Recently a video went viral online of Bill Gates losing a blitz chess game in nine moves to the new World Chess Champion Magnus Carlsen, getting checkmated quickly:
The truth is deeper than it seems
Non-chess-playing friends of mine enjoyed forwarding me links to this. Their general impression was, ha, of course the nine-move checkmate makes sense given the gross mismatch in strength between Bill Gates and the world champion! You can also look at the YouTube video’s comments to get an idea of what a lot of people thought of the video.
I watched the game and noticed some very interesting features, on multiple levels, and came up with the idea of “9 lessons for 9 moves”, because it is hard to get an accurate picture of what serious chess is like through typical media portrayal that shows moves being made, and maybe plays up drama, but does not explain what happened, or what could have happened instead, or what players were thinking: all this is left invisible. This Gates-Carlsen encounter is no exception. The comments from non-chess-playing friends made me feel the need to give a deeper picture. The main misconception that bothered me was that a world champion should be able to win a game against someone in nine moves. But it’s not so simple. In fact, I found amazing depths in the truth of what happened.
This article will actually be directed at three different audiences, simultaneously, through a separation of levels of observation:
- Those who do not play chess or vaguely know the rules but do not play regularly
- Those who do play regularly and have chess tournament experience, but are below an expert level
- Those who are at expert or master level
The entire game, along with supporting variations, is embedded here so that you can click through what happened as well as what could have happened but didn’t. My detailed explanations follow after this embedded interactive board.
Lesson 1: Why Carlsen played a bad chess opening
The first observation to make is that Bill Gates probably knows how to play a basic chess opening as White. Very popular, for example, is to play 1 e4, as Gates in fact did (the late Bobby Fischer’s favorite), and then if Black symmetrically plays 1…e5, White plays 2 Nf3 attacking Black’s Pawn, then Black usually defends with 2…Nc6, and then White plays 3 Bb5 attacking that Knight. This would be the famous Spanish Opening, also known as the “Ruy Lopez”, after the Spanish priest of the 16th century (yes, chess wisdom goes way back!) who made a systematic study of this opening.
Carlsen plays both colors of this opening: in the recent World Championship match against Anand, for example, he played the Black side repeatedly with success in three games, drawing two and winning one.
But in this game against Gates, he did not play this opening. Instead, he played an opening that is usually considered unusual and not so good, the Nimzovich Defense. Why?
The short answer: because by playing the best moves, Carlsen would have risked taking much longer to win a game against Gates. If you are a seasoned chess player, especially if you play blitz, you understand this. If you are not a chess player, or do not know much beyond the rules, this may puzzle you: how could playing better lead to winning much more slowly? The answer is that games are not won so much as lost.
Here’s an analogy. Suppose you find yourself being attacked by someone and you need to defend yourself. Do you put up your best blocks, and attack back with proper form, making sure to stay safe? Or do you do one of those risky roundhouse kicks you see in the movies, where you could get in a lot of trouble if you failed? Doing the “right” thing does not make for entertaining and short action movie scenes. So that’s what happened in this game: Carlsen gambled that Gates would not know how to play properly, and played an opening with an intention of possibly unleashing that roundhouse kick.
Lesson 2: Chess is psychological warfare
Gates seemed confused by the Nimzovich Defense. On move 2, he makes a perfectly good move, but the move Carlsen clearly had wanted to see. Gates was on autopilot, playing the Knight out to f3 as though Carlsen had played 1…e5 instead. There were other options. 2 d4 would also have allowed Carlsen to continue his game plan of advancing with 2…d5, but 2 Nc3 would have stopped that plan and made the game probably much, much longer.
Contrary to the popular misconception of chess as pure logic, it is actually primarily a psychological sport, in which knowing how the opponent might react is a valuable part of being competitive. This is true at the highest levels as well as in casual blitz chess. Carlsen won his world title last year largely on the basis of thwarting Anand psychologically, refusing to play “good” openings that he knew Anand would be prepared for, and playing bland openings instead. Here, in a flashy game against Gates, he took the opposite approach, deliberately playing a “bad” opening that expected Gates not to be prepared for.
Lesson 3: Refusing to play the best move because it may win too slowly
Gates played a poor move 3, being defensive in a position in which he could have seized an advantage aggressively by capturing Black’s Pawn on d5.
However, Carlsen refused to punish this poor move. The obvious punishing move, that most club players would see and play, is 3…Nb4, attacking the White Bishop, and in fact winning it in exchange for the Knight, because if the Bishop moves, then White loses the Pawn on e4 that the Bishop is defending.
Again, this goes back to psychology. If Carlsen had played the obvious and best move, the game could have taken much longer to win. It would have been a matter of gaining a permanent advantage by winning the Bishop pair and ruining White’s Pawn structure on the d file, and then pressing further until victory.
If you look back at the video, you can see Carlsen hesitating before playing the “worse” move. In a real game against a real opponent, there is no doubt he would have instantly whipped out the strong 3…Nb4, no hesitation.
Lesson 4: The appeal of open vs. closed games
The gamble worked, because Gates took the Pawn on d5 on move 4, and Carlsen brought out his Queen, as he had intended from the very moment he played his very first opening move!
Note that Gates could have played e5 closing the position, to reach a perfectly decent game for White. In fact, I believe many club players would have instinctively played e5, in analogy with fundamental positions in the French Defense. Click through the embedded game board for a logical continuation of the game if Gates had played e5. Note that after White gets in d4 and Black gets in e6, the resulting French Defense setup is very closed and it will take a very long time for Black to win the game.
Closed games may not be as flashy and media-friendly as open games, but they are very strategically rich and beautiful to play and watch. It would have been interesting if Gates had gone into this French Defense setup and Carlsen had to patiently attack White’s center with f6, get the half-open f file, swing the dark Bishop against White’s King side, free the light-squared Bishop also, and bring the Queen around to the King side via Qe8, etc. There might even have been a thematic exchange sacrifice on f3 to break up White’s King side.
But this would not have been a nine-move media spectacle.
Lesson 5: The unexpected complexity of chess
On move 5, Gates correctly attacked Carlsen’s Queen with his Knight, and Carlsen played the Queen over to the King side. Although a thematic plan in this opening, under circumstances where Black has already played the light-squared Bishop out to g4 and is preparing to castle Queen side, this move actually looked to me like a blunder! But upon looking at it more, with the help of the computer, I saw that it was not.
(The following discussion is mainly for those of expert or master strength.)
The reason Qh5 looked like a blunder was because it seemed that White could play the very strong-looking Nb5, threatening to win the Pawn on c7 with check and also pick up Black’s Rook. Black could defend by Kd8, but that forfeits castling and guarantees White an advantage. Nd5 to defend instead looks possible, but then c4 chasing the Knight looks strong. After Black’s Nf4, White retreats the Bishop to defend, but renews the threat. However, after Black plays Bg4, the position then becomes a bit crazy, with Black having huge threats against White’s Knight on f3 and offering to give up both castling and the Rook on a8 for a big attack against White. After White plays d4, attacking Black’s Knight on f4, things look bad for Black.
It turns out that Black can castle Queen side, abandoning the Knight but then creating threats with e5 in order to regain the piece. So far, so good: an expert or master player could see all this because of the forced nature of the calculations.
But when I ran the position through a chess engine, the computer spit out a completely different approach, having Black not castle Queen side, but instead play e5 immediately, still offering to give up the the Pawn on c7, castling rights, and the Rook on a8. I did not see this possibility. Nor have I analyzed the full ramifications. All I know is that yet again, chess engines reveal the surprising complexity of chess, one that the top world human players grapple with regularly. As an amateur, I am still repeatedly surprised by how intricate chess positions can be. If you are curious, you may wish to work out the possible continuations after the crazy-looking 9…e5.
It would have been fascinating if Gates had seen Nb5 and played it, forcing Carlsen to enter into these wild complications.
Lesson 6: Playing a good move can lead to losing quickly, but it’s not the fault of the good move
On move 6, Gates castled King side. This was a good move. A general principle that chess novices often neglect is King safety. Many games in chess are lost because one side did not protect the King by castling in time. Beginners are taught to castle as early as possible, as a rule of thumb. It is usually a great principle to follow, and it was in this game.
Yet, it was because Gates castled King side that he lost so quickly. Paradox? Yes. What happened was that he lost because of things that he did wrong after castling. If he had played a worse plan, bringing the Bishop back to e2, then developing the Queen side with d3, and castling Queen side, the game would have been considerably longer for sure.
Sometimes it is tempting for a less experienced chess player to attribute a loss of a game to a move that was actually a good move, instead of noticing the true blunder that happened well after the good move. In particular, it would be very tempting to reason, “Black’s Queen is already aiming at my King side, so I shouldn’t have castled there into the attack.”
This happens in real life all the time: it is the logical fallacy of post hoc, propter hoc, or “after this, therefore because of this”. For example, someone might hear about a runner dying of a heart attack and wrongly conclude that running is bad for your health, or someone else might hear about a runner losing weight and recovering from diabetes and wrongly conclude that running lots of marathons is the key to health.
Chess is a game in which, in the long run, you cannot succeed if you fall into the trap of logical fallacies: all psychology aside, it is still a game of logic, in which if someone plays a trick on you, you can learn the truth from it for the next time you face that same opponent or a different one, and perform better. That is the beauty of the game.
