Chess Chats

I’ve not been feeling up to playing chess lately, sadly, but I still want to stop by to drop in random neat games people may not have seen before, or neat studies, if nothing else. I mean, most people have seen Morphy’s Opera House Game or Steinitz vs. von Bardeleben, but I wanted to highlight some that I think people would enjoy seeing. I’ll try to do this once a week to see how it goes.

White brings his pieces out very sensibly and has a slight edge because of …Nxd4?! which lets White’s queen hang out, un-harassed, in the center. Then he builds the pressure.

Then, boom! The fireworks start on move 19.

graphic of move 19, white to move

Korchmar vs. Poliak, Kiev, 1937

1. e4 e5 2. Nf3 Nc6 3. Bb5 d6 4. d4 Bd7 5. Nc3 Nf6 6. O-O Nxd4
7. Bxd7+ Qxd7 8. Nxd4 exd4 9. Qxd4 Be7 10. Rd1 O-O 11. e5 Ne8
12. Bf4 a5 13. Rd3 Ra6 14. Re1 Qf5 15. Nd5 Bd8 16. exd6 Nxd6
17. Rg3 f6 18. Bh6 Rf7 19. Nb4 axb4 20. Qxd6 Qd7 21. Qd5 Kf8
22. Rxg7 Qxd5 23. Rg8+ {analysis: Kxg8 24. Re8+ Rf8 25. Rxf8#}

Very cool! Have you seen Maia? There’s a Maia bot on lichess (maia1 : Activity • and published research & code:

1 Like

I’m well aware of Maia, as I’m one of the main lc0 developers (the engine it runs on). The issue is that it is meant to be used with a single node evaluation, which easily misses simple tactics, that a small search finds, and also plays deterministically. Our approach for the lichess LeelaLevel5 experiment is different, we start with a full strength neural net but we detune the search and add randomness to reduce strength and approach human play (but we are still not there).

Edit: This came out more negative than I intended, Maia is an excellent and worthwhile project. My criticism was for the use case of offering a bot in a chess server.


Maia looks very cool! That’s something I’ve always want to do. Thanks for sharing!

IMO, those things are of no concern. :grinning:


Sounds like a chess engine I could beat every once in a while… :wink:

1 Like

Here’s a game in a chessable course (the “Short and Sweet” a.k.a. free sample of Kamil Plichta’s King’s Indian Attack)–mostly just the moves. Starting with 1. Nf3 doesn’t have to be boring!

mate in 4 puzzle

1.Nf3 c5 2.g3 Nc6 3.Bg2 g6 4.c4 Bg7 5.Nc3 d6 6.O-O Nf6 7.d4 cxd4 8.Nxd4 Bd7 9.Nxc6 bxc6 10.c5 d5 11.e4 dxe4 12.Nxe4 Nd5 13.Nc3 Nxc3 14.bxc3 Qa5 15.Rb1 Rd8 16.Rb7 Qxc5? 17.Ba3 Qxa3??
…Qa5 Qb3 and Black finds it hard to move. …Qxc3 Bb4 Qd4 Bxc6! is a winner after …Bxc6 Re7+ Kf8 Re4+
18. Qxd7+ Rxd7 19. Rb8+ Rd8 20. Bxc6+ Kf8 21. Rxd8#


Interesting. I was able to solve that mate in 4 without difficulty. Thanks for sharing.

1 Like

30 seconds on move 4. I don’t know whether this is an argument for learning openings, or against.

Not taking the rook with the pawn is a mistake.


Oh no. I had plans for games to show everyone once a week, but they went down the drain. Well, temporarily.I had a few backlogged. Then I found this.

This really recent “trust me, this is wild and you will enjoy it” game reminded me about this topic, though. It’s just the sort of fun, neat thing where you sit back and enjoy the fireworks.

I don’t know who here knew about Agadmator before… He lives in Croatia and has 1 million + YouTube subscribers. I think he got to a million before Levy Rozman, who now has 3 million, mainly due to an insane work rate and being very good at making video titles that grab search engines’ attention.

Agadmator’s video of this impressive game is here:

Also, I found it cool that he made it on TV Tropes.

And finally, this is not chess-related, but the first 45 seconds of this video are worth watching even if you hate chess. Again, trust me, though past the first minute won’t be interesting to non-chess fans.


Because of the puppy behind him doing air-zoomies on its back?


Somehow, all my hobbies lead me back to chess! I’ve weaned myself off it as a player (something about the structure of online play/tournaments and ELO ratings unlocks a hypercompetitive, stressed-out side of me that I’d rather not entertain, for my own sake and that of the people who have to interact with me regularly :joy:) but I do like keeping up with fun happenings on the board like this:

(And my own tangentially-related addition/plug for one of my favourite musicals of all time: Chess (by Björn and Benny of ABBA glory! and Tim Rice, I suppose)). I’ve had some good fun watching how different productions stage the big epic chess showdowns in the script. An actor from the recent Moscow production said in an interview that he and the other male lead tried to replicate the 1984-85 Karpov-Kasparov games during the show, which I thought was an interesting tidbit!


