A Wildly Average USATE

Against my expectations, I was able to make the US Amateur Team East for the third (!) year in a row, and like David, saw a rare opportunity to push for some of the winner glory, with CMU fielding Grant Xu (2396), me (2136), Ryan Christianson (2059), and Alex Hallenbeck (2027) for an average January rating of 2154.5. Also, I had worked my way to 2180 through early February, so I was definitely pushing for master, and the tournament served as an important test of how I could handle that mentally.


There was no contest for the highlight of the event as we pulled off a huge upset in Round 5 and I drew an International Master for the first time.

Ryan and Beilin in Round 5 (photo by Vanessa Sun)

Our opponents were GM Eric Hansen and IM Aman Hambleton of Chessbrah fame, followed by an expert and a 1600 on the bottom boards. Their top-heavy strategy put a lot of pressure on the GM and IM to clean up, although it seemed to mostly work as I was the only one to score against either of them. On the other hand, it seemed like we would have to win the bottom two boards to stay alive in the match.

Alex delivered an early win on Board 4, while Grant looked to have a good Alapin against Hansen, but fell for an early tactic that left him in a positional bind for the rest of the game. Against Hambleton, I ended up on the Black side of the closed Rubinstein Nimzo-Indian for the first time. I was really nervous because I was playing an IM and had no idea how theoretical that line was, but I played it solidly, stayed even throughout the game, and we agreed to a draw on move 29 to close out my third encounter against an IM/GM. See the full game here!

At 1.5/3, this left the match in the hands of Ryan, who had struggled against some lower-rated players, but nevertheless defeated his expert opponent easily. For the first time, CMU was 4.5/5 going into the last round, and tied for 2nd with a legitimate shot at winning. With my unexpected draw, I also had a great chance to make master if I could win my last game.

Unfortunately, we didn’t quite get the ending we hoped for as Grant, Ryan, and I were simply crushed by IM Justin Sarkar and young experts Eddy Tian and Nico Chasin in the last round. To his credit, Alex easily beat his young expert opponent on Board 4, but it’s hard to win the match when your teammates are all lost by move 20.

Nevertheless, our score of 4.5/6 was good enough for Top College and Top Pennsylvania team (the second was later removed due to duplicate prize policies). And after three long days, I had gained one (!) rating point. I had mixed feelings about this; the result was disappointing given what I did in Round 5 and the chance I had after that, but on the other hand, 2180 is around the point where many National Master contenders collapse, and I avoided that for the most part.

In hindsight though, it was kind of a fitting result for some interesting moments earlier in the tournament.

Round 1: Lucky Misses

We played down in Round 1, and the outcome of the match was never really in doubt. But at the time, I felt like my game was a lot more chaotic than it needed to be. My 1850-rated opponent quickly ended up on the wrong side of the 4. Nc3 Advance Caro-Kann:

Weinstein – Li: after 11…Nc6

I made some mistakes and tried to simplify too early. Eventually, I bailed:

Weinstein – Liafter 18. Be3

Thinking I’d at least grab a pawn for my troubles (great logic, right?), I ventured 18…Bc5 19. Qa4+ Kf8 20. O-O-O Bxe3+ 21. fxe3 Qxe3+ 22. Kb1 Nf6, but White won his pawn back with 23. Qb4+ Qc5 24. Qxb7:

Weinstein – Li: after 24. Qxb7

Black’s king is actually safe now, but it’s not easy to find a way to break through the kingside. White gave me an opportunity a few moves later, but that’s where the misses started. While I was better (or winning) throughout the sequence, it was a little disconcerting to have missed some of White’s moves, lest one of them be a spoiler!

Weinstein – Li: after 27. Ne4

I thought I calculated through everything, and went ahead with 27…Nxe4! After the forced 28. Qxf3 Qxc2+ 29. Ka1 Rb8 30. Rb1 Nd2 I missed 31. Qf4, gaining time by attacking the rook on b8. Same thing after 31…Rb7 32. Qh2. Again, neither move actually saved White, but I was a little nervous at having missed both of them.

I faced the final test after 32…Nb3+ 33. Ka2 Nd2 (a repetition to get closer to move 40) 34. Ka1 Nb3+ 35. Ka2 Qc4! 36. Rbd1 Nd2+ 37. Ka1.


It took me 18 of my remaining 19 minutes to find 37…Rxb2! and it’s mate in 3 (not including the Qb8+ intermezzo) because Black threatens to mate on a2 and 38. Kxb2 Qb3+ mates next move. Evidently I don’t do any tactics puzzles.

After the hiccups (if only mentally), that was over and our team followed suit, winning 4-0.

Round 2: Another Miss or a Blessing in Disguise?

Seeded 40th out of about 300 teams, we weren’t expecting to play up so early. Alas, pairings (accelerated, maybe?) are pairings, and we were paired against the 3rd seed team of IM Dean Ippolito, NM Eric Most, and a 2100 and 2000. Grant drew Ippolito comfortably, Ryan’s opponent offered a draw in a slightly better isolated queen pawn position, and Alex and I looked to be winning our games. Unfortunately, neither of us could convert our wins and we drew the match 2-2.

One could reasonably make the case that this saved us from some future obstacles, as it set us up for our remarkable Round 5 win and it was the only half point we gave up before the last round. But it was really hard for me to not be disappointed at my own game.

Li – Most: after 20…Qg5

The opening wasn’t great, but my opponent thought he would just trade queens en route to destroying my queenside. I was very happy to find 21. Nf4!, threatening Nxe6 and exploiting the loose knight on c4. The game continued 21…Nxf4 22. gxf4 Qf6 23. cxb4 Qxd4? (23…Ba6 put up a lot more resistance) 24. Rd1 Qxb2 25. Qxc4, leaving me up a piece for two pawns.

From there, it was a series of rash yet timid simplifications on my part. I eventually bailed into an ending, which admittedly wasn’t the easiest to win, though with an hour more on the clock, I should have done way better.

Li – Most: after 39. Nxb2

And after Black tried to break up the e- and f-pawns, I got a passed d-pawn:

Li – Most: 45…Bd7

The simplest route seemed to be 46. Nc6! forcing 46…Bxc6 (otherwise, 46…a6 47. Bh3+ g4 48. Bxg4+ Kxg4 49. Ne5+) 47. dxc6 Ke6. And I naively assumed that: 1) I had to deal with the a-pawn before anything else, and 2) the bishop would hold the kingside pawns easily. One issue at play is that White has the wrong-colored bishop should I end up with only the h-pawn.

I was very, very wrong on both points. If the king is away, the kingside pawns are not nobodies, as I found out the hard way. And while the a-pawn is a bit of a nuisance, White can simply tie Black’s king to the passed c-pawn while marching the king over to the kingside. White may end up with only the h-pawn left, but if calculated correctly, White should have enough time to take Black’s pawns and prevent Black from reaching the h8 corner. For example, 48. Bd5! Kd6 49. Ke4 a5 50. Kf5 a4 51. Kxg5 a3 and White takes the f7 and h7 pawns while Black is distracted with c6.

But after 48. Kd4 Kd6 49. Kc4? f5 50. Kb5?? g4 it’s a dead draw:

Li – Most: after 50…g4

I played this out for completeness, but Black’s pawns are too fast by one tempo: 51. h3 h5 52. hxg4 hxg4 53. Ka6 Kc7 54. Kxa7 f4 55. Be4 (otherwise 55…f3 wins!) 55…f3 56. Ka6 f2 57. Bg2 f1=Q and we wrapped up the game at 12:30 am.

Round 3: Durkin’s Folly

We went back to playing down for two rounds, this time against a team of kids each rated around 1500. I’m embarrassed to say that this was more interesting that it should have been, but…

On Board 3, Ryan tried the Durkin Attack (1. Na3) because the tournament was giving out prizes to the best games in a few openings, which included the Vienna, the King’s and Queen’s Gambits, and the Durkin Attack. Ryan’s game started with 1. Na3 e6 2. Nc4 d5 3. Ne5 f6 (later, we decided that 2. c4 followed by Nc2 was a better plan). Meanwhile, I had ruined a good position by miscalculating a tactic and had to trade into a microscopically worse rook ending with seemingly no winning chances. It seemed that we were going to have to resort to the bare-minimum 2.5-1.5 win against a team we outrated by 600 points.

Mak – Li: after 16…Na5 =+

But things turned really hairy later, when White tried a last-ditch h-pawn storm on the kingside:

Mak – Li: after 28. h5

I immediately played 28…Qa2?, which I thought was forcing 29. Rb4 (which he did play). Apparently, 30. Qd1! is good for White, because 30…Qa5 31. b6! threatens Ra1 trapping the queen, and 30…Qa4 31. Ra1 wins the a7-pawn with a winning advantage.

Be that as it may, the game continued 29. Rb4 Be7 30. h6 Rg8?! 31. hxg7+ Rxg7 32. Be5 f6 33. Qxe6, and I made my pre-planned escape with 33…Rxc3 34. Rxc3 Qa1+.

Mak – Li: after 34…Qa1+

Although this is not an easy find, 35. Rb1!! likely wins after 35…Qxb1+ 36. Kh2 (White threatening Rc8+ and Bxf6) and 35…Qxc3 36. Bxf6. Even disregarding that, 35. Kh2 Qxc3 36. Bd6 (I missed this!) forced 36…Bxd6 37. Qxd6 Qc7 and we reached this ending:

Mak – Liafter 38…Rxc7

Somehow, I managed to win this ending as Black. That wasn’t very nice of me, but a win is a win. That happened in several monumental steps. First, White immediately played 39. b6, which led to trading his d-pawn for my f-pawn.

Mak – Li: after 41…Rxd4

Now Black has a passed pawn. White should still draw this without trouble. In fact, I think I could lose this as Black if I was careless enough. Though, White played f4, let me activate my king, and traded his f-pawn for my h-pawn.

Mak – Li: after 47…Kxf4

Now Black’s king is somehow active. I didn’t see that coming from the beginning position, to be honest. Nevertheless, White’s king is very near and with the rook behind the Black pawn and king, it’s still a dead draw. In fact, this is a draw even without White’s g-pawn.

Mak – Li: after 52…Rh2

This might look a little scary, but in fact White is still fine. Even without the g-pawn, this is the famous Philidor position. The next position isn’t so easy for White though:

Mak – Li: after 56…Rh8

And White finally cracked with 57. Rg1?? (57. Rf2 was necessary, to jump to the back ranks for some annoying checks). And after 57…Kd4 58. Rf1 Ra2+ 59. Kd1 Kc3 60. Rg1 Ra2 61. Ke1 Ra1+ he resigned.

This clinched the match with 3 points, but Ryan was still losing against his 1500-rated opponent with the Durkin. However, his opponent offered him a draw in a likely winning ending, so it was a 3.5-0.5 match. One wouldn’t exactly have guessed that from the positions earlier in our games.

Round 4: Finally Clean Sweep

In our last match leading up to our Round 5 upset, we played down once more, but got the job done 4-0. For once, my game didn’t make me completely nervous. I’ll leave it at this:

Li – Gindi: after 14…g6

I switched gears slightly with 15. Qh2!? and it’s actually pretty awkward for Black to defend the c6 and d6 pawns (best is probably the counterattacking 15…b4). Instead, Black tried 15…Qb6 16. Nde2 c5? 17. g5! Nh5 18. Nd5 which wins at least a pawn.

And that’s how you work your way to drawing an International Master!

The Chess^Summit Picture That Wasn’t

With DavidGrantVanessa, and Vishal (and some former guest authors to boot) all at the tournament, it was kind of expected that we’d get some unifying picture of us all. Unfortunately, with one round left, Vanessa insisted on trying to find Vishal first, which didn’t materialize. That left us with a few individual pictures of us, all taken by Vanessa:

Vanessa with David Brodsky
Vanessa with Beilin
Grant playing IM Sarkar in Round 6

Maybe next year!

2 thoughts on “A Wildly Average USATE

  1. Pingback: From Worst Tournament to Best Tournament Ever – chess^summit

  2. Pingback: Happy New Year from Chess^Summit: Looking Back – chess^summit

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s