Better Position != Winning

I don’t get to play Titled Tuesday much on

When I do, I always want to make use of the opportunity to brush off some rust and study new opening ideas.

Here is the Full Game Analysis


To summarize:

I played an interesting opening and got what I wanted out of the opening. In the middle game, I had the tide on my side and would have made small improvements to make it less complicated.

When it turned to the endgame and time scramble, nerves got in the way, and I couldn’t finish off the game.

In the end, mission to brush off rust and study new openings were accomplished for this game, but the result was not satisfactory.

This game helped me to feel even stronger about how blitz can help me improve.

For anyone in the Southeast, CCSCATL will have an in-person Blitz Tourney on Saturday, July 21. Hope to see you there!


Accept Draw, Reflect Later?

Unfortunately, useful statistics on draw offers are hard to find, but my guess is that on people believe draw offers are too common, on average. I am inclined to believe this myself, if nothing else because I have noticed a lot of players (at lower levels) accepting draws in clearly better positions. Of course, when we’re at the board ourselves, our perspectives are likely to change as our natural (for many of us) human risk-averseness kicks in. While I was stuck around 2000 USCF, I couldn’t even hold my winning positions against higher-rated players, so it was hard to imagine I could be upset with a draw against anyone 2200+ or so.

That attitude has changed dramatically in the 3 years since. Nowadays, I play more open sections where I’m the underdog, and thus more opponents that are not inclined to draw without a good reason. This occasionally presents some interesting dilemmas.

Screen Shot 2018-07-14 at 10.02.53 AM
Li (2159) – Velikanov (2403)

In this game (March 2018, Chicago), I managed to surprise my much higher-rated opponent in the opening, and he offered a draw in a very difficult position. Against most players, this wouldn’t be much of a decision, but a rare easy draw on move 12 against a 2400 seemed pretty welcoming. Objectively though, White has a huge advantage, and I concluded that if I couldn’t at least draw this position, I didn’t deserve the draw anyway. So I played on, and there were no regrets over that decision, although I royally messed up a 3-pawn-up rook ending later on and had to settle for a perpetual.

However, my fear of messing up such an advantageous position did not come out of nowhere. It’s quite possible that I simply had bad memories of declining a draw only to lose the game later. For example:

Screen Shot 2018-07-14 at 2.52.27 PM
Linde (2096) – Li (2093)

In the deciding game of this team match from the 2017 Pan-Ams, my opponent offered a draw after playing 19. Rd1. Black has avoided serious weaknesses and is getting close to a freeing …d6-d5 break, although that loosens up his pawn structure a bit and requires care. Probably the position is about equal. It came up in our post-mortem that my opponent wasn’t feeling well at the time, which explains the draw offer. Not knowing that, I decided to play on, as I felt pretty comfortable positionally, and I wanted to play a good game after playing mostly way up or way down the whole tournament. Unfortunately, despite taking the advantage late in the game, I blundered a piece in time trouble and lost.

The same scenario would play out twice more during an otherwise stellar 2018. At the Philadelphia Open in March, I decided to press on in a very symmetrical ending against a 2100, only to hang an Exchange and lose. And a month later in a rapid tournament, I declined a draw – and clear first – against a slightly lower rated opponent in a somewhat better position. This backfired yet again when I went on to reverse a winning position by hanging a piece in time trouble!

So by the time summer rolled around, I was feeling pretty cautious about the whole draw offer concept. Especially true at critical moments of the Chicago Open and National Open, which I talked about in my last two posts.

Although I escaped the humiliation of losing after turning down the draw offers, it started to dawn on me that I might have veered too much in the other direction. In the first game, my opponent took a long think before playing 18…Ne5 and offering me a draw, which I accepted despite a better position where he had only 40 minutes left. The second was in someways more questionable. I thought my only option was to repeat with 15…Qa5+ 16. Bb4 Qc7 17. Bd6 and take a draw in a position where my opponent only had 18 minutes. In reality, after 15…Qa5+ 16. Bb4, 16…Qf5 was completely fine (I was worried about my queen being vulnerable on the kingside, but White is not developed enough to exploit this). Unlike the Chicago game, this created some tournament difficulties as I had 2.5/4 and had little room for winning money.

However, in spite of my excessive caution in both games, they exhibit the point that there is often no right answer in these draw offer dilemmas. Giving up a half point, after all, is much less disastrous than giving up a full point, and I ended up surviving both of my decisions, as in Chicago, I was primarily playing for improvement, and in the National Open, I ended up winning the rest of my games to tie for 4th in my section.