Yup. Medo is something else. He’s calmed down a lot with age but is still in the background for most videos. I looked through my bookmarks and found this:

World Champion Material? | Caruana vs Nakamura | Sinquefield Cup (2018) - YouTube 1 minute

A Deadly Dare to Give a Guy Like Tal || Tal vs Olafsson || 1959. Candidates - YouTube 1:40 or so

Magnus Carlsen's Improved Bong Cloud - YouTube almost right away

India's Golden Duo || Vaishali Scores Her First GM Norm! - YouTube was also in my bookmarks but I didn’t see it early

Very! Karpov-Kasparov '85 you say? I’ll toss out this game for people to look at, in case I forget for another few weeks. I imagine it is one of the first they tried to commit to memory:

It’s relatively famous (well, for a chess game) because I think even people who don’t love chess will be impressed by how Black’s knight just sits at d3 and annoys the heck out of White, who is up a pawn – an advanced passed pawn-- and can’t do a dang thing! Also, apparently Efim Geller (one of Karpov’s seconds) had an idea Kasparov might play the opening but figured Karpov knew the refutation. It’s not an easy one, because it’s not a natural move. Instead of 12. O-O, 12. Be3! as if …Bxc5 Qa4+ wins back the knight.


Break the rules

1. h4. One of two worst opening move.
4. Be3. Knight before Bishop.
6. Nd5. Don't move the same piece twice
8. c3. Don't lose material (pawn)
10. exd5. Control the center.
11. b3. Don't push pawns necessary for castling.
14. O-O-O. Don't castle in the direction of the attack.
15. d4. Blunder. Losing material. 
17. Kd3. Black to mate in 6!
24. Re2. Blunder. Losing material  . Push the King ahead of all the other pieces: King pusher.
25... Bd3. The bishop is both pinned and x-rayed! Losing bishop. (D)
26. Re7. When the going gets tough, the tough gets going! Time: White 1:06, Black 1:28
30-32: 3 pawns in a row! Bishop advantage 
is nullified.
40. Kd5. Finger failure. I wanted to move Kb6.
41. Kc5. Inaccuracy. I should've moved Kc6.
44. bxc8=Q+. Rxc8+ is better. The good is the enemy of the great.
45. b5. d7!! is better. Then trade rook+pawn for bishop, and promote pawn to Queen.
48. Kxb7 1-0. This would've been followed by Kc6, and White wins! Time: White 1:04, Black 0:31.

I played this game instinctively instead of calculations, hence very little loss of time. Black calculated normally, hence steady loss of time.

The text contains actual move number, but was “auto-corrected” by the post processing.

Note: too lazy to put backticks on individual moves.

1 Like

I recently saw this super-miniature pop up in a few places, so I figured I’d pass it on. Some may know it.

  1. d4 Nf6 (very normal so far)
  2. Nd2 e5 (both of these moves are not very good, but they’re plausible. This is a weird Budapest but the Nd2 instead of c4 probably helps to hold the extra pawn)
  3. dxe5 Ng4
  4. h3 (you really should move Ngf3, develop instead of moving pawns. Why?)

…Ne3! wins. The queen is trapped unless fxe3, which ruins White’s pawn structure. Oh. It also allows …Qh4-g3 mate.


On the subject of auto text formatting I see two solutions. Both are kind of hacky.

I put backticks around the moves for this short game I played a long time ago. I felt bad for my opponent. He was in high school and I was in college, and we were chatting before pairings were announced. He was nice. He talked about finally understanding and studying some 15-move-deep lines in the Sicilian. Then this happened.

1. e4 e6 2. d4 d5 3. e5 # this is the French Advance. How does Black avoid getting strangled?
3. ...c5 # oh no! Attacking the base of the pawn chain! How does White’s center avoid collapse? And how does Black keep heaping on pressure?
4. c3 Nc6 5. Nf3 Qb6! # It now makes sense for White to develop his king’s bishop. Bd3!? is a tricky gambit line, and Be2 is safe but a bit boring. a3 is also possible, with fascinatingly weird pawn structures resulting after …c4. But White has no time to reinforce the queen’s pawn as planned below.
6. b3? cxd4 7. cxd4 Bb4+ #disruptive–no good way to block!
8. Bd2 # Ke2 was the only way to save the pawn. It’s bad could be quite annoying in the “how did I not punish his horrible move” way if Black doesn’t recognize they should hit the center with …f6.
8. ... Nxd4 9. Be2?! #Surprisingly, chess engines give Bb5+!? as a way to annoy Black and force a decision. The point is that the opposite-colored bishops make things weird, or Qxb5 Nxd4 gets minimal initiative. White is right to develop the bishop, but d3 would have been better and more aggressive. Still, once you get punched, you can go into a bit of a shell, sometimes.
9. ... Bd7 10. O-O? # The engines say Bd3 is best. Bxb4 just helps Black simplify. Nxd4 Qxd4 Bxb4 Qxa1 loses material but may be a decent practical chance since Black can’t castle.
10. ... Nxe2+ 11. Qxe2 Bb5 # White resigns. Too much material is lost.

He was 1400 here, and he got to 1800, I think. Reason for optimism after your next tough loss.

Oh: the other way?

1, e4 e6 2, d4 d5 3, e5
3, … c5, etc.

etc. This is not so hot for cut-and-paste purposes, as search-and-replace may make annotation grammar weird. But it’s a possibility. Looking on the Discourse meta forum, apparently disabling some sorts of markdown is not supported.

1 Like

Of the importance of minding the little pawns after castling the big king.

The computer marked move 24 as blunder, as it suggests to capture with a knight. It wouldn’t have led to the beautiful finish of this game.


Congrats, again!

As for me, I meant to post a bit earlier … so let’s have a couple games as catch-up.

People may find positional chess boring, and masters are divided as to how revolutionary Nimzowitsch’s My System was. But I always found this game of his hard to forget.

He attacks the c4-square, and when White just barely stops him from winning the c4-pawn, the knight sits there anyway, helping freeze White’s position. White is utterly helpless and resigns on move 23. It’s worth noting mate is nowhere close, but with fewer pieces on the board and Black about to win a bunch of pawns and with more active pieces, Black has no hope at the advanced level.

In fact, if you are a relatively beginning player, you may see this play out with 7 pawns vs 6. It’s “only” a pawn but a pretty clear win for the side with the material edge. You should you could beat even Magnus Carlsen if there’s nothing unusual going on (e.g. passive king or doubled/tripled pawns.) Getting to that sort of position against Carlsen? Quite another matter for even the best of us!

PGN for Mattison-Nimsowitsch
1. d4 Nf6 2. c4 e6 3. Nc3 Bb4 4. Nf3 Bxc3+ 5. bxc3 d6
6. Qc2 Qe7 7. Ba3 c5 8. g3 b6 9. Bg2 Bb7 10. O-O O-O 11. Nh4
Bxg2 12. Kxg2 {?} Qb7+ {!} 13. Kg1 Qa6 14. Qb3 Nc6 15. Rfd1
Na5 16. Qb5 Qxb5 17. cxb5 Nc4 18. Bc1 a6 19. bxa6 Rxa6
20. dxc5 bxc5 21. Ng2 Nd5 22. Rd3 Rfa8 23. e4
Ne5 0-1

But here’s another – a crazy tactical battle that lasts longer. Don’t worry about analysis. Just enjoy the roller coaster. The end will surprise you, but you’ll see what’s going on if you sit and think about it.

I always forget the players’ names but I remember the moves! The opening is now somewhat joking called the Frankenstein-Dracula Variation in some circles.

PGN for Kupferstich-Andreasen
1.e4 e5 2.Nc3 Nf6 3.Bc4 Nxe4 4.Qh5 Nd6 5.Bb3 Nc6 6.Nb5 g6
7.Qf3 Nf5 8.Qd5 Nh6 9.d4 d6 10.Bxh6 Be6 11.Qf3 Bxb3 12.Bxf8
Ba4 13.Bg7 Rg8 14.Bf6 Qd7 15.Na3 Nxd4 16.Qh3 Qxh3 17.Nxh3
Nxc2+ 18.Nxc2 Bxc2 19.Rc1 Be4 20.Ng5 Bxg2 21.Rxc7 Bxh1 22.Nxf7
Bd5 23.Nxd6+ Kf8 24.Bg5 Rh8 25.Bh6+ Kg8 26.Rg7+ Kf8 27.Rc7+
Kg8 28.Nc8 Bc6 29.Rg7+ Kf8 30.Rxb7+ Kg8 31.Rg7+ Kf8 32.Rxa7+
Kg8 33.Rxa8 Bxa8 34.Nd6 1-0
Ending position

Why is this lost for Black? Black has no moves. Black also cannot guard the dark squares!

As for how best to win? White could just walk the a- and b-pawns up, but that lacks flair. Instead, moving the king to e7 puts Black in trouble. Ne4 or Ne8 threatens mate. If the Black bishop goes to c6 or g6 (after moving the g5 pawn) then White can move the e-pawn, forcing the Black bishop to move. Black could move pawns, but White can just take.


The knight moves at the end seems to be unnecessary complication. Is there a reason that was done?

1 Like

That was a super cool winning situation, effectively reducing the board to white’s king and connected passers vs. an impotent light-squared bishop! Thanks for sharing.


How to get 98% game:

Win in the opening. :grin: