Luck or Skill?

by admin on January 6, 2015

It’s an age-old question: is chess truly just a game of skill, or is there luck involved too?

Here is perhaps the most crucial position I faced in the New Year Championship tournament last weekend. It’s round five, and I’m playing back against Ivan Ke (the #29 11-year-old in the country). I got a good position out of the opening, but after a couple of mistakes I’m in a bit of trouble.

ke1Position after 30. Qxb5. Black to play.

FEN: 3r2k1/2q2pp1/1b2pnn1/1Q5p/1P2P3/6PP/1B1N1PK1/1B1R4 b – – 0 30

Ke has just won a pawn on b5, but his kingside is looking rather deserted. How can Black take advantage of this?

Well, here is where either luck or skill, but definitely some creativity, entered the picture. I was in time trouble, with about 9 minutes left to make 11 moves. I had been expecting him to take the pawn, and I didn’t even think twice. I slammed down the bishop sacrifice: 30. … Bxf2?!?

What was my thinking? Well, after 31. Kxf2 h4! I simply didn’t see any clear defense for him. The pin on the d-file is a real killer for him. He can’t defend g3 with either the rook or the knight because he loses material; the bishops are so far away that they’re useless; and the queen also can’t defend g3. So no matter what, it looks as if White’s king is going on an adventure, and I have lots of pieces (queen, rook, and two knights) buzzing around him. My intuition said that Black must at least have a perpetual, with very good practical chances for a win.

Alas, there was one flaw in my thinking! The White queen can defend g3, by playing 32. Qg5! It’s perhaps understandable that I missed this, being in such time trouble. But what’s a little bit more surprising is that my opponent, who had oodles and oodles of time — probably an hour to make 10 moves — didn’t see it either. Perhaps the problem is psychology. If you’re thinking about “defending” g3, you’re thinking about retreating your queen, to e2 or f1 or something. The move 32. Qg5 is psychologically hard to see because it’s a defensive move that isn’t a retreat. At least that’s my theory.

However, there is another possibility, which is that Ke saw the move but simply liked 32. e5? better. Because that’s what he played. In his defense, the computer almost agrees with him. For the first several minutes, Rybka thinks this is almost as good as 32. Qg5. It really has to go deep into the position before it concedes that the position after 32. e5 is equal.

So was I lucky or good? My answer is, “Neither.” Chess is about creativity, and about posing difficult challenges for your opponent. There’s no question that the position is tremendously difficult for both sides. My only luck, I think, is that I was in time trouble and didn’t see 32. Qg5. Because if I had seen that I almost certainly would not have sacrificed the bishop!

Anyway, let’s continue. Ke played 32. e5? and I replied 32. … hg+ 33. Kxg3 (walking into a pin, but declining the pawn turns out to be no better — it really just gives Black another attacker) 33. … Rd5! After this move I really felt optimistic; all my pieces are coming into play and there is no relief in sight for his king.

Here Ke came up with a pretty good move: 34. Rc1. The computer considers it a mistake, but from my perspective it was very clever, unpinning the knight and gaining a tempo by attacking my queen. But before we look at Ke’s move, let’s look at the move that the computer likes better: 34. Qa6!

ke2Position after 34. Qa6 (analysis). Black to play.

FEN: 6k1/2q2pp1/Q3pnn1/3rP3/1P6/6KP/1B1N4/1B1R4 b – – 0 34

What should Black do here? I can tell you what I would have done, because I was already thinking about it: 34. … Nxe5? I mean, who can resist? Winning a pawn, ripping open the diagonal leading to White’s king. But White has a miraculous defense here: 35. Qa8+! Rd8 (Forced! Black’s king doesn’t have a flight square on h7 any more!) 36. Qa5!! Stunningly, Black’s attack is over. The discovered checks are no good because the queen on c7 is hanging. There is no time to move the queen because the knight on e5 is hanging. The only thing to do is acquiesce to the queen trade with 36. … Qxa5 37. ba, but this has to be winning for White.

Instead the correct move order for Black is 34. … Nh5+!, and the “high-level abacus” says this position is equal. I would certainly rather be playing Black, though. I’ll let you have fun analyzing it.

Instead, as I said, Ke chose to play 34. Rc1 Nh5+ 35. Kf3, and now we get to the last puzzle for the day: What should Black do?

ke3Position after 35. Kf3. Black to move.

FEN: 6k1/2q2pp1/4p1n1/1Q1rP2n/1P6/5K1P/1B1N4/1BR5 b – – 0 35

Here I played the natural move that probably any human would play, especially a human in time pressure: 35. … Qd8? But Rybka says this move gives the advantage back to White! Instead Black is close to winning if he plays 35. … Nh4+!! Of course I considered this, but I didn’t really see the point. Why am I putting all of these knights on the rim? What if White just plays 36. Kg4? Well, the point of Rybka’s scheme is to lure White’s king to a square where Black’s queen can give check. That will free his rook to capture White’s queen. It requires sacrificing both knights to accomplish this goal. Here’s how. After 36. Kg4 f5+! 37. Kxh5 (note that 36. ef? leads to instant death: 36. … Qg3 mate!) 37. … g6+ (a little extra flourish that Rybka likes; 37. … Qf7+ is also fine) 38. Kxh4 Qd8+ and Black is winning. Technically, White has three pieces for the queen, but the pieces are disorganized, everything in White’s camp is hanging and his king is still exposed, so there’s no doubt of Black’s huge advantage.

Of course 35. Kg4 isn’t forced for White, and I suspect it’s more likely that White would have played 35. Ke2 or 35. Ke3. But the computer says that Black is better in those lines, too. I will leave them to you as an analytical challenge.

After 35. … Qd8? he played 36. Qa6!, which Rybka approves of and gives White a slight advantage. I’ll stop here. Obviously the position is a wild mess, and further mistakes were made by both sides — me, because I had so little time left, and Ke, because the position was so darned hard for him. Even during the game I was sure that the computer would find all sorts of mistakes, and I was right.

Although we should always strive to analyze more accurately, I don’t think it’s worth getting too upset about the errors by both sides. It’s true that we both should have seen 32. Qg5. But allowing for that one blind spot, I think that sacrificing the bishop was absolutely the right decision. It scrambled the position and made the position vastly more difficult for my opponent than for me.

Once again, I say: Was it luck or skill? It was neither. It was creativity. That is the reason that I am proud of this game.

Print Friendly, PDF & Email

{ 5 comments… read them below or add one }

Matt January 6, 2015 at 10:12 am

An intuitive sacrifice is always a combination of luck and skill. If you can’t see a clear win after the sacrifice (usually because to do so would require calculating too many moves ahead) but can at least see compensation or the potential in the position, then it took skill to find the sac and then you trust to your experience, power of your pieces and, yes, a little luck, in order to justify it. Bxf2 was the first move I looked at in the position and it looked tempting. It’s hard to say whether it was “correct” because, on the one hand you felt the position was turning sour for you and you had to make something happen quickly. On the other hand, you mentioned being short on time and it’s rarely a good idea to radically transform the position when in time trouble. There is a good chance you will have missed something (not having had enough time to calculate all the variations) and your opponent is going to have more time than you to figure his way out of the ensuing mess.

Still, I think Bxf2 was a very nice sac. I expect Tal would have played it. 🙂


admin January 6, 2015 at 11:09 am

Nice to think that I might have played like Tal! One virtue of a move like Bxf2 is the “bolt-from-the-blue” effect. I’ve mentioned that White had lots of time, but sometimes it doesn’t matter how much time you have. If your opponent plays a move like … Bxf2, there’s a tendency to say “Oh my god, I didn’t think of that,” and even if you use 5 or 10 minutes, you never get back into a calm, objective frame of mind.