For most ambitious and alert players, premature draws clearly should not be too regular an occurrence. But it is always important to keep one’s ambitions in check (as I learned early on!) and realize as I did later, that thinking about draw offers does not have to be agonizing, merely a learning experience. This is, of course, important not only to offering and accepting draws, but to all of chess.

How to Bomb the World Open 101

DISCLAIMER: This article is NOT an endorsement of the strategy exhibited by the author. Furthermore, the author has no intentions of writing a “How to Bomb the World Open 201” article. Ever.

The World Open is one of the largest open tournament in the US, and it’s quite an event. The $20,000 first prize in the open section attracts dozens of GMs whose participation brings norm hunters like me flocking. I was feeling fairly optimistic after three good tournaments in a row, and I was looking forward to playing 9 rounds of good chess, staying near the top boards, and fighting for a GM Norm.

Instead, I followed a different strategy. Here’s a quick summary:

  • Draw the first game
  • Do your best to lose the second game but save it in the end
  • Get destroyed in the third game
  • Do all of that against significantly lower rated opponents
  • Your tournament is ruined in a record 3 rounds, and there’s no need for more butchery
  • You may now resume playing your regular chess

Simple, isn’t it? Here’s how I pulled it off.

In round 1, I got black against Eddy Tian (2223 USCF, 2083 FIDE). I got a comfortable edge out of the opening with black, but I wasn’t able to exploit it.

Tian 1

All of black’s pieces appear to be active and well placed, while the same can’t be said about white’s counterparts. Unfortunately, black doesn’t have a clear plan to exploit them, but white doesn’t have anything convincing either. When my opponent played 26.h4, I decided to open a second front with 26… h6!?. It’s not a bad idea, but I butchered the execution. The game went 27.Re1 g5 28.hxg5 hxg5 29.g4 Kg7 30.Kg2

Tian 2

I thought the natural move 30… Rh8 would be met with 31.Rh1, leading to a rook trade. Black is on top over there, but I thought I had something better. I went 30… e5? with the idea of meeting 31.Rh1 with 31… e4!, after which white’s rook isn’t very useful. However, I ran into 31.Qf5! which is the flaw in my idea. 31… Qxb3!? may be black’s best. Despite being seemingly suicidal, white doesn’t have anything concrete besides taking the g5-pawn which will lead to complications. Instead, the game went 31… e4 32.Nxg5 Qxf5 33.gxf5 Rd5

Tian 3

Black is winning the pawn back, but after 34.f3! he has no advantage. I tried to win for 25 more moves but didn’t get anything concrete. That was a blow to my tournament plan. For one thing, this was going to hurt my opponents’ rating average for norm chances. Still, this was only one game, and I felt I hadn’t played so badly.

I thought my round 2 game against Prateek Mishra (2166 USCF, 1992 FIDE) would be a fairly smooth win, since I had a large rating edge and the white pieces. I was dead wrong. I stumbled into old theory which I had to figure out over the board, and I didn’t do a very good job. To be more precise, I was practically lost my move 25. Eventually we reached this position:


White’s position is on the verge of collapse here, and a lot of moves (32… Rd8, 32… Qxb4, 32… f3, etc.) would’ve finished me off. Fortunately, my opponent let me off the hook by playing 32… Qxd4?. After 33.Qxd4 Rxd4 34.Rxc2 Rxb4 35.Rxc7 white has enough activity to equalize, and the game soon ended in a draw.

This was so not my plan. Half of me was relieved that I hadn’t lost, but  the other half of me was utterly disgusted with myself for playing such an awful game.

In round 3, I got black against Zhaoqi Liu (2381 USCF, 2118 FIDE), and I was desperate to win. Watch me get crushed.

Zhaoqi 1

This position is fairly imbalanced. The d5-square is a juicy outpost for a white knight, but black has the bishop pair and is fairly active. Objectively speaking, white may have a slight edge, but there’s plenty to fight for.

With his last move, my opponent attacked my rook, and I replied with 15… Rd7?. 15… Rde8! was stronger. 16.Qxd6?? loses to 16… c4+, and white will likely play a normal move such as 16.Ng3. But wait, what’s wrong with 15… Rd7? You’ll see…

16.Qf5 hits the rook and clears out the d5-square for the white knight. I replied 16… Qb7, thinking that everything was under control. 17.Nd5?? hangs a piece, and 17.Nf4 is met with 17… Nd4! 18.Qg4 f6, after which black is on top. However, I hadn’t considered my opponent’s next move seriously: 17.Bf6!! was a nasty surprise.

Zhaoqi 2

17… gxf6 is suicide on account of 18.Nd5!, and 17… Bxc3 18.Nxc3 doesn’t seem to help black. Because white is threatening 18.Qg5 with mate coming soon, I played 17… Ne7 which was met with 18.Qg4 Ng6 19.Nf4!

Zhaoqi 3

The second wave of attack comes, and it’s really powerful. Black won’t survive after 19… gxf6 20.Nh5, as on top of everything, white has two knights that can join the melee. What else to do? Next up is 20.Nh5 or 20.Ncd5 smashing my kingside. There was one move that I had, 19… Bxc3!. After 20.Bxc3 I totally hated my position, but after 20… f5!? there may be hope. This was necessary, but I couldn’t get myself to do it. I had gone into a complicated position with the hope of winning, and now I was groveling for chances to stay in the game by move 20!?

After a long think, I played 19… Qc8? which loses fairly quickly. My idea was to trade off queens with Rb7, but that doesn’t help much. After 20.Nd5, I realized that 20… Rb7 runs into 21.Nfxg6 Qxg4 22.Ne7+! winning an exchange, and white has similar alternatives that also win. Therefore, I played 20… Re8 21.Nh5 Rb7, but after 22.Bxg7 I’m dead lost. I decided to allow checkmate by playing 22… Qxg4 23.Ndf6#

Zhaoqi 4

Ouch!! That one still hurts!

I had 1/3, and my tournament situation was awful. I was losing a lot of rating, and of course my norm chances were beyond extinct. I wasn’t playing well at all. I considered withdrawing, but my gut told me not to. I felt that once I got the bug out of my system, my chess would be back to normal.

Fortune smiled down upon me, and I managed to win my next three games which were fairly decent overall. I had 4/6 with 3 rounds to go. This was a fairly decent score by my standards, except that how I got there was the problem. In round 7, I did get a chance to play up against IM Gabriel Flom (2551 USCF, 2515 FIDE). I was hoping to hold my own this game, but things went wrong from the very start.

Flom 1

After an offbeat opening, I thought I should be fine here after the recapture 8.exf4 dxe4. Unfortunately, I completely missed 8.Qg4!, and I quickly realized that I had messed up big time. Black is going to lose a pawn. 8… dxe4 9.Qxg7 Ke7 10.exf4 didn’t look fun to me, 8… 0-0 9.Qxf4 dxe4 10.Qxe4 looks even worse, and 8… g5 is nonsense. My best shot was 8… Bd6! 9.Qxg7 Ke7. Though I’m a pawn down, and my king is on e7, it isn’t that bad for me. I can always trade queens with Qg8. Instead of that, I went 8… Qf6?! which was met with 9.Bxd5! 0-0 10.Qxf4 Qxf4 11.exf4 exd5

Flom 2

Black is a pawn down here, and he has a lot of suffering ahead of him in this endgame. Not surprisingly, I went down. It was really unfortunate that I blundered like this and spent practically the entire game on the back foot. Despite this humiliation, I managed to win two fairly smooth games against 2300+ USCF opponents to finish the tournament with 6/9.

That was actually higher than I’d ever finished at the World Open—in 2016 and 2017 I got 5.5/9. Nevertheless, thanks to my disastrous start, my misadventures cost me quite a few rating points.

My summer adventures continue on the West Coast at the US Cadet which starts tomorrow. Fingers crossed.

GM At Last!

After a long, suspense-filled journey, chess phenom Praggnanandhaa has officially become the second youngest to ever achieve the grandmaster title.  To say the least, it was a very long road.  Nevertheless, the achievement is still magnificent, although possibly bittersweet for him.

Image result for praggnanandhaa chess
Praggnanandhaa at the 2017 World Junior Championships

Let’s take a trip through time to examine how Praggnanandhaa came to reach such an achievement at this incredibly young age.  Praggnanandhaa first broke onto the scenes as a FIDE Master just before turning 8 years old when he won the U8 Open section at the Asian Youth Chess Championships in 2012.  Fast-forward three years later, and Praggnanandhaa became the youngest-ever International Master in 2016 at the age of 10 years and about 10 months after gaining the third and final IM norm at the KiiT International Chess Festival.  At this point, the chase for the GM title was officially on, and to break Sergey Karjakin’s record at 12 years and 7 months, Praggnanandhaa had about one and three-quarters years to gain three GM norms (performance rating of 2600+) and peak his rating above 2500.  Plausible, right?

