The 2019 Canadian Memory Championship

The 2019 Canadian Memory Championship

By Matthew S. Wilson

I almost didn’t go.

You have to train to get better. This is certainly true in the long run. But it’s not always true in the short run. I had been training for about 2 hours per day, but right before the tournament, my results suddenly plummeted. For about a week, I was 10-20% worse in everything. I didn’t understand why. It was so frustrating. I had been thinking about going to the tournament, but I didn’t want to drive all the way to Montreal and then suck. A few days before the event, I tried to memorize a deck of cards and failed. Again. I decided not to go.

But on the next day, I looked at the situation from a different perspective. I had been making a lot of progress this year. Even on a bad day, I was still better than the Matt Wilson who qualified for the US championship finals. So if I had been willing to go to US championship in the past, then I should also be willing to go to this tournament. I signed up. Three days before the championship, my poor results vanished as quickly and mysteriously as they had arrived. In training, I posted my highest scores ever. The tournament wasn’t going to be a disaster!

Since I’m not a Canadian, I couldn’t become the Canadian Champion. But there was a section for foreigners (the Open section) that I could win. I considered myself to be the clear favorite unless Livan Grijalva showed up. He is a top American mnemonist who sometimes goes to Canadian tournaments. I can compete with him in some of the disciplines, but he is much better at memorizing cards and numbers. In the Canadian section, Braden Adams should win easily. Since we were in different sections, we weren’t competing against each other, but of course I still hoped to outperform him (though I knew that was very unlikely).

I drove to Montreal a day before the event. I might have gotten a Canadian Border Patrol agent interested in memory techniques. He enjoyed testing me a bit (e.g., “What was the second question that I asked you?”). I was afraid that he wouldn’t believe me when I said I could memorize a deck of cards in under 5 minutes. But he let me through. Maybe he’ll be at the competition next year.

Since my entry was so last minute, I wasn’t fully prepared for all 6 events. Luckily, Livan wasn’t there, so I would probably win the Open section anyways. The first discipline was Images and I wasn’t quite ready for it. A few years ago, there was a split in the world of memory competitions. One set of tournaments used abstract black and white images, like the ones here. The Canadian championship and many other tournaments used pictures of everyday objects like bicycles or computers. But all of the other events were the same in both kinds of tournaments. That’s why I focused on the other events. As a result, I’m not very good at memorizing pictures. A talented beginner might be able to beat me. We get 5 minutes to memorize as many rows of pictures as possible. Each row has 5 pictures. Then in the recall session, each row is scrambled and we have to put them back in the proper order. I attempted 25 rows, which is typical for me. One of the competitors had requested an extra sheet, since he planned on memorizing more than 40(!) rows. I thought it was going well, but during recall, I found myself guessing a couple of times. My guesses must have been wrong; I only got 21 of the rows right. But I didn’t know that during the tournament. The results weren’t posted until a few days later.

My poor preparation affected me again in Words and in Numbers. Usually when I train, I have 5 minutes for each of these two disciplines. But in the Canadian championship, we had 10 minutes. I got off to a good start in numbers, but then I struggled to maintain my focus in the second half. I wasn’t used to having an extra five minutes. After attempting and reviewing the first 288 digits, I still had an extra minute or two. Should I review more? Or memorize more? If I had just picked one and stuck with it, I probably would have been okay. Instead, I tried a bit more review, changed my mind, switched to more memorization, and changed my mind again. Combine this with my loss of focus in the last 100 digits, and we have a disappointment. I got 17 rows of 12 digits correct and half of the last row. This yields a raw score of 210 (17x12 + 0.5x12 = 210). This is barely any better than my results with 5 minute numbers. The other 7 rows had mistakes; the extra 5 minutes had been wasted. I was planning on memorizing 90 words in 10 minutes and found myself in the same situation: what do I do with this extra time? Again I was indecisive. I tried an extra column of 10 words mixed with review. I felt okay, but then during recall I realized that several details were hazy. For example, I wasn’t sure if one of the words was “magic” or “magician.” Then there was a part where I knew the word was similar to “cut,” but it wasn’t actually “cut.” I went through as many synonyms as I could think of, but none of them felt right. Mistakes like this can be quite costly due to the scoring system, so I only scored 72 points out of the 100 I attempted. Only slightly better than what I usually do in 5 minute words.

I was ready for Names and Faces. We had 5 minutes to memorize as many as we could. Then during the recall session, we saw just the faces and had to write down their names. I had trained for this. Usually I spend the first four and a half minutes on memorization and then quickly review the first part for the last thirty seconds. Inexplicably I got too cautious and spent more time on review. I think I remembered nearly everything I tried to memorize, but I didn’t memorize enough. Just 45 names! Usually I get more than 50. There is no penalty for writing the wrong name, so there is no reason for spending a bunch of time on review trying to get everything perfect. You probably see the pattern here: I don’t handle the pressure well. I have to be AMAZING in training so that even when I collapse in a competition, my results are still respectable.

I wasn’t ready for the Exam Cramming discipline, an event unique to the Canadian tournaments. They give you notes and 15 minutes to study. It has information about made up historical events, geography, and other subjects. It’s all fictional to ensure that no one has an advantage from studying these areas ahead of time. I had attempted this sample exam at home. My practice result was a disaster: just 51%. A lot of it was just hard to memorize and I got flustered. The format of the practice test took me by surprise as well. To make matters worse, my printer had run out of color ink, and trying to distinguish the black and white photos of different praying mantises was tough. It’s very hard to practice for this event, so I just hoped to avoid a complete meltdown. Thanks to the practice exam, at least I knew what to expect. When I came across something that looked hard to memorize, I just skipped it instead of getting flustered. Then I came back to it later. Again I found myself with extra time. But I stuck to the plan: don’t try the bonus questions, just review again. All the pictures were in color, which helped. I felt good about my performance, but in view of how badly the practice test had gone, I didn’t want to get too optimistic. If I got 70%, I was happy.

I scored 90%!!! I had not dared to even hope for a score that high. Even Braden Adams, who had dominated the previous 4 events, couldn’t keep up with that. It feels great to beat an entire country at something.

The last event was Speed Cards. We had to memorize a deck in 5 minutes or less. I was a bit nervous about this discipline. Even though I had largely recovered from the disastrous results I had a few days before, I was still wobbly in cards. Just a day before the tournament, I had screwed up a deck. Maybe I’m still traumatized from botching the first deck at the 2018 US championship. There were 2 attempts, but what if I messed up both times? I decided to be extra slow and careful in the first round. I memorized the deck and then reviewed it once. That took 70 seconds. Slow by my standards, but it dazzled that Canadians. While I waited for everyone else to finish, I went through the deck in my head. I was pretty sure I had it. During the recall session, I took another deck and rearranged it to match my recollection of the first deck. I double checked and was confident that I had it. We had 5 minutes to recall, but I finished early. I didn’t want to start second guessing myself and ruin a deck that was almost surely correct. But when the judge was checking my results, I suddenly got nervous near the end. It was outside of my control now. But I had recalled it perfectly. Now that I had secured a decent result in round 1, I could afford to take risks with the second deck. By “take risks,” I mean going faster, which makes a mistake more likely. Since only your best result counts, this is the right strategy. I was hoping to memorize it in under a minute – something that I routinely did at home, but doing it during a competition with all the pressure is another matter. I zipped through the deck. But during review I suddenly realized that I didn’t know if a certain card was the 10 of Hearts or the 10 of Diamonds. I lost a few seconds double checking, but it didn’t throw me off. I finished in 62.333 seconds. If I hadn’t stumbled, maybe I would have done it in under a minute. But I was happy with this time. Usually I don’t perform well under pressure, but here I was just a few seconds slower than I was at home. I reviewed the deck in my head while waiting for the rest of the competitors to finish. A few details were hazy, but I thought I would probably get it. What I like about cards is that you can use the process of elimination. For instance, in this video this video, I wasn’t sure if a certain card was the King of Hearts or the King of Diamonds. I went through the rest of the deck in my head and realized that the King of Hearts came later, so this card must be the King of Diamonds. The same trick worked in Canadian Championship. Everything fell into place. I had recalled the deck perfectly. Meanwhile, Braden took 88 seconds to memorize his deck. Everyone else was much slower. Once again, I had beaten an entire country at something!!

I clobbered my rivals in the Open section (full results here). Though I was ineligible to be the Canadian Champion, I still hoped that I could beat the top Canadian. Unfortunately, Braden’s dominance in the first 4 events outweighed my victories in Exam Cramming and Speed Cards. He still would have won even if we disregard all the extra bonus he got that I was ineligible for. I finished 2^nd overall. We’ll get ‘em next time!

IM Jay Bonin defeats e4stat

I spent Labor Day weekend at the New York State Championship. My first game was against a well-known IM. It didn't go very well. Here's how to beat a 2100:


Click here if the full games are not displaying properly

[Event "New York State Championship"] [Site "Albany"] [Date "2019.08.31"] [Round "1"] [White "Bonin, IM Jay"] [Black "Wilson, Matthew"] [Result "1-0"] [ECO "A61"] [WhiteElo "2408"] [BlackElo "2120"] [Annotator "Wilson,Matthew"] [PlyCount "117"] 1. d4 Nf6 2. Nf3 e6 3. c4 c5 4. d5 exd5 5. cxd5 d6 6. Nc3 g6 7. Bg5 {I knew about this move, but I did not consider it to be dangerous. Wrong assessment - the pin turns out to be quite annoying. I didn't have any prep for this line, so I just played natural moves. That didn't go very well.} Bg7 8. Nd2 O-O 9. e4 Re8 10. Be2 (10. f3 $4 Nxe4 $1 { But of course you don't catch IMs with a simple trap like this}) 10... Na6 11. O-O Nc7 12. a4 Rb8 13. f4 a6 14. e5 $6 ({I was anticipating} 14. a5 Bd7 { . The engine loves White's position, but I felt that I still had chances. Nc4 can be met by ...Bb5}) 14... dxe5 15. Nde4 h6 $2 (15... exf4 $1 {It's counterin tuitive to open the f-file, but this is best. In variations with Rxf4 and d5-d6, Black can pick up a tempo on the rook with ...Ne6 or ...Ncd5. And in sharp positions, every tempo matters a lot.} 16. Rxf4 $2 (16. Bc4 {is the engin e's choice, but clearly this wasn't what White had in mind when he played 14. e5. It leads to complications that eventually fizzle out:} Bf5 (16... b5 17. d6 Ne6 18. axb5 axb5 19. Bxe6 (19. Nxb5 Nxg5 20. Nxg5 Rf8 21. Rxf4 Bh6 22. h4 Bf5 23. Nxf7 Rxf7 24. Rxf5 gxf5 25. Ra7 Kh8 26. Rxf7 Be3+ 27. Kf1 { (27.Kh1 or 27.Kh2 would walk into a mating net)} Ne4 28. Qd5 Ng3+ 29. Ke1 Qa5+ 30. Nc3 Qa1+ 31. Nd1 (31. Qd1 $2 Qxb2 {and Black wins}) 31... Qa5+ 32. Nc3 { with a repetition}) 19... Rxe6 20. Rxf4 Bb7 21. Qf1 (21. d7 Bxe4 22. Nxe4 Qb6 { and Black's defenses hold together}) 21... Bxe4 22. Nxe4 Rxe4 23. Rxe4 Qxd6 24. Rf4 Nh5 25. Rxf7 Qd5 26. Re7 Rf8 27. Qc1 { and Black has enough compensation for the Exchange}) 17. d6 Ne6 18. d7 Rf8 19. Bxe6 fxe6 20. Qd6 h6 21. Nxf6+ Bxf6 22. Bxf6 Qxf6 {is equal according to the engine. Black is 2 pawns up, but his position is loose and the passer on d7 is very strong. One sample variation is} 23. Rxf4 e5 24. Qd5+ Rf7 25. Ne4 Bxe4 26. Rxf6 Bxd5 27. Rxg6+ Kh7 28. Rd6 Rd8 29. Rxd5 Rdxd7 30. Rxe5 { with a drawn rook ending}) 16... Bf5 17. g4 {During the game, this move deterred me from playing 15...exf4. However, Black has a defensive resource that is hard to see from afar. Unbelievably, White's e2-bishop is undefended in some variations, and that changes everything!} (17. Nxc5 Qd6 $1 { Black breaks the pin with tempo and the d5-pawn will fall}) (17. d6 Ne6 18. d7 Nxg5 $1 {Stockfish's move. My notes suggested 18...Rf8 and White's attack runs out of gas, but this is stronger} 19. dxe8=Q+ Qxe8 20. Nxg5 Qe3+ $1 { and White loses material}) (17. Nxf6+ Bxf6 18. Bxf6 Qxf6 19. g4 Qg5 $1 { I think I saw this during the game. White can try to seize the f-file as compensation, but it doesn't work:} 20. Qf1 Rxe2 $1 21. Nxe2 { and now Stockfish likes} Bd3 $1 {This is much better than my suggested 21... Bxg4. White is tied up and d5 is going to drop off.}) 17... Bxe4 18. Nxe4 Ncxd5 19. Bxf6 (19. Rxf6 Rxe4 $1 {and White doesn't have any good discoveries.} 20. Rxg6 (20. Re6 Rd4) (20. Rf5 Bd4+ $1 {shields my knight with tempo}) 20... Qd7 $1 {and the rook is stranded}) (19. Nxf6+ Bxf6 20. Rxf6 Nxf6 21. Qxd8 Rbxd8 22. Bxf6 {The e2-bishop is hanging, but White replies with a counterattack against my rook. However, after} Rd6 $1 {one of the bishops will fall.}) 19... Nxf6 20. Qxd8 Rbxd8 21. Nxf6+ Bxf6 22. Rxf6 { and now the undefended bishop on e2 drops off:} Rxe2 {and Black wins}) ({ During the game, I had been counting on} 15... Bf5 {, but then I noticed 16.d6. I didn't like what I saw. After rejecting 15...exf4 and 15...Bf5, I played 15.. .h6? without much further thought and missed a simple tactic.} 16. d6 ({ During the game, I saw that} 16. fxe5 {isn't dangerous.} Rxe5 17. Nxf6+ Bxf6 18. Bxf6 Qxf6 19. g4 Qg5 20. h4 Qe3+) 16... Ne6 (16... Ncd5 17. Nxd5 Bxe4 18. d7 $1 {Stockfish's move} (18. fxe5 $2 Bxd5 19. exf6 {is in my notes. Due to tactics, neither one of Black's bishops is hanging. However, I couldn't find a satisfactory move for Black.} Bc6 (19... Be6 20. d7 $1 Bxd7 { (both 20...Qxd7 and 20...Rf8 run into 21.fxg7)} 21. fxg7 Qxg5 22. Qxd7 { Again, the e2-bishop is hanging, but here that doesn't save Black. The clergyman is immune because of 23.Qxf7#}) (19... Re5 { I thought that this just lost to} 20. Bf4 {, but Stockfish sees further:} Qxf6 21. Bxe5 Qxe5 {and actually Black is fine}) 20. Bc4 { and inexplicably I missed the obvious} Bf8) 18... Re6 19. fxe5 Bxd5 20. exf6 Bf8 (20... Rd6 21. Bg4 $1 (21. Bf4 {lets Black off the hook.} Qxf6 { Black can afford to give back the Exchange since he will round up the d7-pawn.} ) 21... Bf8 22. Bf4 Qxf6 23. Bxd6 Qxd6 24. Rf2 $1 {followed by Rd2 ensures the survival of the d7-pawn. Black is down material and busted.}) 21. Qxd5 Rxe2 22. Rae1 $1 {Black loses control of the e-file. White will follow up with Re8 and there isn't much that I can do about it.}) 17. Bxf6 (17. fxe5 Nxg5 18. Nxf6+ Bxf6 19. exf6 {spooked me during the game because it looks like} Qxf6 20. g4 { wins a piece. However,} Qd4+ $1 {saves the day}) 17... Bxf6 18. d7 Rf8 19. Nxf6+ Qxf6 20. g4 { Stockfish's choice. I gave this move a question mark in my notes.} (20. Nd5 Qg7 21. g4 Be4 22. fxe5 Qxe5 23. Nf6+ {In my notes, I thought White was doing very well here. Knight on f6, pawn on d7, control of the f-file, etc. But Stockfish is unimpressed. Black also has his trumps. White's king is exposed and the e6-knight is itching to go to d4. And the e4-bishop is a strong piece}) 20... Nxf4 {Here my notes correctly observe that Black gets too much compensation after 21.gxf5 Qg5+. However, I overlooked} 21. Rxf4 $1 exf4 22. gxf5 { White's monster on d7 outweighs Black's material advantage.}) 16. Bxf6 Bxf6 17. d6 Ne6 18. d7 $1 Bxd7 19. Qxd7 Bg7 20. Nd5 $2 { Stockfish frowns on this move. Apparently Black is still alive after 20...Kh8!} Qxd7 $2 {Now it's over.} (20... Kh8 21. Qxd8 Rexd8 {and Black picks up a third pawn for his piece. However, White remains in control with a big advantage.}) 21. Ndf6+ Bxf6 22. Nxf6+ Kg7 23. Nxd7 Rbd8 24. Nxe5 Nxf4 {A little trick that recovers some of the material, but White's bishop outweighs the pawns.} 25. Rxf4 Rxe5 26. Rd1 Rxd1+ 27. Bxd1 b6 { Or else 28.a5 fixes the pawn on a light square} 28. Bb3 f5 29. Bc4 { Well played. Now my queenside majority loses its mobility.} a5 30. Kf2 Kf6 31. Rf3 Re4 32. b3 Ke5 33. Re3 g5 34. g3 f4 35. gxf4+ Kxf4 36. Rh3 { Winning a pawn. But I found some counterplay to keep the game going.} Rd4 37. Rxh6 Rd2+ 38. Ke1 ({I was hoping for} 38. Be2 Rb2 39. Rxb6 c4 $1 {The engine says that White is still winning, but this would give me real counterplay.}) 38... Rb2 39. Kd1 g4 40. Kc1 Rg2 41. Bd5 Re2 42. h3 { It looks like a mistake at first glance, but the IM has worked everything out.} g3 43. Rg6 Rh2 44. Rg4+ Ke5 45. Bb7 Rxh3 46. Bg2 {Black can't keep his passer. So 42.h3 was not a blunder - it just liquidated the kingside pawns.} Rh6 47. Rxg3 Rh4 48. Bf1 {Stopping ...c4} Rh1 49. Rf3 Ke4 50. Rf7 Kd4 51. Kd2 Rh2+ 52. Be2 Rh3 53. Rd7+ Ke5 54. Bd3 Rh2+ 55. Kc3 Rh4 56. Rb7 Rh6 57. Rg7 { Threatening to force a trade on g6. I stop it} Kf6 58. Rh7 { But now I'm forced to trade on h7 instead.} Rxh7 59. Bxh7 {I resigned.} 1-0

After taking a half point bye, I scored a nice win over a master.

[Event "New York State Championship"] [Site "Albany"] [Date "2019.09.01"] [Round "3"] [White "Wilson, Matthew"] [Black "Howard, NM Dean"] [Result "1-0"] [ECO "B19"] [WhiteElo "2120"] [BlackElo "2005"] [Annotator "Wilson,Matthew"] [PlyCount "81"] 1. e4 c6 2. d4 d5 3. Nc3 dxe4 4. Nxe4 Bf5 5. Ng3 Bg6 6. Nf3 Nd7 7. h4 h6 8. h5 Bh7 9. Bd3 Bxd3 10. Qxd3 e6 11. Bf4 Qa5+ 12. Bd2 Bb4 13. c3 Be7 14. c4 Qc7 15. O-O-O Ngf6 16. Kb1 O-O 17. Ne2 e5 $6 {At first I thought , did Black just equalize? But then my knights begged me to occupy f5.} 18. Ng3 $6 {Right idea, wrong execution} (18. dxe5 Nxe5 19. Nxe5 Qxe5 20. Bc3 { is more accurate, since Black doesn't have the ...Qf4 defense that we will see later. True, Black can force a queen trade with} Qe4 {, which greatly reduces White's attacking prospects. But Stockfish points out that White retains a big advantage after} 21. Ng3 {. The f5-square remains weak.}) 18... Rfe8 (18... Rfd8 $6 19. Qe2 $1 (19. Nf5 e4 $2 (19... Bf8 { keeps White's advantage to a minimum}) 20. Nxe7+ Kf8 21. Qa3 $1 c5 {and White h as lots of ways to win. My notes suggest 22.dxc5, which should do the job, but Stockfish prefers the simple} 22. Nf5 exf3 23. Qxf3 { White is up a pawn and his attack is just getting started}) 19... Bd6 20. Nf5 { In my notes, I thought that Black was okay after} exd4 (20... e4 21. N3h4 { is very good for White}) {But I missed} 21. Bxh6 $1 gxh6 (21... Re8 22. Qc2 gxh6 23. Nxh6+ Kf8 24. Nxf7 {transposes}) 22. Nxh6+ Kf8 23. Nxf7 Re8 24. Qc2 Kxf7 25. Ng5+ Ke7 (25... Kf8 26. Qg6 Re7 27. h6 $1) 26. Rhe1+ Kd8 27. Rxe8+ Kxe8 28. Qg6+ Ke7 29. Qf7+ Kd8 30. Ne6+ {and White wins}) (18... exd4 { is suggested by the engine} 19. Nxd4 Rfe8 20. Ngf5 Bf8 { and now the prophylactic} 21. Qc2 {preserves an edge for White}) 19. dxe5 Nxe5 20. Nxe5 Qxe5 21. Bc3 (21. Rhe1 {was my original intention, but I feared that Black could swap off all the rooks in some lines. It's not hard to imagine ... Rad8 and ...Bf8 in the near future.} Qc7 (21... Rad8 $2 22. Qxd8 Bxd8 23. Rxe5 Rxe5 {and I missed that I can just grab a pawn with} 24. Bxh6 {(thanks, Stocky) }) (21... Qc5 {is Stockfish's suggested defense.} 22. Bc3 { pursues the same plan as in the game, but it's not effective here:} Rad8 (22... Nxh5 $4 23. Re5) (22... Qxf2 $2 23. Rf1 { followed by Nf5 with overwhelming compensation for the pawn}) 23. Qf3 Rxd1+ 24. Rxd1 {and now Black can clip an important pawn with} Qxc4 { , which also takes control of d5}) 22. Bc3 (22. Nf5 {and Black has time to play } Bf8 {without fearing Bxf6}) 22... Red8 {(clearing e8 for the knight)} 23. Qe2 {My notes recommended 23.Qf3, but Stockfish prefers this move} Rxd1+ 24. Rxd1 Ne8 {Hoping to contest the f5-square with ...Nd6} 25. Be5 $1 { And now we see why it was important to play 23.Qe2.} Qb6 (25... Bd6 26. Rxd6) 26. Bxg7 $1) (21. Qf3 { is the best chance to preserve an edge, according to Stockfish. But after} Qe6 22. b3 Bc5 23. Rde1 Qg4 {, Black is only slightly worse}) 21... Qc7 $2 (21... Qg5 22. Nf5 {and the g2-pawn is obviously taboo}) (21... Qf4 {was the only move } 22. Nf5 Qxf2 {I spent a bunch of time trying to refute this} ({I overlooked} 22... Qe4 {during the game. In my notes, I thought that it solved most of Black's problems, but Stockfish thinks that White still has a clear advantage}) (22... Ne4 $1 { is Stockfish's solution and it refutes 22.Nf5. The tactical justification is} 23. Nxg7 $4 Nxc3+ 24. Qxc3 Bf6 { But if 22.Nf5 is unplayable, then White has nothing}) 23. Rdf1 Qc5 { The g2-pawn is still immune} (23... Rad8 24. Nxe7+ $1) 24. Qg3 $1 { and Black is defenseless}) 22. Nf5 Rad8 23. Qf3 { Setting up tactics against the f6-knight} Kh7 { It's hard to find a decent move for Black} (23... Bf8 24. Nxh6+ $1) (23... Rxd1+ 24. Rxd1 Rd8 25. Re1 $1 Rd7 26. Nxh6+) 24. g4 Rxd1+ {Relatively best} ( 24... Bf8 25. Rxd8 {Stockfish's move} (25. Nxg7 {is similar to the game}) 25... Qxd8 (25... Rxd8 {meets a similar fate}) 26. g5 $1 {This idea is more powerful with the White rook on h1. That's why it's best for Black to swap rooks.}) 25. Rxd1 Bf8 {Falling into a tactic, but Black was already busted. All he can do is decide which tactic he wants to walk into.} (25... Rd8 26. Re1 Rd7 { Dreaming of playing ...Ne8 to hold his position together} 27. Nxg7 $1 { (Stockfish)} ({My notes suggest} 27. g5 $1 hxg5 28. h6 $1 { , which also does the job}) 27... Kxg7 28. Rxe7 $1 Rd1+ 29. Qxd1 Qxe7 30. Qd4 c5 31. Qf4 $1 {and Black is paralyzed}) (25... Bd8 {and now Stockfish loves} 26. Qd3 $1 ({My notes suggest} 26. Nxh6 { , which is also strong. The tactical justification is} Kxh6 27. g5+ $1 Kxg5 28. Bd2+) 26... Kg8 27. g5 $1 hxg5 28. h6 $1 g6 29. Ng7 Rf8 30. Ne6 $1 fxe6 31. Qxg6+ Kh8 32. Rxd8 $1 {Very impressive, but that's a lot to calculate}) 26. Nxg7 $1 (26. Nxh6 {seemed to be harder to calculate; Black has options like 26. ..Ne4. But the engine assures me that it was also good. Bizarrely, neither one of these moves is Stockfish's top choice. It likes 26.Kc2. Apparently Black can't improve his position, so White can take a few tempi to strengthen his hand before capturing on g7 or h6.}) 26... Bxg7 27. Qf5+ $6 { Objectively not a bad move, but it makes White's task far more complicated.} ( 27. Bxf6 Bxf6 28. Qxf6 Qe7 (28... Re6 {and now we see the idea behind the "precise" 27.Qf5+. In the Qf5+ variation, the Black king is on g8, so ...Re6?? loses immediately to Rd8+. But in the 27.Bxf6 line, ...Re6 is a viable option since Rd8 does not come with check.}) 29. Qf5+ Kg8 30. f3 Qe2 31. Qd3 { is much simpler than what happened in the game. White is up a pawn for nothing. }) 27... Kg8 28. Bxf6 Bxf6 29. Qxf6 Qe7 30. Qxh6 {I spent a long time on this move. If only the Black king were still on h7! Then I would have time to consolidate with 31.Qf5+ followed by 32.f3. I regretted playing 27.Qf5+ Kg8.} ( {The main alternative was} 30. Qxe7 Rxe7 {, leading a complex rook ending. White has an extra pawn, but will it be enough to convert?} 31. f3 Re3 32. Rd7 Rxf3 33. Rxb7 Rf4 (33... a5 {is Stockfish's move} 34. Rb6 f5 (34... Rf6 35. Ra6 {rounds up the a-pawn}) 35. Rxc6 fxg4 36. Rg6+ Kf7 37. Rxg4 Rh3 38. Kc2 Rxh5 { and Stockfish keeps wanting to play the anti-positional} 39. a3 {, which allows ...a4. But the engine has calculated that everything works out in the end:} Rh2+ 40. Kb3 a4+ 41. Kxa4 Rxb2 42. Re4 {cutting off Black's king} h5 43. c5 { and White wins}) 34. b3 ({My notes dismiss} 34. Rxa7 {because of} Rxc4 { ; the g-pawn isn't running away. But Stockfish prefers this over 34.b3, since it gives White a head start in the pawn race. Its main line continues} 35. a4 Rxg4 36. a5 f5 37. Kc2 f4 $2 (37... Rg1 {is a slower way to die}) 38. a6 f3 39. Kd2 Re4 40. Ra8+ Kg7 41. a7 f2 42. Rg8+ $1) 34... Rxg4 (34... a5 35. Rc7 Rxg4 ( 35... Rf6 36. Ra7 {- one of Black's queenside pawns drops off anyways, and he doesn't get the g4-pawn in return}) (35... f5 {is Stocky's choice, which initially makes no sense. The f-pawn is destined to be a mighty passer; why exchange it off for a g-pawn that was doomed to die for nothing?} 36. gxf5 Rxf5 37. Rxc6 Rxh5 {And now we see that there was a method to Black's madness. His goal was to create an outside passer on the h-file rather than the f-file.} 38. Rb6 Kf7 39. Rb5 Rh1+ 40. Kb2 h5 41. Rxa5 h4 42. a4 Kg6 43. Ra8 h3 44. Rh8 h2 45. a5 {and White prevails}) 36. Rxc6 Kg7 37. Rc5 { and if Black continues with 37...Rg5, then he loses the pawn ending}) 35. Rxa7 {White's a-pawn is very dangerous, but so is Black's f-pawn. During the game, this was unclear to me.} Rg2 { The best practical chance. Black tries to keep my king at bay.} (35... f5 36. Kc2 {and White wins since his king can stop Black's passer, while the Black potentate is powerless to do the same.}) 36. Kc1 (36. a4 f5 37. a5 $2 { So natural, but this throws away the win!} f4 38. a6 f3 39. Ra8+ Kg7 40. a7 f2 41. Rf8 Kxf8 42. a8=Q+ Kg7 {Amazingly, there is no way for White to win!}) 36... f5 37. a4 f4 38. Kd1 (38. a5 $4 f3 39. Kd1 Rg1+ $1 { My notes give 39...Re2, but this move from Stockfish wins on the spot}) 38... f3 39. Ke1 Rb2 40. Rb7 c5 41. a5 Ra2 42. Rb5 Rb2 43. a6 Ra2 44. Rxc5 Rxa6 45. Kf2 {and White will win}) (30. Qd4 Qe4+ 31. Kc1 c5 32. Qxc5 Qxg4 {is inferior to the game, and Stockfish points out that 31...Qe2 may be even better}) 30... Qe4+ 31. Ka1 Qd4 $2 ({I thought he was going to take on g4, but he said that he was afraid of opening up the g-file.} 31... Qxg4 32. f3 (32. Qc1 {was my int ention back when I played 31.Ka1, but then I noticed that this move is stronger}) 32... Qg3 {(the move he missed. Black keeps my rook off the g-file.) } 33. Qd2 $1 (33. Qc1 Re2 {and now I got bogged down trying to calculate} 34. Rd8+ (34. a3 {and now my notes like} Kh7 {- even awarding it an exclamation point - since it takes the sting out of Rd8. But Stockfish replies with} 35. Qc3 $1 {There is no ...Re1 or ...Qxf3 and Black has to find a way to deal with Rd7. Because 35.Qc3! is so strong, Stockfish recommends that Black play 34... Qe5 instead.}) 34... Kh7 35. Qb1+ Kh6 36. Rh8+ Kg7 37. Qh7+ Kf6 38. Qh6+ Kf5 { and White doesn't have a knockout}) (33. a3 Re1 { Without the rooks, White's attack runs out of gas}) {Since White can't convincingly punish 33.Qc1 Re2, it's best to play 33.Qd2. The tactical justification is} 33... Qxf3 $2 34. Rg1+ Kf8 (34... Kh7 35. Qg5 $1) 35. Qd6+ Re7 36. Re1 Qf6 37. Qd8+ {. If Black avoids this trap, the evaluation for 31... Qd4 is only a little bit worse than for 31...Qxg4. But White's job is significantly more difficult if the g-pawn falls. Miscalculate something and then the h-pawn may drop off and then White has nothing.}) 32. Qg5+ Kh8 33. Qh6+ {Repeating the position in order to get closer to the time control. I had about 20 minutes at this point.} Kg8 34. Qc1 { He resigned here - a pleasant surprise. The game might have concluded with} Qxf2 35. Qg5+ (35. h6 { was my intention during the game. Here Stockfish suggests} Qe3 {and White shoul d win, though rook endings are notoriously drawish. I thought he would play 35. ..Qf6, when White still has a lot of work to do}) 35... Kf8 36. h6 Re1 $2 37. Qd8+ Re8 38. h7 $1 Qh2 39. Qd6+ $3 Qxd6 40. h8=Q+ Ke7 41. Qxe8+ $1 1-0

2019 Fall Chess Classic

The St. Louis Chess Club is hosting another strong 10-player round robin. The forecast for the A Group:

Another GM beats e4stat

E4stat played in the Philadelphia Open. I won a prize and gained a few rating points, but it was not because of this game. GM Bryan Smith gives a lesson on how to beat 2100 players; hopefully you can learn something from this.

[Event "Philadelphia Open"] [Site "Philadelphia"] [Date "2019.06.29"] [Round "2"] [White "Wilson, Matthew"] [Black "Smith, GM Bryan"] [Result "0-1"] [ECO "B57"] [WhiteElo "2112"] [BlackElo "2512"] [Annotator "Wilson,Matthew"] [PlyCount "92"] 1. e4 c5 2. Nf3 d6 3. d4 cxd4 4. Nxd4 Nf6 5. Nc3 Nc6 6. Bc4 Qb6 7. Nb3 e6 8. O-O Be7 9. a4 { This move is less effective if Black hasn't played ...a6. 9.Be3 is better. I forgot my prep. } O-O 10. Be3 (10. a5 {is more consistent}) 10... Qc7 11. f4 b6 (11... d5 { is playable, according to the engine:} 12. exd5 Nb4 $1) 12. Be2 Bb7 13. Bf3 Rac8 14. g4 (14. Nb5 Qb8 15. c4 {doesn't work} a6 {Stockfish's recommendation} (15... Nb4 { is in my notes, but Stockfish finds} 16. Bd2 { and Black shouldn't go pawn grabbing:} Rxc4 17. Na3) 16. Nc3 Nb4 17. Nd2 Rfd8 { and ...d5 is coming}) 14... Nb4 15. g5 Nd7 ({ Stockfish briefly likes a weird tactic:} 15... Nxe4 16. Bxe4 Bxe4 17. Nxe4 Nxc2 {. However, it soon realizes that this doesn't end well for Black} 18. Rc1 Nxe3 19. Rxc7 Rxc7 20. Qe2 Nxf1 21. Kxf1) 16. Bg2 Rfd8 17. Rf2 {A multi-purpose move. White can double on the f-file, swing the rook over to d2, or play Qh5 without dropping the c2-pawn. But these plans are too slow. Stockfish recommends 17.Nd4.} Nc5 18. Rd2 {Preventing ...d5} ({ After the game, I thought that} 18. Nd4 { was more accurate, but in my notes I realize that} d5 19. e5 Ne4 { is still miserable for White}) (18. Nb5 Qb8 19. Nxc5 bxc5 {is Stockfish's recommendation. It thinks that the position is equal since ...d5 is far less dangerous now}) 18... d5 {Oh...I guess I didn't prevent it after all.} ({ Stockfish points out that} 18... Nxb3 19. cxb3 d5 {is more accurate, since it prevents(!) White from playing c2-c3 in some variations}) 19. e5 $6 ({ I saw his trick:} 19. exd5 $6 Nxb3 20. cxb3 Bc5 $1 { and Black regains the pawn favorably.} 21. Bxc5 Qxc5+ 22. Kh1 Nxd5 23. Nxd5 Bxd5 24. Bxd5 Rxd5 25. Rxd5 exd5 {and Black is much better}) (19. Nb5 { was the only way to maintain the balance.} Qb8 20. c3 { Stockfish's recommendation} ({My notes dismissed 19.Nb5 due to} 20. e5 Ne4 { but Stockfish thinks that it's equal after the awkward 21.Re2. Instead, it thinks that Black should meet 20.e5 with 20...Nxb3 followed by ...Ba6, with a slight advantage.}) 20... Nba6 21. e5 (21. exd5 Nxb3 22. Qxb3 Bc5 $1) 21... Ne4 {and the engine goes to triple zeros land}) 19... Nxb3 20. cxb3 Bc5 21. Bxc5 $2 ({During the game, I saw} 21. Bd4 Bxd4+ 22. Rxd4 Qc5 $1 23. Kh1 Nc2 $1 { and White is busted}) (21. Bf2 Bxf2+ (21... d4 { leads to complications and throws away most of Black's advantage} 22. Nb5 Qe7 ( 22... Qd7 23. Nxd4 Bxg2 {(Stockfish's move)} (23... Nc2 {is in my notes, but I missed that} 24. Bxb7 $1 {wins for White}) 24. Kxg2 Bxd4 25. Bxd4 Rc2 26. Rxc2 Nxc2 27. Qxc2 Qxd4 28. Rf1 {Black retains the initiative, but White is hanging on}) 23. Nxd4 Bxd4 24. Bxd4 (24. Rxd4 $2 Rxd4 25. Bxd4 Rd8 26. Rc1 (26. Bxb7 Qxb7 27. Rc1 ( 27. Qd2 Nc6 28. Qg2 Rxd4 29. Rc1 Nd8 {and Black keeps his extra material}) 27... Qe4 $1 (27... Nc6 $2 28. Qf3 $1) 28. Rc4 Nc6 {and Black wins a piece}) 26... Bxg2 27. Kxg2 Qd7 (27... Qb7+ $2 28. Qf3 $1 { and tactics hold White's position together}) 28. Qg1 Qd5+ { and at the very least, Black can recover his pawn with a big advantage}) 24... Bxg2 (24... Nc2 25. Rxc2 Rxc2 26. Qxc2 Rxd4 {and White is okay}) 25. Kxg2 Qb7+ 26. Qf3 {and White hangs on}) 22. Rxf2 Qc5 (22... d4 $2 23. Nb5 Qe7 24. Bxb7 Qxb7 25. Nd6) 23. Nb5 Ba6 24. Bf1 { is bad for White, but it's better than what happened in the game}) 21... Qxc5+ $1 ({I was expecting} 21... bxc5) 22. Kh1 Qe3 23. Nb5 (23. Rd4 Nc2 (23... Nc6 $1 {is a much simpler refutation}) 24. Rd3 { leads to some very entertaining variations:} Qxd3 25. Qxd3 Nxa1 26. Qd1 { (My notes suggest 26.Ne2 as an improvement, but Stocky prefers 26.Nd1)} d4 $1 27. Ne4 (27. Bxb7 dxc3 28. Qxa1 c2 $1 {and Black wins}) 27... d3 28. Nf6+ (28. Qxa1 Bxe4 29. Bxe4 d2) 28... gxf6 29. Bxb7 d2 30. Bxc8 Nc2 31. gxf6 { and it looks like} Ne3 {wins, but White turns tables with} 32. Qg1+ $1) (23. Ne4 Qxf4 24. Nd6 Rc7 25. Nxb7 Rxb7 26. Rd4 {picks up the b4-knight, but if Black plays 23...Ba6, then it's clear that White is just setting traps}) 23... Ba6 $6 {A natural move that I passed over in my notes} ({Stocky prefers} 23... a5 $1 {The point is that after} 24. Qe1 Qxf4 25. Rd4 {, Black's knight is protected. So White can't save his f4-pawn and his position collapses.}) 24. Nxa7 $4 {I felt like I was almost in zugzwang. At the moment, ...Qxf4 is not a threat due to Rd4, but it seemed like every move worsens my position, e.g.} ( 24. Rd4 Nc2 25. Rd3 Qxf4 26. Rc1 Nb4 (26... Ne3 $1 {is even more powerful according to Stockfish} ) 27. Rxc8 Rxc8 28. Rd4 Rc1) (24. Rc1 Qxf4 {and now Rd4 hangs a rook}) (24. Nc3 Nd3) (24. Qe2 Qxf4 25. Rd4 Qxd4) (24. Rb1 Bxb5 25. axb5 d4 $1) (24. Qe1 $1 { was the only move} Qxb3 ({Stockfish prefers} 24... Qxe1+ 25. Rxe1 Bxb5 26. axb5 Rc2 {. White is clearly suffering, but there is still hope}) 25. Nd4 Qc4 26. Rad1 {Stocky's move. My notes originally suggested 26.Bf1, but then I noticed that 26...Qxf1+ was strong. White has lost a pawn, but he has a firm grip on d4 and he has mostly sealed up the holes in his position.}) 24... Ra8 $1 { I missed this idea. I thought he was going to double on the c-file} 25. Rd4 ( 25. Nb5 $4 {drops a piece}) 25... Nd3 26. Rxd3 Bxd3 27. Nc6 Re8 28. Qf3 { This is Stockfish's top choice, but now White goes down without much of a fight } (28. Qg1 {is recommended in my notes} Qxf4 29. Qxb6 {and White can dream of pushing his passed pawns. Of course it shouldn't work, but you have to create opportunities for your opponent to go wrong}) 28... Qc5 $1 (28... Qxf3 $2 29. Bxf3 {will be much harder to win. White will withdraw the knight to the wonderful d4-square.}) 29. Qxd3 Qxc6 30. Rd1 Rac8 31. Kg1 (31. f5 Qc2 { and White's "attack" is not going anywhere}) 31... Qc2 32. Qxc2 Rxc2 33. Bf1 Rec8 34. Rd4 g6 35. Rb4 R8c6 36. Bb5 (36. Rd4 Rxb2 37. Rd3 {is a slower death, but it doesn't give me any counterplay. With the text, I hope to pick up the b6-pawn and create queenside passers.}) 36... Rc1+ 37. Kf2 R6c2+ 38. Be2 { Losing a piece, but my position was already hopeless} Rxb2 39. Rxb6 { The bishop cannot be saved} (39. Ke3 Re1) (39. Kf3 Rc3+ 40. Kf2 Rcc2) 39... Rcc2 40. Rb8+ Kg7 41. a5 Rxe2+ 42. Kf3 d4 43. b4 Re3+ 44. Kg4 Rg2+ 45. Kh4 Rxh2+ 46. Kg4 Rhh3 0-1

Grand Chess Tour - Croatia

The 12-player round robin begins next week. Magnus Carlsen has put a lot of distance between himself and his rivals, so he is the big favorite in the forecast.

Poikovsky Karpov tournament

Artemiev has gained a bunch of points lately; he is the top seed in the 10-player round robin.

Altibox Norway Chess

World Champion Magnus Carlsen has been performing very well lately. He enters the tournament as the big favorite.

The Memory Competition (part 3)

One last post about memory. Click here for Part 1 and Part 2.

____________________________________________________________________________________

The 2019 National Memory Championship

            I had made it to the finals last year! This gave me a bunch of motivation. I had been training for about two and a half hours per day. But the national championship had a different format than the tournaments I was preparing for. Both of them had cards and numbers, but in the qualifier, we also had to memorize poetry and names. So I added those to my routine. But I only had a few months to train for names and poetry, so I knew I wouldn’t be great at them. However, I would run up the score in cards and numbers, where I had made enormous progress. I thought that I would qualify for the finals comfortably. If all goes well, I could even finish in the top 5. I faced a tougher field this year. All of the best memories in America were there except for former World Champion Alex Mullen. And only 8 of us would advance to the final. But I had improved so much that I was still very confident.

            The first event: Names. They passed out sheets with names and headshots. We had 15 minutes to memorize as many as possible. Then we got new sheets with just the headshots. We had to fill in the names from memory. I memorized 100 names. Decent, but not great. Four-time US champ Nelson Dellis set a national record by recalling 235 names – an astounding performance. But I didn’t need to beat Nelson at names. I just had to stay close to the top 8 and then compensate with cards and numbers. So far, it was going according to plan. I had been hoping to break my personal record (104 names), but 100 was good enough. It was way better than my performance last year, which was just 56 names.

            The second event: Numbers. We had 5 minutes to memorize as many as possible. In training, I was usually getting about 160-200. There were two attempts. My plan was to try 160 in the first round. That way I would have time to review it thoroughly and I should nail it. In the second round, we get a new sheet of numbers and I would shoot for 200. Only the best score counts. Because of that, it’s common to memorize a smaller amount in round 1. Once you’ve done that and secured a decent score, you can get ambitious in round 2.

            I turned the sheet over and started memorizing. The font was noticeably smaller than what I was used to. I tried not to let it bother me. The five minutes zipped by faster than I had expected. I turned in my sheet and then tried to write down the 160-digit number from memory.

            The small font, the pressure – it must have gotten to me. There were several places where I just drew a blank. And the scoring system severely punishes stuff like that. Make one mistake in a row of 20 digits, and you get zero points for that row. You have to get the row perfect in order to get 20 points. I messed up 2 rows. Five rows were perfect and in the last row, I only wrote down the first 14 digits because I wasn’t sure about the last 6. I scored just 114 points. That’s reasonable by the standards of the national championship, but I really needed to do well here in order to compensate for names and poetry. Defending champ John Graham memorized more than 300 digits.

            Then they passed out the second sheet of numbers. I tried 160 digits again, but I really needed to get it right this time. I was able to go faster; perhaps I had adjusted to the font size. I reviewed it twice. When I was writing the number down from memory, there was just one row that I was uncertain about. Was there a “1655” in the middle of that row or was it a “5516”? All I could remember is that I had made a mental note to be careful about that part. I put down “5516” on my sheet. I recalled the rest of the number comfortably, so I went back to double check. Then I erased the “5516” and replaced it with “1655.”

            It was 5516! No points for that row, but the other 7 rows were perfect. I scored 140. A bit disappointing. I was in 10^th place and only the top 8 would advance to the final. One of the high schoolers had put up a phenomenal score in names and then there was a new guy who somehow memorized 260 digits in his very first tournament. There were still two more events in the qualifier, but suddenly I was not sure if I could make it to the final.

            Event #3: Poetry. They passed out sheets of paper with an unpublished poem. We had 15 minutes to memorize as much as we could. In a last minute rule change, the author read the poem aloud at the beginning. I knew that there was no way I could keep up; people read faster than I can memorize. I tried to ignore it and focus on the paper. I just hoped it wouldn’t throw me off. And given my precarious situation in the tournament, I really could not afford to screw up. I budgeted 9 minutes for memorization and 6 minutes for review.

            Time’s up! I handed in the poem and started writing down as much of it as I could remember. I was pretty sure that I had all the words right. But I was a bit unsure about some of the punctuation. Is there a comma at the end of that line? If I got it wrong, I would score zero points for that line; you only get points for the lines that are perfect. I had to make some guesses. Only a day ago I thought I would sail through the qualifier comfortably; now my sloppiness with punctuation might eliminate me.

            Event #4: Cards. We had to memorize a deck in five minutes or less. They were still scoring the poetry event, so I didn’t know where I was in the standings. I did know that I had to hit it out of the park in order to qualify. Just like last year’s situation, I thought. That was not very comforting. Last year, I screwed up the first deck. The pressure. But I had improved at cards and it was one of my strongest areas. But that’s what you told yourself before the numbers event and look at how that turned out. Take it slow on the first deck, make sure you get it right this time. I started my timer. I got through the deck and reviewed it. I put it down and stopped the clock. Immediately I was disappointed by what I saw. A minute and thirty-nine seconds?! In practice, I could go at least 15-20 seconds faster without losing any accuracy. The pressure – it must have gotten to me again. I didn’t feel nervous. But memorizing in a competition and memorizing at home are two very different things. If you look up a memory tournament and scroll down to the bottom of the standings, there will probably be somebody who memorized just 6 or 7 cards in 5 minutes. You know they can do better than that. But the pressure gets to people. I reviewed the deck in my mind while waiting for everyone to finish. Everything was clear. At least I wouldn’t mess up like I did last year.

            I got it right. Though it was slow by my standards, a minute thirty-nine might be enough to propel me to the final. My dreams were still alive. As with numbers, there were two rounds and only your top score counts. Since I had already secured a good result, now I could go at top speed with the second deck and not worry so much about making a mistake. I zipped through the deck and reviewed it quickly. Usually “top speed” means under 70 seconds for me. But in a competition, there’s so much pressure… I stopped the clock. A minute and twenty-five seconds. Mildly disappointing. I recalled the deck perfectly. Now we just had to wait for the judges to tally up the scores. Maybe I would make it to the final. Maybe I wouldn’t.

            During the lunch break, I talked to a few of the other competitors. Some of last year’s finalists stumbled in the cards event. The pressure! It affects everyone. When I finished up lunch, the organizers were still working on the results. All I could do now was wait and hope.

            A spreadsheet appeared on the big screen. “The names highlighted in green are the finalists,” the organizer said. I was sitting in the back and couldn’t read it. I got up and went over for a closer look. Predictably, the “Big Four” made it. That’s defending champion John Graham, 4-time champion Nelson Dellis, world record holder Lance Tschirhart, and Livan Grijalva. I knew I couldn’t beat them. New-guy-who-memorized-260-digits-in-his-first-tournament made it. A pair of outstanding high schoolers qualified. And so did Matt Wilson! I finished 7^th.

            When they passed out the awards, I got a special mention. The announcer told everyone about the Boston Globe article I appeared in last year. I reenacted the pose in my famous photo.

            But there was not much time to celebrate. The final would begin shortly. They took the finalists to a separate room and handed us a sheet with 300 random English words. We had 15 minutes to memorize as many as we could. Then on stage, the finalist in the first seat would say the first word. If they got it right, then the next person would have to say the next word. But if the first person screws up, they’re out. In that case, the second person would have to say the first word. This continues until 3 people are eliminated.

            I was prepared for this event. You never have to memorize all 300 words; in the past, enough people would mess up early on that it never went past 100. But the field was stronger this year. In training, I had been targeting 120 words in order to be safe. But I had lost confidence after my performance in the qualifier. At the last moment, I decided to do just 100 words. Then I would have plenty of time to review and I should nail it. The 15 minutes flew by. We went back to the stage and took our seats.

            I was a bit slow at recalling the first words, but then I got into the groove. The new guy stumbled quickly. In his training, he had focused on the qualifier, so he was not ready for this event. Seven finalists were left. Only five of us would advance.

            It was John Graham’s turn. The defending champion. He hesitated. You only have 15 seconds to answer. I know he can memorize a ton of words. But on stage, in front of a large audience, facing high expectations…the pressure is so huge. He drew a blank. The champ was eliminated!

            I only had to outlast one more finalist in order to advance. Whenever it was my turn, I recalled everything quickly and accurately. But I only knew the first 100 words. Would that be enough in today’s hypercompetitive championship? Around the 95^th word, I realized that I might be in trouble. No one else was stumbling.

            Livan correctly recalled Word #101. He passed the mic to me. My turn. “That’s as far as I got – I’m out.” I got down from the stage. I left the room to get some water. A few kids in the audience congratulated me. I have fans!

            Seventh place in the qualifier, sixth in the final. That sounds great, but I’m mildly disappointed. Officially, the Random Words event finished when I was eliminated, but they kept going to see how much everyone had memorized. One of the finalists messed up in the very next round. So if I had memorized just two more words, I would have outlasted her. And the pressure got to me in the qualifier and I didn’t do as well as I had hoped. The 1655 will haunt me for a while. They say that baseball is a game of inches. I feel the same way about memory tournaments. I knew I was not going to win the championship. But finishing in the top 5 was very much within reach.

How far could I have gotten? On a good day, I might have forced a tiebreak (but it’s more likely that I would have collapsed under pressure). The next event was the Tea Party. Six “tea party guests” read a script with a bunch of personal information. Name, birthday, phone number, etc. The finalists had 15 minutes to memorize it all. They could also review a sheet with the same information printed on it. In training, I had struggled with this event. But at the last minute, the organizers changed it to just 5 guests instead of 6. However, the finalists still had the full 15 minutes to memorize. Due to that rule change, I might have survived this event if I hadn’t been eliminated earlier. Livan was eliminated quickly, but Lance, Nelson, and the two high schoolers advanced.

