself no slouch at strategy, gradually improving her position and grinding the game out.

Both JannLee and CrazyAra missed 22. N@f6+ with a forced mate in 3! But the actual finish was much more spectacular. Game 2 will always be remembered for the stunning tactical motif with which JannLee unclasped CrazyAra’s defences: 30. N@g6!! followed by 32. Q@d6! (although as you can see in the study, 32. Q@f6!! with the further idea of attacking g7 was even stronger).



Game 3 was a masterclass from the engine.



In game 4, Jann spotted a mating idea with Q@f7 followed by Qxe7+ but not being able to see the full continuation he did not follow through on his plan, and let slip what would have been a good win.



In the given position in game 5, CrazyAra had a mate in 4! But she missed it and proceeded to let slip a dominant position as JannLee defended like only JannLee can. But just as JannLee was regaining command, he sacked a queen (which although objectively the best move it gave CrazyAra a lifeline) and then missed the critical 57. ... Q@g4+! mopping up CrazyAra’s attack. He instantly regretted it, and even more so when the discovered check on move 64 that he had calculated not to be dangerous came with a back rank threat that he had not foreseen. CrazyAra had won with a very human-like swindle.

Check out this CrazyAra study if you would like to explore these lines and variations further or for some of the other games CrazyAra has played in training or on its two other outings to Lichess (3 weeks and 3 months ago).

As we saw, CrazyAra 0.3.1 still misses mates (not delivering a mate in 4 on its turn in game 5 as well as hanging a mate in 3 in game 2). To try to alleviate this flaw, a “blunder check” will be incorporated in future versions, running the MCTS from the opposite perspective to CrazyAra so that having chosen a candidate move to play, the program checks (as a human would do) to make sure that there is not some forcing line that loses. Also, CrazyAra has yet to undergo any self-play training, or training by playing against Stockfish or watching Stockfish play, so we imagine this will be the next big step. 

Some further ideas for improvement are (1) to adjust the probability distribution (the p-values) so as to expect the unexpected, paying a little more attention to seemingly less likely moves, (2) a test database of critical positions (where there is one clearly best move), so with each new improved version of CrazyAra, the developers can test if the time taken to solve the critical positions does indeed decrease on average, (3) a deeper analysis of forcing variations (especially checks & captures), to avoid the engine missing mates or winning lines, (4) in the Monte-Carlo tree search, stopping ideally at child nodes which are peaceful positions (positions with many possibly equally good moves) at which the network’s q-values assessing the position are likely to be more accurate and finally (5) the introduction of a free move counter. In a given position, if I do a null move, then does my opponent have forced mate? If so, I have 0 free moves as I must defend the mate in some way. If not, then can I do two null moves without being force-mated? In this case I would have 1 free move. Turning the board round, if I have more free moves than my opponent does after my move, then this gives me a license to attack. Informally speaking, if both players ignore and try to mate each other, then the player with the greater number of free moves in a given position is likely to win, especially if this number is small. This kind of counter might better hone CrazyAra’s strategic sense for when to defend and when to attack which would complement its tactical and positional strengths.

Wishing everyone in the crazyhouse community a very merry Christmas!

—okei & IQ_QI

2nd Season of the 10+10 ZH-er0 League

As a spin-off of the Lichess4545 League organised over Slack, there is also a 10+10 crazyhouse league organised by Zher0. Each week, players play two games of 10+10 crazyhouse, one Black and one White, against their assigned opponent. Both 1.5-0.5 and 2-0 count as a win, while 1-1 is scored as a draw. The second season was recently completed, and this time I was the outright winner with 6/7, with a draw against Antic in second place (which I live streamed at the time) as well as against yago666 in third. The third season is just beginning this week, but if anyone would like to join with a first-week bye, then they are welcome to sign up by contacting Zher0 on the Lichess4545 Slack.

Final Standings (compiled by Zher0 & Highonpotnuse)

We enclose in this blog a study with 16 mating puzzles which you can practice against the computer, as well as a video in which I share the puzzles on stream. There is also a solutions study with shortest variations which I will share at the end of the blog. I reproduce images of the study puzzles here:

Puzzle 1: yago666 - okei White to move

Puzzle 2: Isachess - Seb32 Black to move

Puzzle 3: Isachess - Seb32 Black to move

Puzzle 4: FiveKnights - Zher0 White to move 

 Puzzle 5: okei - yago666 White to move

Puzzle 6: yago666 - colwem Black to move

Puzzle 7: Zher0 - gokuba White to move

Puzzle 8: Antic - yago666 variation Black to move

Puzzle 9: Antic - okei White to move

Puzzle 10: FiveKnights - Isachess White to move, but finish the mate for Black

Puzzle 11: Zher0 - okei variation Black to move, but finish the mate for White

Puzzle 12: okei - Antic White to move

A solution study with full variations included may be found here.

—okei

IM opperwezen is the 2018 Crazyhouse World Champion!

The CWC Final capping off the year-long Crazyhouse World Championship was won by the Challenger! While Carlsen-Caruana saw 12 draws of classical chess between them, opperwezen and JannLee played 60 games of 3+2 spanning 3 weeks with only one draw. At the end of week 1, JannLee led 11-9. opperwezen pulled into the lead on week 2 taking it 12-8, so 2 ahead 21-19 going into week 3. Needing 9½/20 on the final day, he secured a draw to almost guarantee a decisive result and left Jann needing a streak of six to hold his title. Jann won the first two, but in the end the opper proved too strong taking the series 32½-27½ overall. Below are highlights of some of the best games, and we were joined by the man himself for the last forty minutes:

To re-live the action live, you can watch the live streams:

JannLee (without commentary)
Day 1 
Day 2
Day 3

opperwezen: (silent)
Day 1
Day 2
Day 3 

Day 1 Flourish & Antic (live commentary)




Day 2 Flourish & Antic (live commentary)




Day 3 okei, MMichael, Flourish, Fumitoks, and others from the House Discord



Finally, if you missed anything from the 2018 CWC, check out this CWC 2018 playlist of over 90 streams of commentary and analysis. Follow the streamers you like on YouTube. And check out all the recaps of each round posted earlier on this blog with embedded puzzles and annotated games.

—okei

Spot the Move! #2 Opper to Move!

Today (Saturday 24th November) is the final day of the 2018 Crazyhouse World Championship and Day 3 of the Final between opperwezen and JannLee is at 1800GMT. It will be followed by an Elite Arena (>2100 zh on lichess) at 2100GMT and opperwezen has expressed interest to team up with Jann for some hand-and-brain crazyhouse this weekend or over the coming few days. This is also the crazyhouse week of the FischyVishy variant tournaments on Lichess. See the crazyhouse calendar in the right sidebar for timings. But who will be crowned the Crazyhouse Champion in less than 8 hours?

Day 1 saw JannLee edge ahead 11-9 in the first 20 games. On Day 2, opperwezen bounced back winning 12-8. The score therefore stands at 21-19 in favour of opper with 20 games to play. If scores are tied at 30-30 tonight, then two more games are played and again until the deadlock is broken and someone is ahead. The event will be live-streamed by JannLee (with 3-minute delay and without commentary), by flourish_zhouse with Antic, and also by myself (okeizh on Twitch) with MMichael aka TheFinnisher from the House Discord. It can also be watched directly with in-game chat on Lichess by clicking on either JannLee TV or opperwezen TV.

For an in-depth analysis of game 1 of 60 of the CWC Final, check out the video on my YouTube and do subscribe! Also, following day 1 of the Final a fortnight ago, there was a re-enactment for the CWC of the leaked video prep of Caruana in the World Championship in this very instructive 90-minute long video sharing opening ideas on how to attack weak squares in crazyhouse from beginning ideas to advance. Along the way, at 1:03:25 - 1:09:10 the camera pans to a study of CWC novelties and games, specially created by opperwezen and myself for the purpose of this video.

As a prelude to the action on Day 3, I would like to share with you some puzzles from opperwezen’s games with JannLee, followed by hints, solutions and a link to a lichess study at the bottom with full variations. These are not necessarily mating studies, but critical positions in which opper either did or did not find the critical move (sometimes under time pressure), and we challenge you to do better! Before we begin the puzzles proper, let me share two examples where by contrast it was Jann to move against opper, with solutions included. These two are embedded on lichess so you can move the pieces, but the puzzles you will have to calculate in your head.

White (opperwezen) has just promoted on d8 and it is Black (JannLee) to move. This was from a game back in 2016, but opper still remembers it. Jann did find the critical move! It both stops mate, as well as carrying an imminent mate threat of its own which indeed came to pass. The second example is from a recent game, and perhaps the players were unaware, but there was a critical winning move which Jann missed:

Now, in all the following positions, you are opper at the board and it is your move. You are up against the World Champion JannLee. What do you play? Write down the first move. Hints, solutions and a study with full variations follow.

Puzzle 1

Puzzle 2

Puzzle 3

Puzzle 4

Puzzle 5

Puzzle 6

Puzzle 7

Puzzle 8

Puzzle 9

Puzzle 10

Puzzle 11

Puzzle 12

Hints

Puzzle 1: 

Don’t give up the centre!

Puzzle 2: 

Develop with tempo!

Puzzle 3: 

The threat is sometimes stronger than the execution!

Puzzle 4: 

Defend with a counter-attack!

Puzzle 5: 

Take advantage of the opponent’s lack of development!

Puzzle 6: 

Maintain initiative!

Puzzle 7: 

Sacrifice to make a breakthrough!

Puzzle 8: 

Pawns are the matches to ignite a mating attack!

Puzzle 9: 

Distract the defender!

Puzzle 10: 

Attack the back rank!

Puzzle 11: 

Create weaknesses!

Puzzle 12: 

Know your mating patterns!

Solutions

Puzzle 1: d4! opper found the move twice in 2017, but subsequently missed it once in 2018. 8. Bxd4+ is the sharpest continuation for Black, followed by Qh4+, although this has never been played yet.

Puzzle 2: Bd6! opper missed the move! Nxd3 Bxd3 @b7 0-0 bxa8=Q is the sharpest continuation for White, but then Black has QxBa4!! QxQa4 B@b5 with multiple threats and Black is better.

Puzzle 3: NxNe5! followed by N@f7 recycling the threat. NxQd8 may also be equally good, so let’s count that as correct also. But opper missed the move, taking the rook which is certainly worse.

Puzzle 4: N@e2+! followed by @f3! attacking the White squares on which the White king must hide, dxQe7 can be met with fxg2+ Kxg2 @f3+ Kh1 BxPe7. White may have a queen on the board and a queen in hand, but RxNe2 is practically forced after which @g2 Kg1 fxRe2 and Black has three pawns on the second rank with a rook in hand and a fourth pawn threatening cxb2. opper missed the move.

Puzzle 5: Nd5! Black’s only defence of c7 is Kd8, after which Nxc7! anyway. Kxc7 falls to a mate in 11! There is a sharper defence in the study, but it still loses the Black queen. opper found the move!

Puzzle 6: Qxg5! Bxg5? falls to a beautiful knight mate. Can you spot it? opper found the move!

Puzzle 7: N@c5! dxc5 allows a stunning queen sacrifice Qxc7+!! followed by Nb5 which is mating. Ke8! is the best response from Black, but after Q@d7+ Kf8 QxNc6, White is clearly better. opper missed the move.

Puzzle 8: @f3+! Qxf3 or Kxf3! are both met by dxe4! and the former is worse because the threat is N@g1!! opper missed the move.

Puzzle 9: Q@h4! Taking the queen lets go of defence of f7, but not taking leads to a queen trade and the potential Q@g8+! followed by a double-take smother on f7. opper found the move!

Puzzle 10: Q@e7! BxQd1 falls to QxRf8 followed by R@h8+, R@e8+ Nxh7+ and @e7#, or even better R@h8+!! Kxh8 QxRf8 followed by Nxf7 with the smother. opper found the move!

Puzzle 11: N@h6+! is crushing for White, creating a weakness on g7. White may later even sacrifice his queen on g7 and finish with a minor piece mate. opper missed the move.

Puzzle 12: N@h3! White’s sharpest defence to stop R@g1# is B@g1, but then Black has NxBf4 and an inevitably mating attack via h3 and g2 although it could take over 20 more moves with best defence. This final puzzle was taken from a 97-move 3+2 game in April 2018. opper missed the move and Jann ultimately clinched the game.

In conclusion, opper hit a home run on 4 of these 12 critical positions. Bear in mind, a lot of these games were 1+0 zh and the opper of today would no doubt score much higher. How many did you manage? Check out the Opper to Move! Puzzle Study to play through any variations you might have been curious to delve deeper into. 

—okei