A couple things I didn’t mention in the original post, for reasons of space. Even after 32. Qg5 the game is far from over. After 32. … hg+ 33. Qxg3 Qb6+ Black does win the the b-pawn, and it will be hard for White to actually win even though he’s up a piece for a pawn. All the pawns are on one side, and his king is still much more insecure than Black’s.

The other thing that surprised me in Rybka’s analysis was that in the original position, it thinks that 30. … h4! gives Black full compensation for a pawn. I didn’t think that was so strong, because it’s trying to batter down the front gate of White’s castle, while 30. … Bxf2 goes around and attacks the castle from behind. But Black can shift his threats among three diagonals, and it’s hard for White to keep them all covered. If he brings his queen back to defend the kingside, the b-pawn often falls and then we’re back to even material.


Mike Splane January 7, 2015 at 7:23 am

Hello Dana. Hope you won’t mind if I offer some constructive criticism.

“My only luck, I think, is that I was in time trouble and didn’t see 32. Qg5. ”

I don’t think luck has anything to do with you being in time trouble and missing a tactic.. I’ve seen similar comments in your blogs on many occasions. There is something wrong with your thinking processes that causes you to be in time trouble far too often. If you can figure out what that is and fix it you’ll be a much stronger player.

Maybe you’re over-relying on calculation.

Here are some questions to consider using that could speed up your choices of candidate moves. Of course after you answer the questions you still have to verify your conclusions with calculation.

The Mike Splane question: How am I going to win this game?
The Gjon Feinstein question: What are the targets?
What does my opponent want to do?
Which side of the board should I be playing on?
Who is better in the ending?
Whose pieces are better; which pieces are active (good) and which are inactive (bad)
How does the human element affect the game? ( time pressure, relative playing strength, age, experience, emotional state, fatigue, hunger, etc.)

I like the computer’s recommendation of 30… h4. Answering the Gjon Feinstein question tells us that there are targets are g3, f2, and d2 and potential outposts for your pieces on e3 if the f2 pawn moves and on f4 if the g3 pawn moves. Eliminating one of the dark square defenders really improves your chances. It also takes away a target of the White queen.

Maybe we can have a chat offline about this particular position.


admin January 7, 2015 at 9:39 am

I never mind constructive criticism! Indeed, calm after-the-fact reflection suggests that there was a lot of room for improvement in my play last weekend, even though my result was good. I only managed my clock well in one game out of five. I actually reviewed your above questions before the tournament, so I don’t think that is the problem. I do think, though, that you put your finger on one of the two main causes of the time pressure. The cause you correctly identified is that I am still trying to compute when I should be thinking schematically. (Ironically, in the above game that left me with NO time to compute when I really did need to compute, namely when I played the bishop sac!)

The second cause, which you couldn’t know about because you can’t read my mind, is lack of confidence. Early in the game I might think about a line three times before I finally decide to play it. That’s not efficient or good. (Kotov in “Think Like a Grandmaster” says to look at each line once.) Later in the game I’d LIKE to triple-check but I don’t have time any more! Sometimes that strangely works in my favor (that’s one point of my provocative statement that I was “lucky to be in time trouble”). In particular games time trouble can help by forcing me to reach a decision faster, and even by rattling my opponent. However, it is of course essential to realize (and I do realize) that over the long run time trouble can only be detrimental.


Matt January 7, 2015 at 11:06 am

Time trouble is a big problem for me personally. I estimate that at least 75% of my losses are due to time trouble and that in at least half of those I had winning positions at one point and then spoiled them by having to move too quickly. It’s been an issue for me for a number of years now, one that I am gradually trying to correct but it’s not easy. I marvel at the higher rated kids at my chess club who are able to rattle off highly accurate moves in seconds. True, some of the time that’s because they are in book but, often, I think it is because they are so used to playing blitz that they have become accustomed to playing quickly.


Leave a Comment

Previous post:

Next post: