How Caruana Conquered the Candidates

After more than two weeks of grueling competition between eight of the world’s best players, Caruana emerged victorious by a full point in the 2018 edition of the Candidates tournament, meaning that he will challenge current World Champion Magnus Carlsen in November later this year.  The tournament is structured as a 14-round, double round robin, and winning it is no simple feat.  In many cases, these players have been preparing for this tournament in lieu of playing in other events in recent months.

This year, Caruana rode to a +4 performance, scoring 9/14 with five wins, one loss, and eight draws.  The only loss came in round 12 against Karjakin, which made the race much more interesting as it sprung both Karjakin and Ding into the race; Mamedyarov, the runner-up at the time, lost as well.  Still, Caruana was always at least tied with the lead after round 4.

So, how was Caruana able to stay head and shoulders above the rest of the competition in this tournament?  We may (unfortunately) not be able to pick his brain directly, but we can look at his games in the tournament and see which ones had the greatest significance.

  1. Caruana – So, Round 1

This game was significant mostly for its embodiment of a “hot start.”  Fighting off dangerous counterplay from So in the middlegame, Caruana was able to catch a break after the critical move 23. … Ba6?!, giving him the opportunity to build up pressure on the kingside.  Caruana wasted no time in opening lines for his major pieces, and before long, he had a mating attack.  This huge turnaround likely had a substantial boost to Caruana’s confidence, setting the tone for him for the rest of the tournament.

  1. Kramnik – Caruana, Round 4

This game was significant for its situational implications.  Prior to this round, Caruana was trailing Kramnik by half a point in the tournament standings.  With Kramnik having White in this game, it was a critical moment for these two players.  If Kramnik was to win, it would set Caruana back 1.5 points behind the leader, making the rest of the tournament an uphill battle.  Thus, it was important for Caruana to keep pace in this game, and he did more than enough.  This game was another instance of the double-edged queen-less middlegame, somewhat of a commonplace in the Petrov.  Comically, the critical move was once again on move 23, when Kramnik lunged forward with 23. c5, attempting to undermine the support of the knight on e5.  However, when looking at the bigger picture, this endeavor was flawed, as Black was able to capture twice to get a passed pawn on g2 in exchange for a couple queenside pawns.  Compounded with back-rank issues, this became a troubling issue for Kramnik to defend.  Caruana won a piece, and with sufficient dark square control to prevent the promotion of White’s d7 pawn long enough and mating net ideas on the queenside, he was able to secure the win.  Especially with the Black pieces, this win was monumental for Caruana, as it also gave him the lead in the tournament.

  1. Caruana – Aronian, Round 13

After losing the previous round and allowing Karjakin and Ding to gain ground, this game was perhaps the most crucial of them all, as Caruana needed to right the ship once again with two rounds to go.  His opponent was Aronian, who, despite is poor performance in the tournament up to this point, was a solid player.  After a slow start with much maneuvering, Aronian ventured to sacrifice a piece for White’s three kingside pawns.  However, Aronian missed the key move 31. … Nxb4, which would have kept the pressure on White in protecting the kingside.  The misstep with 31. … e4 allowed Caruana to build up his pieces on the kingside and get an attack of his own, and he never looked back after that, eventually ending the game with a nice tactic.  This win proved crucial as Caruana was able to negate the loss from the previous round and keep at the front of the pack with one game to go.

Caruana fittingly ended the tournament in a Carlsen-like style, pushing and winning a game when he only needed a draw to clinch the tournament victory.  Nevertheless, he won the tournament and the right to challenge Carlsen for the World Championship title later this year.  Interestingly enough, the tournament comes with a quick turnaround as many of the players will play in the GRENKE Chess Classic in London, and Caruana-Carlsen in round 1 will play out as a preview for the upcoming World Championship match.

As always, thanks for reading, and I’ll see you next time!

Advertisements

Thinking about Chess Again: I Mean was I Ever Not?

IMG_0450
Yeah, they weren’t kidding about Gaucho. Hands down the best restaurant in Pittsburgh!

I mean do you even know me? Of course I’ve been thinking about chess this whole time! But with less than a month until this semester finally finishes, the difference is that I’m thinking about chess again – like for real …what?

This semester has been hectic at best for me: changing majors, managing the Pittsburgh Pawngrabbers, becoming a chess.com streamer, and on top of that, the usual class course-load. Tied down with all of the commitments, I had to put my goal of earning the National Master on hold, and in doing so, I have only managed to play one tournament game since the Cardinal Open (I’ll get to this later). So while I’ve thought about chess in some capacity every day, I haven’t dedicated as time to my own chess as much as I would have liked.