The last event was the Double Deck. The finalists would have 5 minutes to memorize 2 decks of cards. Even at my slow pace of 1 deck in a minute twenty-five, this isn’t hard. But in the tiebreak, I would have been busted. There would be two decks again, but this time we would only have 3 minutes. That is beyond me. Memorizing a deck in 1:25 is very different from memorizing 2 decks in 2:50. It doesn’t scale up like that. For example, memorizing a single card in 1 second is easy, but doing 52 cards in 52 seconds is not. I can’t do 2 decks in 3 minutes. But Lance memorized a single deck in under 30 seconds(!), so he could do it. My plan had been to do a deck and a half and hope my rivals stumbled.

The high schoolers were knocked out quickly. Only Lance and Nelson were left. I was almost certain that it would go to tiebreaks, but Nelson messed up on the 102^nd card. The pressure! Lance became our very deserving national champion.

The Memory Competition (part 2)

Click here for Part 1. After this, there will be one more blog post about memory, and then we will go back to writing about chess.

_________________________________________________________________________________

The 2018 National Memory Championship: Finals

I had barely qualified and knew that I didn’t have the best memory in America. I trained harder than ever, but most of my rivals were still well ahead of me. My goal was to survive Round 1 to prove that I belonged there.

But my biggest worry was about the night before the tournament. Would I sleep? Two weeks before, I had traveled to a chess tournament and didn’t sleep at all. I tried another tournament in Chicago a week later. Almost no sleep! I forced myself to play a game anyways. That was a very unwise decision. It was so hard to calculate anything. I ended up drawing against a much weaker player. My doctor prescribed some pills to use when traveling. “These will knock you out,” she promised. I wasn’t certain. Would they really work?

I stayed in a beautiful hotel close to the MIT campus. Right before bed, I took a pill. A few minutes later, a muscle would occasionally twitch. Is that supposed to happen? The next thing I remember is waking up after getting the best night’s sleep that I had ever gotten in my life. That’s exactly how you want to feel before a competition! I downed some oatmeal and walked to the auditorium on campus. Most of the other competitors arrived around the same time. We had come early in case the media wanted to talk to us. If I thought I had a serious chance of winning, I would have declined. Just get some extra rest and focus. But since I was almost certain to lose, I thought I might as well enjoy the time in the spotlight. Alas, the media had no interest in talking to Guy Who Finished Ninth. So instead I chatted with the other competitors. The first event wouldn’t start until a few hours later, so no one seemed too tense yet.

Lunch was different. A quiet and serious affair. Round 1 would start almost immediately afterwards. Focus. Get in the zone. I don’t even remember what I ate. My thoughts were elsewhere.

The finalists from the qualifier were joined by a wildcard nominee and a few stars from the high school championship. Thirteen of us in all. They lead us downstairs. We each had a sheet with 300 random English words. Memorize as many as possible in 15 minutes. Make 2 mistakes and you’re eliminated. I tried to sit away from everyone else so that there would be no distractions. Concentration is critical. Like most of the competitors, I had earplugs to block out any distracting noise.

We turned over our sheets and started memorizing. There was no way that I could get through all 300 words. But I didn’t have to. In the past, knowing the first 100 was always more than enough. But in the past, you were eliminated after just 1 mistake. Maybe we would have to know more than 100 this time. I tried 104, since I knew I couldn’t do much more than that. Words like “climb” and “laptop” were easy, since they are very visual (memory techniques are all about visuals). I struggled to come up with an image for “intuition.” I went with Jose Capablanca, a former world chess champion known for his excellent intuition. I reviewed the sheet twice. They lead us up to the stage.

The wildcard was in the first seat. He had to say the first word in the list. Then the second person had to say the next word, and so on. I was tenth. Recalling a list words in the comfort of my apartment is not so bad, but recalling them on stage could be very different. Bright lights, a large audience and the media, and with only seconds to state the right answer. But my nerves were steady. Professors are used to performing in front of large groups of strangers. I closed my eyes to better focus on the images in my mind. There was a scoreboard with the number of mistakes that each of us made. I was going to ignore it. Just make sure you get the word right when it’s your turn. That’s all that matters.

My plan to ignore everyone else was almost immediately derailed. Four-time national champion Nelson Dellis was the first to be eliminated. My eyes popped open. He had won the qualifier and he owned several national records. How on earth was he the first one to go?! Focus, Matt. I closed my eyes again and got back in the zone. I just had to outlast two more competitors.

When it was my turn, the next image and word sprang to mind quickly. Someone else stumbled and got eliminated. I didn’t know who it was. It didn’t matter to me – just get your word right when it’s your turn. Now I only had to outlast one more rival and then I could go on to Round 2.

A high schooler messed up and was knocked out. I had sailed through Round 1 with no mistakes! During the break, I took a short walk. I had achieved my goal! If I somehow survived Round 2, that was just an added bonus. We’ll worry about winning the championship next year.

I returned to the auditorium.

“Good luck,” a woman said as I went down the stairs.

“Thank you,” I replied.

“I’m also from Missouri, so I’m rooting for you,” she said.

I have fans!! How cool is that?!

Round 2 was the Tea Party. Four “tea party guests” told us personal information about themselves. Name, birthday, job, etc. Make 3 mistakes and you’re out. The event would end when the 10 remaining finalists were winnowed down to 7. I had barely trained for this event, but I wasn’t nervous – the mission had already been accomplished.

The tea party guests talked so fast! I was overwhelmed. I hardly caught any of it. We had five minutes to study the same information written on paper, but it was still overwhelming. I knew I was busted. But I don’t lose motivation when the situation is hopeless. As chess players know, keep fighting hard even when the position is lost; maybe the opponent will make a mistake. I blocked out the rest of the world and focused all my attention on the sheet. Soak up as much information as possible. The Boston Globe noticed the intense concentration on my face and they decided to put my picture in the paper! The headline was "Competition Puts Their Memory to the Test". I’m famous!!

It turned out that many of the competitors were just as overwhelmed as I was. I botched the first question, but my rivals were also making mistakes left and right. One of them didn’t get anything right. He was the first one to go. Then a former national champion stumbled. It suddenly occurred to me that I might actually survive! I only had to outlast one more rival. I was still struggling and making mistakes, but I got one correct answer. Then I had to recall the name of one of the tea party guests. In my mind, it was blended with the names of the other people and a muddle of other semi-memorized information. I got it wrong. I was eliminated. The round ended and the 7 remaining finalists proceeded to Round 3.

I could relax. I had accomplished my goal and had been so close to making it through the next round. I took a seat in the audience and watched the rest of the finalists sweat through Round 3.

The organizers had dreamed up a new form of mental torture for this round. “Long Term Recall.” A month earlier, they had sent out spreadsheets with a vast amount of information. The Academy Award winners, the Football Hall of Fame, the Rock and Roll Hall of Fame, and the periodic table of elements. That was too much for me. I only got through about half of it. It was good training, but I really hope that I never have to do that again. So even if I had survived Round 2, I would not have made it much further. The wildcard was swiftly eliminated. However, a pair of high schoolers performed very well; it took a long time to knock them out. John Graham, the second seed from the qualifier, advanced to the last round. He was joined by prodigy Claire Wang and newcomer Avi Chavda, who had been seated next to me in the qualifier.

In Round 4, they had 5 minutes to memorize two decks of cards. Claire stumbled quickly and was knocked out. Avi seemed to be recalling the cards easily while John took his time before answering. I wondered if it would go to tiebreaks. At this level, memorizing 2 decks in 5 minutes is not very hard. But Avi messed up in the second deck, so John Graham became the national memory champion.

Nice work, John. I’ll try to beat you next time! We went out to dinner with the other competitors and the organizers (they’re wonderful people). Our generous hosts from MIT paid the bill. Since the tournament was during summer break, I got to stay in Boston for a few extra days and enjoy the history and art in the city. There is so much to do in Boston! This is the second time I have visited and I’ve only seen a small fraction of the city. No one recognized me from the Boston Globe photo; I guess I’m not really a celebrity yet 😃

Farewell, Boston. I hope I get an excuse to visit you again!

The Memory Competition (part 1)

GM Timur Gareyev used memory techniques in order to set his blindfold chess world record. I also have an interest in memory and I compete in tournaments. This my story from last year's championship. Next week, I'll post an update with this year's event. To regular readers of this blog: don't worry, I'll go back to writing about chess tournaments soon.

                                                                                                                                                                   

The 2018 National Memory Championship

By Matthew S. Wilson

The journey began in early 2011. I was a first year grad student at the time. I was browsing the news online and there was a New York Times article about memory. Some guy named Joshua Foer had won a national memory tournament and wanted to tell us about it. I started reading the article.

            As I was reading, two things came to mind. First of all, a better memory might help me pass the very difficult qualifying exams (about half of last year’s class had failed, forcing them to drop out of the Economics PhD program). Second, I wonder how my memory would stack up against a former champion? He talked about memorizing the sequence of an entire deck of 52 playing cards. So I found a deck of my cards in my apartment and shuffled. The first few cards were pretty easy. Anyone can memorize 5 cards without much effort. But by the time I got to roughly the 20^th card, it was tremendously straining. There was no way I could get through the entire deck. I stopped and tried something similar to his technique. He had a system that assigned each of the cards to a celebrity or someone he knew; then he mentally placed all the images into a building he was familiar with. I had no interest in developing such a system; I was just wondering if his principles would help me get through the qualifying exams. So instead I placed images of the cards into my parent’s house. Something like the King of Spades knocking on the door, which was answered by the Five of Clubs while the Eight of Hearts swung from the chandelier, etc. It took a long time to get through the whole deck. But I was stunned at how easy it was to remember it all. I was hooked.

            I practiced nearly every day. At first, it took more than 5 minutes to get through half the deck. When I tried to cut that down to 4 minutes, I felt rushed. But I progressed steadily. In less than six months, I was already memorizing the whole deck in under 5 minutes. My efforts started bearing fruit in Spring Quarter 2011. My GPA shot up to 3.9!

            After I passed the qualifying exams – a long story for another time – I decided to keep practicing. My results were good enough that maybe I had a chance at the national memory championship if I kept improving. But I was stuck in a routine where my training was only with decks of cards. For competitions, I would also have to memorize numbers and names and other things. I started with numbers a year or two later, but even then, my main focus was cards.

            Fast forward to late 2017. My memory had been on a plateau since 2014. I had been thinking about going to the national championship for years, but I never got around to it. It was always on the East Coast; the distance deterred me. I had signed up for a memory competition in Canada but cancelled because there was a chess tournament on the same weekend. So I registered for the national championship and figured that I could withdraw if something else came up. I didn’t buy my plane ticket yet.

            In mid-January, I realized that the tournament was just two months away. It was time to get serious with my training. I had only been doing cards and numbers; I needed to memorize poetry and names as well. My friend Joey Frantz recommended a poetry book. For names and faces, I signed up on Memocamp, a memory training website.

            I had a good chess tournament at the end of January, but after the euphoria faded, I was back to being depressed over my failing job search (another long story for a different time). But after a few weeks, I started seeing almost daily improvement in my memory. It was exciting. I forgot(!) that I was supposed to be unhappy about my failed career. As the memory tournament approached, I compared my scores to the results from previous championships. The top competitors would advance to the finals. And one of them will be crowned as national champion. Though I was improving quickly, it didn’t look like it would be enough to qualify for the final. If I matched my best recent results in training for all four events and if the finalists from 2016 failed to improve and if there weren’t any newcomers who could beat me, then I might just barely qualify. There was certainly a chance that I would collapse from the pressure of competing in my first memory tournament. Also, I often don’t sleep very well when I’m not at home. To make matters worse, the first event (poetry) began at 8:45am Eastern Daylight Saving Time. That’s 6:45am Central Standard Time (the event was just a week after the clocks were changed). And I am not a morning person. It was a long shot.

            I had a plan to make the most of my chances. First, gradually go to bed earlier. For a couple of weeks, I moved my schedule back by 5-10 minutes each day. Second, I arrived at the hotel two days early, just in case I didn’t sleep well on the first night. Then there was the plan for the 4 events in the tournament. For random numbers, my practice results were probably on pace to qualify. For poetry? Probably not. For names and faces? Definitely not. Stay within striking distance for numbers, poetry, and names. Then run up the score in my strongest area: cards.

            The big day arrived. I didn’t sleep as well as I would have liked. But I felt okay. To save time, I ate a breakfast bar during the announcements. Good news: Alex Mullen wasn’t there. He was the world champion and he can memorize a deck of cards in under 20 seconds. If he’s absent, the rest of us have a better chance. Unfortunately, Nelson Dellis and a few other stars were present. But there was more good news. According to the website, only the top 7 would qualify for the final, but in the announcements, they changed that to the top 8. And I wasn’t feeling nervous. I still had a chance.

            We had 15 minutes to memorize a poem. Then we had 20 minutes to write down as much of it as we could remember. But when recalling the first stanza, I realized that I was not 100% sure how to spell “easel.” I had never been a good speller. Due to the way the scoring works, a little error like that can be costly. I guessed. I got it right. The rest was fine except for one thing. I wasn’t entirely sure if a certain line ended with a period or a comma. I had to make a guess. The remainder of the 20 minutes was spent cleaning up my handwriting; I didn’t want to risk losing any points if they misread something. I felt that I could have done a bit better. But my performance was still reasonable. I just had to stay close to the top 8 for now. Then in the final event (cards), I would shine and hopefully catch up.

            Event #2: Speed Numbers. We had 5 minutes to memorize as many random numbers as possible. Then we had 10 minutes to write it all down. I collapsed. The pressure must have gotten to me. I only did half as well as I did in practice. To make matters worse, the guy sitting next to me memorized more than 100 digits. And he was a newcomer. I had expected that the stars from the previous championships would fill most of the top 8 spots, but if this new guy also gets a spot, I’m probably eliminated! Fortunately, there was a second chance at this event; we got another sheet of random numbers and only your top score counted. I steadied my nerves and memorized 103 digits in 5 minutes. When I finished writing them down, I was nearly certain that I had nailed it. It was perfect. But my neighbor edged me out, memorizing 104. My chances were slipping away.

            Then an MIT professor gave a presentation while the organizers tallied up the scores. I was around 10^th place so far. That’s okay. Just survive until we get to the cards.

            Then we had 15 minutes for the names and faces event. We were given sheets of color photos; their names were printed below their picture. There was a rustling of papers as the competitors turned over the sheets. Just try to ignore it. I tied my personal record. Fifty-six points. (1 point for each first name, 1 point for each last name). But it wasn’t good enough. Some of my rivals were scoring more than 100 points.

            Speed Cards: the final event of the day. Memorize the order of 52 playing cards in 5 minutes or less. Not many people can do that. But in practice, I was doing it in about a minute and a half. This was my opportunity to stage a comeback and qualify for the final. I got through the deck in 2 minutes and 9 seconds. Slow! But there would be a second chance and only your top score counts. And it was possible that 2:09 would be good enough. Or maybe not – my neighbor’s timer beeped when he stopped the clock. He, too, could memorize a deck in under 3 minutes. If he finishes ahead of me, it’s going to be harder to qualify. I reviewed the deck in my mind while waiting for the rest of the 5 minutes to elapse. Everything was clear. Almost. After the Six of Spades, there was the Three of Hearts. But was this pair placed before or was it placed after a certain sequence? I wasn’t entirely sure. I guessed.

            I was wrong! This mistake happened on the sixth card in the deck, so I was credited with memorizing just five cards. Totally embarrassing! And this was supposed to be the event where I would shine. Yes, there would be a second chance, but I was almost surely busted. The gentleman sitting next to me memorized his deck perfectly. He also bested me in numbers and poetry, I’m bad with names – he’s got to be ahead of me. And if he qualifies, and if the former champions qualify, and if the other stars qualify, there isn’t much space left for me. Maybe I can overtake him if I go at top speed. At my best, I can get through a deck in under a minute. But then there is a very high chance that I’ll make another mistake. It was too risky and besides, I knew at the beginning of the tournament that I probably wouldn’t qualify. I had really really wanted it, but I had been prepared for failure. Finishing in the top 10 would still be respectable. At the very least, it was possible to avoid humiliation and correctly memorize more than 5 cards!

The decks were shuffled and we tried again. I wished that I could go to the bathroom first, but now I just had to ignore it. I was extra slow and careful. After going through the deck I reviewed it twice. I put the deck down and stopped the timer. Two minutes and forty-nine seconds. Very slow by my standards, but it could still provide a big boost to my score. But only if I didn’t make any mistakes. I reviewed the deck in my head while waiting for the rest of the 5 minutes to pass. There was a very tiny bit of uncertainty about one part. Everything else was clear.

At the end, I raised my hand and a judge came over to check. I had recalled the deck perfectly! She recorded my result. I raised my arms in victory. Now we just had to wait for them to calculate all the scores.

The two former champs would certainly qualify. Avi (the guy sitting next to my right) had probably defeated me in every event. Then there were a couple of experienced competitors at the tables in front of us who sounded confident; from what I overheard, their scores seemed higher than mine in several of the events. But if we add all that together, only 6 or 7 of the qualifying spots would be taken. Is it possible that I finished in the top 8?

We all had lunch together and chatted. Then we returned to hear the results. Now I was nervous. I thought that they would start with 8^th place and then build up to 1^st place; that seemed like the most dramatic way to do it. But they started with 1^st place. Former champion Nelson Dellis had the highest score. No surprise there. A few other stars qualified, Avi qualified. But two of the other finalists had not been on my radar screen. Add them to the other people who would probably qualify, and all 8 places would be filled. I thought it was over.

“In eighth place – also a newcomer –” the announcer began.

“ ‘Also a newcomer’? I’m a newcomer! Is it me?” I thought.

“Matthew Wilson!” the announcer said.

I had done it! I walked up to the stage. After a few more remarks, the announcer passed the mic to the finalists. Starting with me. I hadn’t prepared for this. I was just surprised and happy. “So glad to be here – didn’t think I would make it to the final – this is awesome!” –my victory speech.

But there was one more plot twist. I stuck around to watch the conclusion of the high school team championship. Then there was an announcement. They had miscalculated Kyle Matschke’s score; actually, he was the one who finished 8^th and I was 9^th. Would they take my spot away from me?

“ –so we will proceed with 9 finalists instead,” the announcer concluded.

I was still a finalist! You could say that I qualified due to a scoring error. But if I had done as well in the cards event as I had in practice, I may have edged him out. In July, we will reconvene in Cambridge, Massachusetts and compete for the national championship.

Capablanca Memorial

Ivanchuk is the top seed in the 6-player round robin. Will he return to the 2700 club?

Shenzhen Masters and Grenke Chess Classic

I am a bit late in releasing my forecast for the Shenzhen Masters. It is a 6-player double round robin; the top seed is Ding Liren.

The Grenke Chess Classic starts in a few days. It will include World Champion Magnus Carlsen and his 2018 challenger, Fabiano Caruana.

Gashimov Memorial

Gashimov was a 2700 player who died at a young age. Now there is an annual tournament named after him. Top seed: Magnus Carlsen.

2019 US Championship

There is a national championship that features 5 players in the 2700 club. No, I'm not talking about Russia. The Big 3 (Caruana, So, and Nakamura) are joined by defending champ Sam Shankland, who has maintained his 2700 rating, and Dominguez. But according to the model, there is a clear favorite among these top 5. Caruana is coming off a very successful 2018, while Nakamura and So are well below their peaks.

E4stat has played Akobian before. I have not faced "Blindfold King" Gareev, but we share an interest in applying memory techniques, so I hope he does well. The forecast:

Prague Masters forecast

I'm a bit late with the forecast for the Prague Masters. It is a 10-player round robin starring Navara as the top seed. We are not used to seeing Gelfand as the lowest rated player. Not that long ago, he played Anand in the World Championship.

Spring Classic forecast

The Spring Classic will take place at the St. Louis Chess Club in a few days. Ivanchuk is the top seed in the A group, but he will have a lot of strong opposition.

GM swindles e4stat

Another loss, but I feel like I am making progress. At first, my duels with grandmasters were quite lopsided (GM clobbers e4stat and E4stat loses to another GM). Then I played GM Var Akobian in a rapid tournament and actually had a decent position until I noticed that my clock ticked down to 10 seconds. This time, the GM was lucky to escape!

I was competing in the Port of Burlington Open. On the first day, I beat a pair of 1800's. I took a half point bye in Round 3. Then I faced GM Fidel Corrales Jimenez in Round 4. I had played him before in the 2018 Jane Addams Memorial.

Click this link if the game does not display properly

[Event "Port of Burlington Open"] [Site "Burlington, Iowa"] [Date "2019.02.03"] [Round "4"] [White "Wilson, Matthew"] [Black "Jimenez, GM Fidel Corrales"] [Result "0-1"] [ECO "B28"] [WhiteElo "2108"] [BlackElo "2633"] [Annotator "Wilson,Matthew"] [PlyCount "80"] {We had faced each other before in January 2018. He won convincingly. I wondered if he remembered me. Or maybe he didn't because he clobbered so many amateurs every year.} 1. e4 c5 2. Nf3 a6 { Hey, that's the move I played against you in 2018! He definitely remembered me! } 3. c4 d6 {Back in our game from 2018, I had played 3...e6.} 4. d4 ({ White shouldn't delay d2-d4 any longer:} 4. Nc3 e5) 4... cxd4 5. Nxd4 Nf6 6. Nc3 e6 7. Be2 Qc7 8. O-O b6 9. Be3 ({Back when I was experimenting with different move orders, I remember getting a similar position with Black. Then the computer crushed me with} 9. f4 Bb7 10. f5 $1 { I looked at the same idea here. However, I didn't see enough compensation after } e5 11. Nc2 Nxe4 12. Nd5 Bxd5 13. Qxd5 Qc6 (13... Qc5+ 14. Ne3 Qxd5 15. Nxd5 { and White recovers his material favorably}) { If I had kept calculating, I might have found} 14. Qxc6+ Nxc6 15. Bf3 $1 { and White wins. But Black is not forced to take on e4. If he plays 11...Nbd7, he is probably okay. So the move I played in the game is fine.}) 9... Bb7 10. f3 Nbd7 11. Rc1 Rc8 12. Qd2 Be7 13. Rfd1 Qb8 {Black's ...b5 and ...d5 pawn breaks almost never work when his queen shares the c-file with White's rook, so the queen steps aside.} 14. Bf1 { A normal move in the Hedgehog. White prepares to slide the queen over to f2.} h5 (14... O-O {is more typical, but this is also playable.}) 15. Kh1 { This is a standard move in the Hedgehog, but it is not the most accurate. For reasons given in the next note, I wanted to meet ...Be7-d8-c7 with Bg1. So my king scoots over to make space for the bishop. But this does allow Black to set up tactics on the h-file and the long diagonal.} (15. b3 { was slightly better}) 15... h4 16. b3 {A decent move, but the reasoning behind it was flawed. At first, I was not afriad of ...Nf6-h5-g3+. But after the standard ...Be7-d8-c7 maneuver, it could be deadly. Once the bishop reaches c7, Black threatens ...d5 followed by ...Bxh2. To prevent this, I wanted to play Bg1, but then ...Ng3+ is almost mate! I calmed down once I saw that Nde2 neutralizes ...Ng3+. Then I thought my c4-pawn was hanging after ...Be7-d8-c7 Bg1 ...Nh5 Nde2. So after a long think, I decided to play 16.b3 to protect it. But it's all based on a hallucination: after ...Bc7, Black can't play ...Rxc4 because it's illegal! Nevertheless, in some lines it is handy to give the c4-pawn extra support, so my move is fine.} h3 $2 {Sending the pawn on a suicide mission. Previously, the pawn prevented me from playing g2-g4; now my g-pawn is free to advance. Black had a couple of alternatives. One of them was the 16...Bd8 followed by ...Bc7. Another one is to reroute the d7-knight to g6, where it would support the h4-pawn so Black could castle.} 17. g4 ({ I seriously considered} 17. gxh3 {. Black can't punish it:} d5 18. exd5 exd5 19. Nxd5 Nxd5 20. cxd5 Bxd5 21. Bxa6 {. The advantage of playing 17.gxh3 is that the bishop can come to g2 in some variations. However, the move that I played in the game is better. White can win the h3-pawn without compromising his structure.}) 17... Ne5 (17... d5 $2 18. exd5 {Stockfish's choice} (18. g5 $2 {is recommended in my notes, but Stockfish replies with} dxe4 $1 19. gxf6 Nxf6 20. fxe4 Nxe4 21. Nxe4 Bxe4+ 22. Kg1 Rh4 23. Be2 { . Black has two pawns and a powerful attack for the piece}) 18... exd5 19. g5 $1 {After Black moves the knight, the d5-pawn falls}) 18. g5 Nfd7 19. Qf2 { Simplest and best. I just want to round up the h-pawn.} (19. f4 { leads to chaos. Here's a sample line from my notes:} Ng6 (19... Bxe4+ { is actually playable, according to the machine} 20. Nxe4 Qa8 21. Qc2 Nc5) 20. f5 Ngf8 21. g6 fxg6 22. fxg6 Nc5 23. Qf2 Nxg6 24. Qg3 Ne5 25. Qxg7 Nxe4 $3 26. Nd5 (26. Qxh8+ Kd7 {White has gobbled up a whole rook with check(!), but now his queen is under attack and his king is feeling very unsafe}) { I thought White was winning here, but Stockfish finds} 26... Bf6 $3 27. Nxf6+ Nxf6+ 28. Kg1 Rg8 {and instead White has to resign!}) 19... Nc5 20. Qg3 g6 { I thought that this was a good move and he played it quickly. I had been dreaming of pushing back the Black knights and then unleashing g5-g6! However, the computer disapproves of his idea. It thinks that the g6-square should be reserved for the knight.} ({Stockfish prefers} 20... Kf8 21. Kg1 Ng6 22. b4 ( 22. Bxh3 $2 d5 $1) 22... Nd7 23. a3 { . Now Black should not try to break out immediately:} Nge5 $6 (23... Kg8 { is better, but White retains a big advantage}) 24. Na4 d5 $2 25. exd5 exd5 26. cxd5 Bxd5 27. Bxa6) 21. Rc2 {Now Black doesn't pick up a tempo with ...Nd3 after I snap off the h3-pawn. The pawn isn't running away and Black can't do much, so I can get away with this slow move.} Qa8 22. Kg1 ({There is no hurry to take the pawn, so first I avoid any shenanigans on the long diagonal. Stockfish says that White can get away with grabbing the pawn immediately, but it gets messy:} 22. Bxh3 b5 23. cxb5 Nxf3 24. Qxf3 Nxe4 25. Nc6) 22... Rh5 23. Bxh3 b5 {A desperate bid for counterplay} ({In my notes, I thought that} 23... Kf8 {offered more resistance. Black intends to continue with ...Kg7 and ... Rch8. But Stocky shows that ...Kg7 doesn't work in many variations - White can reply with f3-f4-f5-f6+}) (23... Kd7 { looks like a faster way to get the c8-rook into the fight, but it runs into} 24. Nd5 $1) 24. cxb5 Ncd3 {Once he played this, I knew I had a winning position } ({During the game, I was unsure if} 24... Nxf3+ {was sound. I thought that the point of 23...b5 was to open the c-file so that this tactic worked.} 25. Qxf3 ({In my notes and during the game, I rejected} 25. Nxf3 Nxe4 26. Nxe4 Rxc2 {. But White is still winning here. Stockfish finds} 27. Nf6+ $1 Bxf6 28. gxf6 Bxf3 $2 29. Qxd6 {with mate in 6}) 25... Nxe4 26. bxa6 $1 {I missed this move during the game. It disrupts Black's battery on the long diagonal and leaves me with a completely winning position.} ({Much stronger than} 26. Nxe4 Rxc2 27. Nxc2 Bxe4 {, but according to the engine, White wins here as well})) 25. Bf1 Nb4 26. Rcc1 axb5 ({A weird tactic:} 26... Kf8 27. a3) 27. Bxb5+ Kf8 28. a4 ( 28. f4 Nec6 {is also winning}) 28... Ba6 (28... Kg7 29. f4 {My notes stopped here since White is completely winning, but Stockfish extends the variation by one more move:} Nec6 30. f5 $1 { Black is down material *and* facing a powerful attack}) 29. Bxa6 Qxa6 30. Ncb5 Ned3 31. Rxc8+ Qxc8 32. f4 $4 ({Why didn't I just take the pawn???!!!?!??!} 32. Nxd6 {and Black is just busted:} Bxd6 33. Qxd6+ Kg7 34. Nc6 $1 {Somehow I didn' t even consider 32.Nxd6. Time pressure is no excuse; the pawn grab is just too obvious. You could have beaten a grandmaster!!}) 32... e5 $1 33. fxe5 $2 (33. Ne2 Rh3) (33. Nf3 exf4 34. Bxf4 Rh3) ({Stockfish recommends} 33. f5 {, a move t hat I missed in the game and in my notes. White is 3(!) pawns ahead, so he can afford to sacrifice a piece. Nevertheless, Black is still in control after} exd4 34. Nxd4 Ne5) 33... dxe5 $2 (33... Rh3 $1 34. Qg2 Rxe3 35. exd6 Bd8 { looks chaotic at first, but if you look a bit longer, you will see that White can't do anything. Black is completely winning, according to Stockfish. 33... dxe5 looks simpler, but as we will see shortly, it could have lead to extreme complications.}) 34. Nf3 $2 (34. Na7 Qb7 (34... Qc3 $2 { lets White off the hook. Stockfish finds an incredible variation:} 35. h4 exd4 36. Qb8+ Kg7 37. Nb5 $1 Qc6 38. Bxd4+ f6 39. Rf1 Kh7 40. gxf6 Qxe4 41. Qh8+ $3 Kxh8 42. fxe7+ Kh7 43. Rf7+ Kh6 44. Bg7+ Kh7 45. Bf6+ Kh6 46. Bg7+ {with a draw }) 35. Ndc6 {leads to another long forced variation from the machine:} Nf4 36. Bxf4 exf4 37. Qc3 {(or else 37...Bc5+)} Qb6+ 38. Kf1 (38. Kh1 Qf2) 38... Rxh2 39. Rd4 Nc2 40. Kg1 Rh5 41. Qxc2 Bc5 42. Qc3 Bxd4+ 43. Qxd4 Qxb3 44. Qb4+ Qxb4 45. Nxb4 Rxg5+ 46. Kf2 Ra5 {and Black is clearly the one who is pressing, but maybe White can hold. I don't think I could find all that even with plenty of time on my clock}) 34... Rh3 $1 {I thought he was going to play 34...Qc2 and pick up my queenside pawns, but this is much more powerful} 35. Qg2 Bc5 $1 { Eliminating a defender and activating the queen with tempo} 36. Bxc5+ Qxc5+ 37. Kf1 Qe3 $1 38. Ng1 {The rook is untouchable, of course} (38. Ne1 Nxe1 39. Rxe1 Rf3+) (38. Nd2 Rh4 $1 39. Qg3 Rf4+ 40. Kg2 Qe2+) 38... Qf4+ 39. Ke2 Re3+ 40. Kd2 Rg3+ 0-1