The First GM Norm

Praggnandhaa earned his first GM norm at the 2017 World Junior Chess Championships.  This was a big story, but it was probably overshadowed by the potentially bigger story – one that never materialized.  This event was (and still is) one of several around the world that offer immediate titles to the top finishers of each section.  In the top section, the first-place prize was, in fact, an immediate GM title.  Unfortunately, Praggnanandhaa fell half a point short with “only” 8/11.  Still, he had finally gotten the first norm that he needed, and he had about five months to gain two more in order to beat Karjakin’s record.

The Second GM Norm

After winning his first norm in November 2017 and with a large number of events coming up, many believed Praggnanandhaa could feasibly gain his last two norms before March 2018, when he would become the same age as Karjakin when he won his title.  Unfortunately, Praggnanandhaa never seemed to catch a lucky break, and he came very close on many occasions, but could never seal the deal.  Thus, the March 2018 deadline came and passed.  But, it definitely wasn’t the end of the world for him, since Praggnanandhaa still had six months to beat the second-fastest time, which was held by Nodirbek Abdusattorov of Uzbekistan.  Indeed, about a month after the deadline in April of 2018, Praggnanandhaa gained his second GM norm at the 4th Heraklion “Fischer Memorial” GM-Norm tournament, finishing in clear first half a point ahead of the rest of the field.  He only had one more to go.

The Third GM Norm

The band Linkin Park has a song that goes: “Night gets darkest right before dawn / What don’t kill you makes you more strong / And I’ve been waiting for it so long.”  Indeed, that seemed to ring true for Praggnanandhaa, who had one of his worst performances in early June at the Schaakweek Apeldoorn GM tournament in the Netherlands, going a frightening 3/9.  Yet, all seemed to work out in the end, as later in the month, he played in the 4th ad Gredine Open in Italy and captured the last GM norm and the GM title in the 8th round of the tournament.  To put on the finishing touches, he won the last round as well and tied for first place in the end.  A fitting finish.  Congrats to Praggnanandhaa!

While he wasn’t able to break Karjakin’s record in the end, Praggnanandhaa’s journey was still fascinating and fun to follow.  And, what’s bad about being the second fastest to GM?!

Some now even call him the 2nd Tiger from Madras, as his hometown is in Chennai, India, and it happens to be the same as former World Champion Viswanathan Anand.  Those are big shoes to fill, but it seems very possible as he is yet to 13.  Once again, congrats to Praggnanandhaa, and I’ll see you next time!

Product Development for AI and Chess

Product and feature refinements are part of the development cycle for every technology company.

As chess gets popular and into the mainstream tech and AI world , it has many potential to be integrated as part of the software product features.

To build on top of Chess and AI in general, here are three ideas I had brainstormed

  1. Automation of tedious tasks
  2. Image recognition to seamlessly integrate physical and online world
  3. Chess Self-learning from the machines


There are thousands of games played daily in the chess world, and a few hundred will be published in TWIC.

Stronger players would like to dig thru all the latest trend in the openings, and new players would like to improve their favorite tactics from all the games played recently.

Human can do both of the above tasks, but it is much more efficient to have the machines to automate these tasks, and have chess players access the information needed with a click of a button.

Image Recognition

Historically, chess students learn history or famous players by reading thru books. In today’s world, books are still more organized than online platforms.

What if an image recognition software can take a picture of a chess diagram, and immediately sync-up with any chess database, then we can merge any information from the physical world into the web sphere.

Essentially we want to build a chess encyclopedia crowdsourced by chess fans.

Self-Learning from Computers

DecodeChess is helping chess players translate Stockfish evaluations. What I’ve been thinking ahead is to have AI learn from historical famous chess players, and build a program to imitate that particular player.

Once the program is activated, chess fans around the world can choose their own favorite player, and have that program imitate the style.

We are in an exciting time to help chess gain popularity thru technology.

Chess players with tech interests can gain more by leveraging the technology advancement to improve their own games.

When you need a break from studying a complicated rook-endgame, take a walk outside, think about how AI and technology can help your chess and popularize our favorite game to more interested players!