Admittedly, with less than four weeks until the end of the semester, I’m thinking about playing tournament chess again, and I couldn’t be more excited. I’ve started running regularly, eating healthier, and gotten back to regularly solving tactics. This alone won’t get me back to my best form, but it’s a manageable start – especially since I still have finals to study for. Speaking of tactics, I found a nice tactical shot at the end of my most recent chess.com stream:

What am I working towards? I’m planning on competing in the Chicago Open, and despite my eligible rating, I’m bypassing my chances of scoring big in the U2300 section to swim with sharks in the Open. In all seriousness, I’m going to be a massive underdog in nearly all of my games, but I want a chance to see how much I’ve improved since I last tried something like this at the 2016 World Open.

DSC_0352
Playing against Maggie Feng last January in Columbus (Photo Credit: Kelly Bloomfield)

For those of you who’ve been Chess^Summit readers for a while, you may recall how the 2016 World Open was not exactly pleasant to me. In the aftermath of my 1/7 score, my coach GM Eugene Perelshteyn had a field day finding weaknesses in my play, and while enduring six consecutive losses is an ego-bashing no chess player should be on the receiving end of, I learned a lot from the experience, and it parlayed into my later success in Europe.

So I’ve got to start somewhere to get ready for Chicago, and last week I built up the nerve to play my first tournament game in months without any prior preparation – and by preparation, I mean any studying. I’ve got some tournaments planned in early May, but I really didn’t want to wait anymore and guaranteed White against another 2100-rated player was just too good to pass up.

Wanting to sidestep any my opponent’s preparation, I chose 1 e4 for just my fifth time (in a standard game) since the 2017 Reykjavik Open, and it was clear I had succeeded once we reached the conclusion of the opening with 11. Qf3 – the Scotch Four Knights:

Screen Shot 2018-03-28 at 01.48.14
Steincamp–Park, position after 11. Qf3

While I failed to win, I built a reasonable advantage before squandering it after time control. Even with a few mistakes, I was more or less fine with my game – honestly, I was just happy to be at the board again. I’ll need to improve if I want to perform well in May, but knowing that I can play an opening I’ve never played before and do reasonably well is a good sign.

And with a somewhat amicable result, my preparation for the Chicago Open had begun. With just two months to go, I have a lot to do – but I’m mentally ready to make a comeback to tournament play.

 

K-1 Nationals: I love playing chess

I love playing chess, I play all the time on chess.com.”

This is the quote by 2017 K-1 Grade Nationals Champion Andrew Jiang.

Here is the full analysis of Andrew’s clinching game.

As kids grow up playing chess, at least in the beginning, many kids are more excited to show up than caring about the results.

However, as we grow, the pressure of winning becomes a baggage.

sunset-3225828_960_720

What differentiates kids and adults tournament is often the Excitement vs Result Spectrum.

Younger age: Excitement to play chess is 90%, results matters about 10%

As we get older, the reverse becomes true.

Many adults, including myself, would have nerve wrecking moments before the round, given the stake at hand. But for the K-1 warriors, it might be just another moment of an exciting chess day like any other.

Learn from kids, be excited to SHOW-UP.  Enjoy the process.

 

Draw Offers

Draw offers are an important part of the game of chess. Now, I won’t enter the debate about whether or not draw offers should be banned, but I’ll discuss my opinion about draw offers during the game.

I have a few stories involving draw offers. Once upon a time I got yelled at by my (higher rated and significantly older) opponent during the game for offering a draw. Long story short: I was freshly 10 and rated 1900+; the position was a dead draw to my eyes, and I offered a lot of draws. He was right. I indeed shouldn’t have offered draw repeatedly. I was right about the position being a dead draw, and we eventually drew. I never did that again. Lesson learned.

At scholastic tournaments, I’ve seen kids offering draws on practically every move, and that is simply obnoxious. Please don’t do that!! Another principle I learned was that if you’re defending or pseudo-defending, you shouldn’t be the one to offer a draw. Once your opponent gives up hope of winning the endgame and offers a draw, you should take it.

I’ve made a few hilariously quick draws (2 moves, 7 moves, 9 moves, etc.) in the last round of different tournaments, usually because either my opponent or I would win or tie for first place. That’s pretty typical for last rounds, and I’m quite sure just about everybody has done that at one point or another. Nevertheless, I’m not here to talk about my personal experiences, as those are usually pretty boring and have little or no instructive value.

First I would like to discuss the technical aspect of draw offers. The “proper” etiquette is to make your move and offer a draw immediately after it. Though there’s nothing illegal about offering a draw say 5 minutes into your opponent’s think, it’s not cool. If, on the other hand, your opponent offers a draw when it’s his move, then it’s perfectly within the rules for you to ask him to make his move, after which you can either accept or decline the draw offer. I’d recommend doing that, especially if your opponent has a tough decision to make.

As for draw offers themselves I’d split them into a few categories:

It’s equal draw offers

If the position is completely equal, then offering or accepting a draw is a natural thing to do. If, however, there is still life in the position and you want to play on, then play on if you feel like. This is not a justification to play rook vs. rook for 50 moves, as the only way you could win that in a game without increment/delay is if your opponent has a heart attack. What I’m saying is that if there is a reasonable chance that you will win even if the position is equal, then there’s no reason to agree to a draw.

Quick draws

One player offers a draw fairly early into the game, maybe around moves 15-20, in a position where neither side is clearly better and there is plenty of life left in the position. In my experience, most of these draws are in a tournament situation when neither my opponent nor I would win first place or any prize. I’ve accepted many of these draw offers, and a lot of people do that. Here’s an example I had this summer:

Gorovets

I was black. My position appeared to be more pleasant, but I didn’t have any concrete advantage. My main masterplan would be to orchestrate a minority attack with b5-b4 (after the bishop trade naturally), but white can stop it with Nf4-d3, after which I’ll have a hard time playing b4 and the white knight can land on c5. In view of that, I could play Nd6-e4, after which white would probably play Nf3-d2 trading off a pair of knights. Black has nothing—for the record, my engine evaluates this position as -0.1—and I accepted my opponent’s draw offer.

My stance on these kinds of draws is that as long as it’s a “here and there” kind of thing, it’s fine. A word of caution though: if you agree to a draw in less than 20 moves on a regular basis, then you’re not really learning anything from those games.

One side is worse draw offers

This is a typical scenario: player A is higher rated than player B, but player A is in big trouble and he offers a draw to try and escape. Player A is hoping that player B doesn’t want to risk it and wants to grab some rating points. Here’s an example from my own practice:

King

With his move 16.Bxe4, my opponent offered a draw. Black is clearly better, as white has several weaknesses (d4, h3, b2 could drop). White does admittedly have the bishop pair, but if black simply plays 16… Be7, white won’t preserve the bishop. Long story short, black has a risk-free edge, and white has a tough defensive task ahead of him. Nevertheless I decided to accept the draw. In retrospect I should have played on, but more on that later.

Kudrin

This is another game where I, as white, accepted a draw against a higher rated opponent. Unlike in the previous example, there is fire in this position. It could go both ways. The fact remains that white is better here. Much better to be precise. After 26.Rf1!, white has a variety of powerful ideas: Ng4 attacking the e5-pawn could be very strong. If black moves the king out of the way, than I could go Bh3 hitting the e6-pawn. Meanwhile, what is black going to do? While I can luxuriously play moves like Kb1, black is in dire straits.

Declining those kinds of draw offers can be hard. After all you are playing a higher rated opponent. It all depends how much better you are and how you feel about your practical winning chances. If you’re completely winning, then you really shouldn’t take the draw. If you’re only marginally better, it depends. In principle, you should play on. Even if you end up losing the game, what’s most important is that you learned something from the game. You can usually, however, come up with excuses/reasons to take the draw.

There are also times when the lower rated player offers a draw in a better position. I actually never really did that. There have been times when I’ve felt that if my opponent offers a draw, I’ll probably take it, but in those games I never found a good opportunity to offer a draw. I can’t remember the last time I offered a draw against a GM, and I don’t mind that. Most of my draws against higher rated players have ended after a) I was defending but survived, and my opponent offered a draw, b) it was equal for most of the game and my opponent and I squeezed the life out of the position, c) I was pressing but wasn’t able to win, so I repeated moves, or d) “quick draws” as described above. Where’s the room to offer a draw against stronger opponent? When I’m worse, I shouldn’t offer a draw unless my opponent is short on time. When I’m pressing, I should press.

Philosophical chat aside, what are my thoughts about draws in better positions? Honestly, I think it’s okay to agree to a draw in a better position as long as you don’t make it a habit. If you make a draw every time you get into a better position against a higher rated player, then you’re damaging your chess improvement. To get better at chess, you have to beat higher rated opponents and take risks. Period. If you’re pressing but can’t get through, and the game ends in a draw, then you’ll feel better than if you accepted that draw early on. It feels right. If you have a risk-free advantage, there’s no harm playing on. After all, you’re agreeing to a draw if you’re afraid that you’ll end up losing.

What about the higher rated player end? If you’re completely busted, then there’s no reason why you shouldn’t try offering a draw to a lower rated opponent. If you’re worse and don’t see a way out, offering a draw can alleviate suffering. Strangely, however, I can’t remember doing that in any particular game. I’m serious here. Nevertheless, offering a draw in a worse position as a higher rated player can make your life easier. If you’re in big trouble and your opponent offers you a draw, then you should take it. A draw is better than a loss. No question about it. If, however, you have swindling chances and want to gamble, you can turn down the draw. It’s your decision, after all.

There’s a special case called time trouble where you can forget most of the things I’ve said above. If you’re worse but your opponent has no time, then you could offer a draw. Your opponent might be relieved to accept it, and you’ve just snagged a half point. If you’re better and low on time and you feel things are going to go wrong, then you should consider offering a draw. If the position is unclear and complicated and both you and your opponent have no time, then offering a draw is reasonable. After all, it’s an easy way out of potentially disastrous mistakes/blunders in a time scramble.

For some reason I don’t offer many draws, and it’s hard for me to decline my opponent’s draw offers. Say you’re 250 points lower rated than your opponent, you’re a bit better, and he offers you a draw on move 15. Are you going to really be principled and reject my draw offer? You gain rating, you’re tired, you want to get some rest, etc… The reasons are just piling up.

Going through my game database today, I’ve also found that I’ve had more “chicken draws” in the past. Back in those days playing someone who was 200+ points higher rated than me wasn’t as rare as it is now. Once in a while I’d get a good position, my opponent would offer a draw, and it would be hard to resist… In principle I shouldn’t have accepted any of those draws, but just about everybody breaks those rules. I believe that I did accept too many draws. That’s my feeling looking back at those positions a few years later.

Here is my final conclusion. It’s okay to make chicken draws here and there, but don’t make it a habit. Everybody does it once in a while, but you should play positions out. How else are you going to improve your chess? If you’re not willing to take risks, then why are you playing chess?

Until next time!

How to (NOT) Play the Ruy Lopez

The Ruy Lopez has been considered one of the best openings to play as White due to its solidity and relative simplicity.  The opening has been around for many centuries, and some of the basic concepts and ideas in the opening are known to many.  Still, it’s not a two-result game every time White plays it.  Even with decent play from White, there have been many instances where very accurate play by Black has led to mind-blowing wins.

There was such a game quite recently.  The Candidates tournament is currently in progress in Berlin, where eight of the top players compete in a double round robin for a chance to challenge World Champion Magnus Carlsen to a match at the end of this year.  In the third round, there was a particularly crazy game between Aronian and Kramnik.  Let’s take a look:

Aronian – Kramnik, Berlin Candidates, 2018

That could definitely be one of the best performances by Black in the Ruy Lopez, if not the immaculate performance.  Granted, it did stem from some clever opening preparation, but with the knowledge that Kramnik didn’t go far into that specific line before going out of prep, it was a brilliant performance.  If there’s one thing we could take away from this game, it’s to never underestimate Black’s attacking prospects on the kingside if given the opportunity.   Kramnik didn’t hesitate to start attacking as early as move 7, and he never had to castle as he was always pushing with the initiative.

This also offers another instructive lesson – one cannot play opening moves in any random order, as playing certain moves earlier or later can change the dynamic of the position, allowing certain possibilities to come up that otherwise wouldn’t be possible.  For example, in this game, if Aronian had delayed castling and playing h3 in favor of Nbd2, Nc4, Qe2, etc, he wouldn’t have had to deal with a kingside attack from Kramnik.  It is little things like this that make chess the great game that it is.

And with that, I’ll see you next time!

Closing Thoughts on the Pittsburgh Pawngrabbers’ 2018 Season

What a season it’s been! With the Pawngrabbers falling short against the St Louis Archbishops last Tuesday night in the Round of 16, a historic run for Pittsburgh chess has come to an end. As this year’s team manager, I decided I wanted to retell some of the great moments from the Pawngrabbers’ PRO Chess League season while providing some behind-the-scenes insights as manager.

While I might not cover every detail from the past year, I hope I can share enough to help you appreciate what our players achieved this year.

Pittsburgh Offered Immediate Qualification to the 2018 PRO Chess League Season

2018 Announcement

In the closing days of my European Chess Tour, I received a notification that Pittsburgh was invited to participate in the 2018 PRO Chess League. As the League was consolidating to a more competitive 32 teams, it wasn’t so clear that we would get the bid. While we had an engaged fanbase on social media, the only Grandmaster on the roster at the time was Alexander Shabalov.

After watching us start the 2017 season 0-4, a lot of the League’s spectators bemoaned our invitation, perhaps forgetting that we won each of our final three matches. It may well have been that our narrow win over the Minnesota Blizzard to close 2017 spared us from having to enter the PRO Chess League through the qualification tournament… who knew that match nearly cost one of this year’s quarterfinalists a berth in the League?

Needing to prove our place in the League, the search began for additions to revamp the Pawngrabbers’ roster. With the season’s new expanded local rules, we added NM Mika Brattain from the relegated Columbus Cardinals fairly quickly, while also signing FM Mark Heimann and NM David Itkin from the area. This alone increased our team’s average rating significantly – but likely not enough to avoid relegation. Luckily for us, through extensive note-taking and stream watching, October’s conclusion meant a whole new batch of players were looking for teams.

MikaCard
NM Mika Brattain was one of the first new additions to the Pawngrabbers, and he finished 2018 with a 2430 performance rating.

Following the conclusion of the League qualifier, we were able to continue recruiting 2300+ strength local players, thanks in part to Michigan’s failure to get a team in the League. With that, Pittsburgh acquired IMs Atulya Shetty and Safal Bora, as well as FM Edward Song, who debuted for us in 2017 as a free agent. Now as local players, my managerial team (Beilin Li, Grant Xu, and myself) could begin scouting Free Agent talent to complete the roster.

AtulyaCard
IM Atulya Shetty proved himself to be an extremely solid board 3 for Pittsburgh, scoring 50% against much stronger competition.

From the outside, you all likely know the story – Pittsburgh signed GM Awonder Liang days before the start of the season, and the Pawngrabbers took off. However, for a few weeks it seemed like we were stuck with the line-up we had, as time-zone differences and financial constraints were proving to be a constant limitation for us. It was during this time we developed various match strategies to compensate for a much lower average rating.

Breaking Down the Gs and gs in Line-up

If you’re familiar with the League, you likely know that a team’s line-up for any given match is limited to an U2500 rating. Of course there are certain exemptions from this, but since we’re talking about a generalized match strategy, let’s pretend the highest rated players in the world are exactly 2700. With the exception of our clashes with Webster and St Louis, this assumption held true for all of our regular season matches. To represent this symbolically, I will use ‘G’ to indicate a player roughly rated over 2500, and ‘g’ for all other players.

The GGGg line-up proved to be the dominant strategy in 2017 as St Louis employed it throughout the playoffs and against Norway in the Championship match which they won, 9-7. This makes sense – you field three top tier Grandmasters, and hope your board 4 notches a point somewhere, totaling to at least 8.5/16 to win the match.

2-768x607-2
2018 proved to be quite the year for team chess in Pittsburgh, as Carnegie Mellon won the USATE in February. Photo Credit: Vanessa Sun

But the consolidation of the PRO Chess League to 32 teams also meant the average rating for each team grew. Now with 16 fewer teams (and 64 fewer boards on any given week), the demand for a spot on a team drastically outweighed the supply of talented players. As we saw in the case of the Archbishops and the Gnomes, the GGGg format was not so invincible, as more teams had two (or even three) grandmasters to field on any given week. The match break down between a GGGg and GGgg lineup (assuming both teams optimized their line-up to fit the U2500 constraint) would look something like this:

Team A (GGGg line-up): 2700, 2650, 2500, 2150*

Team B (GGgg line-up): 2550, 2550, 2450, 2450*

*maximum possible rating, given average constraint set by the three preceding boards

Ideal match strategy for Team B is to win each of the four games against Team A’s board 4, meaning that now Team B only needs to score 4/12 for the remainder of the match to avoid a loss. Given the rating gap between Team B’s players and the 2150 on Team A, I’d say this is highly likely – heck, I could probably give the 2150 a reasonable game, and I’m not PRO Chess League material.

To compensate for this, Team A’s 2700 must offset their 2150 by scoring 4/4, which is far more difficult than Team B’s task of beating the 2150 four times with four different players – only one player managed to do that against our line-up this season, and he happens to be the 2013 World Blitz Champion, GM Lê Quang Liem from Webster.

For the sake of evaluating match strategy, let’s give the 2700 the benefit of the doubt and assume he succeeds scoring 4/4, regardless of the actually probability of that happening. Here are the players’ remaining number of games for both teams:

Team A (GGGg line-up): 2650 (4 games), 2500 (4 games)

Team B (GGgg line-up): 2550 (2 games), 2550 (2 games), 2450 (2 games), 2450 (2 games)

With the match (theoretically) tied at 4-4, Team A’s best player is a non-factor to the outcome of the match, while all of Team B’s players are still capable to continue contributing to the score. All the sudden, Team A’s chances are winning the match are drastically reduced – if their Board 3 scores anything less than 2/4, the 2650 needs to make up the difference.

In this set-up, it just takes one player on Team A to have a bad day, and as we saw in 2018, this strategy worked for Pittsburgh and Minnesota against St Louis in the regular season, and Norway even fell shy of qualifying for the postseason. In fact, this is exactly how the Blizzard toppled Webster in the Round of 16 last Tuesday.

AwonderCard
GM Awonder Liang gave Pittsburgh the second “G”, making the Pawngrabbers a much more formidable force in the Atlantic Division

When Awonder signed with Pittsburgh, we had our two G’s, but Atulya was our next highest rated player at 2403, meaning we could never match the desired line-up strategy on paper. Luckily, with Atulya consistently playing above his level, it meant that we just had to find a board 4 who could consistently score more than 1.5 points a game. This took all season, but in the end, we were surprised by how many players on our roster that could fit the role. I’ll discuss some of the shortcomings of the GGgg line-up in a bit, but lets start talking about actual results, and less about theory.

Pittsburgh Gets off to a Strong Start

We got to test our GGgg strategy immediately against the Buenos Aires Krakens, as the Argentineans fielded three strong grandmasters in Federico Perez Ponsa, Alan Pichot, and Leandro Kyrysa. Of course, Buenos Aires would eventually be relegated from the League, but given that they brought the same line-up to the 2017 quarterfinals, we held our breath for much of the season opener.

PittBuenos_FinalScore

Two hours later, Pittsburgh had its fourth consecutive PRO Chess League win, overpowering the Krakens on all boards 10-6. Following the script, Pittsburgh held Buenos Aires’ 4th board to 0/4, and while Federico Perez Ponsa notched 3.5/4, the aggregated total between their boards 1 and 4 was 4.5-3.5 in favor of Pittsburgh. With Shabalov and Awonder each scoring 3 points, Atulya’s 2.5 were enough to clinch the match before taking into account Ed’s tactical shot in his fourth round win:

The win proved to be a confidence booster for the team, and it quickly carried over against the Montreal Chessbrahs in another decisive decision. Awonder won all four games and produced a masterpiece against GM Robin van Kampen – an early sign of things to come for the 14 year old US Junior Champion:

While IM Michael Kleinman notched an impressive 1.5/4, RvK’s 2.5/4 forced Montreal’s second and third boards to perform, and they fell short. Once again a victory for the GGgg line-up. Of course Montreal would also go on to be relegated, but at the time, they too seemed like strong League title contenders given their prior semifinal finish.

The Defining Stretch

The next three weeks proved to set the course for the season, as the then-Atlantic division leaders Minnesota Blizzard, Super Saturday, and the St Louis Archbishops stood in our way of the season mid-point. Three consecutive losses would have likely derailed our playoff aspirations, so the team’s performance was critical in these next three outings.

Despite some late game heroics from the team, Pittsburgh fell short to Minnesota in our only loss of the regular season, 8.5-7.5. The loss was tough, but Awonder produced arguably the PRO Chess League’s most entertaining game of the season with his sac-sac-mate win over surging IM Sean Nagle:

With Super Saturday approaching, we signed bullet specialist IM Tuan Minh Lê to join the team. While Minh’s heroics impressed against superior competition, it was Awonder who muscled the Pawngrabbers to a half point, with critical wins over Nakamura, Dominguez, and a draw against Yu Yuangyi. In just four hours, Awonder became an icon in the PRO Chess League.

GM Eugene Perelshteyn from ChessOpeningsExplained.com and I got a chance to review Awonder’s games days after the event concluded:

A draw was a fantastic result given Shabalov’s absence from the line-up, and the confidence boost proved to come at the right time as we faced off with St Louis. As a manager, this was the real test for the GGgg line-up, as we had designed this strategy specifically because of juggernauts like St Louis and Webster – we were never going to out-rate these players on paper.

St Louis brought top grandmasters Fabiano Caruana, Vladimir Fedoseev, and Alejandro Ramirez – two 2700+ rated players. At one point, chess.com predicted we had only a 9% chance of winning the match!

StLouisLineups

One of the downsides of the GGgg strategy is simply that you don’t know if you’ve succeeded in shutting out the fourth board until the match is over. Thus playing out the match on paper feels a lot different than in real-time, as on paper, you have the knowledge that Boards 1 and 4 cancel each other out with 8 games remaining. This distinct lack of knowledge always favors the GGGg line-up, as going into the last round, the GGGg line-up is always favored to score more points than the GGgg one.

Admittedly, I was pretty nervous as I ran a solo commentary stream that night, but in the end, the strategy worked exactly as it did on paper. With Awonder’s win over Caruana, Fabiano maxed out at 3 points, which when aggregated with NM Forest Chen‘s last round win, cancelled out Fabi’s contribution to the match. This left Vladimir Fedoseev and Alejandro Ramirez with the final 8 games, in which we won the aggregate total 4.5-3.5 to secure victory.

Even better, Pittsburgh had the lead going into the final round to account for the head-to-head on-paper advantage St Louis had. Brilliant. Shabalov put together one of the most crushing positional wins I’ve seen against Vladimir Fedoseev in the third round.

bEIhVGtBMFNndlpKb3c5MGZvOCFibFlRZWdYUGpyNlhjalFJQUpTSnZLNzZhYzVaS3Q2UUJJUElyelhPekk0NWxyUXlJUU90ZHR5aWpUWlRyQmlHUVk1NnRIMHpIR3pHQlFHUFEwIT8wNjk2b3g2WWNZUFlmY1lQYzZUIXhvMlVvSj8yNjUhVEp2VU
Alexander Shabalov–Vladimir Fedoseev (PGH vs STL)

After the match, I congratulated him on his 3/4 performance on the night, to which he responded: “Safal won us the match”. IM Safal Bora had a tough night, only scoring 0.5/4, but his half point made the difference. Shaba was absolutely right – this was a team win. After this match, the focus of the team centered on the over-performing third and fourth boards – they were gaining confidence and had become an asset for the team.

Power of the Bottom Half

As the season progressed, it became clear that Awonder’s efforts needed to be supported by a strong secondary (“g” players). Atulya quickly did his part against Montclair, notching 3/4, including a win over Africa’s first ever 2700 rated player GM Bassem Amin.

bUMwU2x0WUlvdzVRZm8yVWd2OTJlZyEwa3M4IWZlU0tpcVpSanpXT2JsNlNhYjNWbEFYSEF1N1pzQVFCdkJJQnVKMEpDSlNFZHJIQXRBRUxiakJ0bnZWTmN1OThqbEtDdkNMQ29DOENydDQ4bG1aRWdmRXZmZzJC
Bassem Amin–Atulya Shetty (MCR vs PGH)

Atulya had been a strong weapon for Pittsburgh, with a performance rating consistently north of 2500, but now with a playoff bid in sight, who would take on the role of the fourth board? There were too many options and not much time left, so the strategy shifted and the focus for the team became finding the right fit.

Super Saturday saw the Pawngrabbers clinch a playoff berth with a 4th place finish, despite missing both Shabalov and Atulya in the line-up. The early story of the event was the dominance of tandem Awonder Liang and Mika Brattain:

But by the end of the day, FM Edward Song impressed the most, scoring an unbeaten 6.5/8 from fourth board. Admittedly, he barely made the line-up for the event, but he got to spend the following four hours proving me wrong over the board. IM Tuan Minh Lê also returned for the Pawngrabbers, scoring an impressive 5/8 against Grandmaster competition.

His crowning highlight was this jaw-dropper against GM Helgi Olafsson:

Because of the limitations on number of Free Agent players during the playoff matches, this would prove to be Minh’s final event with the Pawngrabbers in 2018. Had we qualified for the semifinals in San Francisco, he would have been on the line-up (as allowed by League rules) – Shabalov, Awonder, Minh, and Atulya, chiming in at a 2496 average rating. Who knows what that line-up could have accomplished?

While upsetting St Louis will likely be remembered as our signature achievement this year, our next two matches were also quite noteworthy.

PittWEB_FinalScore

Despite Webster’s deep roster, our clash featured an all-GGgg match, and Pittsburgh won in a nail-biter, thanks to a 3/4 performance from Ed. His win over FM Joshua Colas is most memorable for me, as he swiftly dismantled Black’s Sicilian:

bUNZSWd2MFNsQklCdkJXT2Z0IVRlZzdZZG1aUmtBMlViczkyQnY1UWNEOCFhY1FLdktSS0R1NlpqelROb3cxTENMU0xzSlk3SlA0NWZkOTFBSTc4UFoxWnRBIT9kWjhaY2RaNnVNNjhtbExEbFpOVFpZVENNNzU0WVg0N2Q3ODdYQzJUQU83bE9tbG
Edward Song–Joshua Colas (PGH vs WEB)

Discounting his (narrow) loss to Lê Quang Liem, Ed was now unbeaten in 11 consecutive games and his time management had improved drastically. The match win broke Pittsburgh’s extended losing streak to Webster – one that extended far beyond the PRO Chess League.

The closing week match-up against Miami presented us with the toughest challenge we had all season as Shabalov, Awonder, and Minh Lê were all unavailable. Thus in one week, we had to plan out the gggg vs GGGg strategy. We knew Miami, trying to avoid relegation, would bring their best possible line-up, which boasted Iturrizaga, Quesada Perez, and Becerra.

In the gggg vs GGGg pairing, it is absolutely critical that the gggg outfit score 4/4 against against the opponent’s lone “g”. Failing to do this makes the likelihood of winning close to impossible, which among other things, is one of the main reasons why this kind of line-up is inferior and not sustainable long-term.

However, rather than focusing on cancelling out the top board, the gggg strategy now calls for every player to score 1+/3 in the remaining pairings to get to a minimum 8 points.

MarkHCard Debut
Unable to play for much of 2018, Mark will enter 2019 as one of Pittsburgh’s newest weapons

The team held on tight, and thanks to FM Mark Heimann’s 3/4 debut, Pittsburgh hit an 8-8 draw, as Miami rightfully avoided relegation. With Minnesota getting bashed by Montclair, Pittsburgh locked up the second seed heading into its Round of 16 clash with St Louis.

House of Cards

Unfortunately, you all know how this one plays out. St Louis got out to an early lead, and with a 7-5 advantage in the last round, and Pittsburgh failed to score three points to take the upset. While a 10-6 loss does seem like a blowout, the match was actually a lot closer on paper. NM Forest Chen’s lone win cancelled out Fedoseev’s 3/4, leaving Akobian and Zherebukh for Pittsburgh.

Even with Ed putting up another monster 3/4 performance on board 4, Zherebukh continued his dominant League form with 3.5/4, and Pittsburgh couldn’t close the gap on Akobian. Shabalov had a tough night on the top board being an underdog in three of his games, and the Pawngrabbers couldn’t seem to catch a break for the entirety of the match.

You can rewatch the match in full here:

This match was humbling as it showed some of the shortcomings of the GGgg vs GGGg match strategy. Thinking long-term, our line-up did ask our players to consistently over-perform, and while they succeeded for much of the season, inevitably there was some burnout. This makes sense – while a player may initially score a few upsets, over time that player’s performance rating will regress to their expected performance level.

For our “g” players, we managed to avoid this by constantly rotating them. Funnily enough this was intended to decide which player we wanted on board 4 for the Round of 16, but it benefited all of our players in the long-run (at least on paper).

Does this mean the GGgg line-up is impractical? No, I don’t think so, and for the PRO Chess League format I think both the GGgg and GGGg lineup have their distinct advantages. As teams continue to get stronger, I think at some point both of these strategies will become obsolete – ratings in long-time controls don’t truly represent how a player does in an online rapid format, and it will be up to team managers to decide which players are too highly rated and which players are diamonds in the rough. At some point, teams will feature a GGGline-up, where “G is a hybrid player – a “g” rated player who consistently performs at a “G” level with no burnout.

Finding such a line-up will take a few seasons of data, and naturally runs the risk of said “g” rated player becoming “G” rated. That is the fun of the U2500 average!

Outlook for Pittsburgh

Okay, this is turning out to be a longer article than I envisioned, so let’s put aside the theoretical “improbability” of Pittsburgh’s success in 2018, and talk about next year’s potential. Even with the abrupt end to our season, I think Pittsburgh will be much stronger force next year in the League. FM Edward Song finished the year on a high, but NM Mika Brattain and FM Mark Heimann also showed they are more than capable of playing at a high level this season. IM Atulya Shetty will continue to give the Pawngrabbers an anchor.

JenniferCard
FM Jennifer Yu only played one match for Pittsburgh in 2018, but she may see more action as she continues to improve at an impressive rate!

At the rate she’s improving, I think FM Jennifer Yu will also become a much more frequent member of the Pawngrabbers’ outfit in 2019… and that only rounds out the potential for boards 3 and 4!

Of course its much harder to predict what will happen on the top boards during the offseason, so I will rightly keep my mouth shut about our options and new enhanced match strategy we are already developing for 2019. Stay scared, PRO Chess League, Stay scared.

Follow the Pawngrabbers on Twitter!

Want to support the Pawngrabbers? Become an official Booster today!

Want some PRO Chess League merchandise to support the Black and Gold? Click here.

 

AI and Chess

Here is a self-composing music using AI.

I watched an AI documentary on my flight back from China, where I learned about the self-composing music using AI.

My immediate question was can this technique applied to chess as well?

The possibilities are certainly there.

ai_pic

AI and Chess

Deep Learning and AI has been the topic in the tech world. Ideas from self-driving cars to language translations have expedited the hype.

Chess had its own moment in the news, namely AlphaZero, where DeepMind stepped aside from the game Go to join the chess research.

AlphaZero not only took down Stockfish in record time, what’s more impressive is the new approach it brought to the game.

AI Applications in Opening Prep

Part of a chess player’s growing pain is how to prepare an opening repertoire. The vast amount of possibilities often overwhelm a strong professional player, needless to say, it’s a much more painful chore for club players.

What if there is a machine that can self-learn opening styles from top players, and then provide a repertoire based on a student’s preference or his/her chess idols?

What if once that repertoire is ready, it can be imported to chess.com or other chess databases and can be easily accessed?

One reason of the popularity for AI is the cool applications, but another is the accessibility for us mere mortals to get our hands on and experiment with the technology.

We are just at the beginning of the AI advancement, and as technology progresses further, the possibilities would only increase!

Happy brainstorming!

———————-

For those who are interested to explore further, below are some references.

Chess Reinforcement Learning: https://github.com/Zeta36/chess-alpha-zero

Tensorflow: https://www.tensorflow.org/

This Week in Chess: http://theweekinchess.com/twic