Lesson 7: The appalling practice of playing a terrible move in hope that your opponent doesn’t see how bad it is
On move 7, Gates played h3, attacking Black’s Bishop on g4. The obvious and good move is to exchange the attacked Bishop with White’s Knight on f3, trading Queens after White retakes with the Queen, and wrecking White’s King side Pawn structure to obtain a clear advantage. The average club player would play this continuation without hesitation since it is obviously good and best.
Unfortunately, with the Queens off the board, it would then have taken probably another thirty moves for Black to win the game.
Instead, Carlsen played a shockingly bad, losing move: he ignored the threat to his Bishop on g4 and simply offered a sacrifice, while making an unjustified aggressive gesture with his Knight to d5 against White’s pinned Knight on f3.
I was not happy to see him do this, because I believe many people believed this was an authentically and well-played chess game, rather than media fluff, and because many chess players actually deliberately engage in this practice in serious games, and I would like to discourage it. I wrote this article largely because of this howlingly bad move. I wrote this article for my friends who were wowed by this “daring” attack but need to know the truth.
The truth is that this was not a serious game in which Carlsen played well. It was one of those games you might see played in urban parks by hustlers (I have played with such hustlers before; they typically let you win one and then try to get you to play them for more money and then start playing more and more for real). Carlsen played like a hustler in this game. Granted, it’s fun to do this sometimes, as long as everyone is in on the joke and knows what is going on. But this game has been forwarded around with no disclaimer that it was for entertainment purposes only.
Lesson 8: The right plan is more important than the right move; but what about the importance of failure?
Gates accepted the sacrifice. Again, it is important to emphasize that even though he got checkmated just one move later, it wasn’t because he was foolhardy in doing the right thing by accepting the sacrifice to achieve a winning position.
He played the right move. But he didn’t play the right followup move. He clearly had not calculated the implications of his move, however right the move was. If you do the right thing, but without really understanding why it was right, you can get into trouble. This is as true in chess as it is in regular life. It is important to make a move not in isolation, but in full context of what might happen afterwards.
For example, if you are stuck in the wilderness, and you know which direction the main road is, the right thing to do may well be to start moving in that direction. But this assumes that you have already seen that you are not walking toward a deep rift, or that you have seen a barrier but know you have the resources to overcome it. If you have not at least figured out some kind of plan, it may be safer not to head to the main road, but to head in a different direction toward terrain that looks less unpredictable, even if the route will take longer to traverse.
Should Gates have taken another path, instead of accepting the sacrifice? This is a hard question to answer. An improving player should always at least experiment with trying to play the best move possible, even if not perfectly equipped for the followup; you learn by failing while doing the right thing. But as a practical matter, there are good reasons to argue that failing too drastically may not help learning, and that playing more safely is a more incremental way of learning. I think it’s a hard judgment call sometimes.
For example, here, Gates could have done some meta-thinking, “Hmm, Carlsen wouldn’t be offering this free piece to me if he didn’t have some trick to play on me, so I should play it safe and decline to accept it, and instead play safely with 8 Be4 to simply defend my attacked Knight on f3.” If he had in fact done that, Carlsen’s trick would have been halted, because there are no more pieces he can bring in against the Knight on f3. Again, objectively the best for Black would be to trade everything down on f3, giving White doubled, isolated f Pawns, but this is the route to a very long game.
Unfortunately, I see this kind of meta-reasoning happen too often in club level chess, where a weaker player is afraid that the stronger opponent must have had a reason when playing a seemingly bad move, and therefore does not play the correct continuation against it. I think there is a fuzzy line between being pragmatic about one’s limitations and being too timid to dare to question a poor move and try to prove it wrong. This psychological tension is definitely an important part of the game, and can be a vehicle for developing self-awareness, trust in one’s own mind (rather than trust in the authority of those better than you), and continuous learning in case of making the wrong call.
I advocate erring on the side of learning from one’s mistakes by going out on a limb at least sometimes.
Lesson 9: Carlsen was really lost; life and death can depend on only one detail
In the final move of the game, Gates played a move that would be very good except for one flaw: it allowed instant checkmate.
But we cannot entirely blame him. Ignoring the checkmate oversight, the move was actually a fantastic one. White takes the Knight on e5:
- If Black recaptures with the Knight, White wins the Queen on h5.
- If Black recaptures with the Queen, White capture the Knight on g4, left unguarded because of the deflection of the Queen that had protected it on h5.
So it looks like a brilliant move, forcing the win of a piece. With Black two pieces down, victory would be in sight for White.
Unfortunately, there’s that checkmate.
In chess, you can do everything right, have everything figured out but one detail, and lose horribly. In fact, many high-level games actually go this way. What are analogies in real life? There are many.
In warfare, missing one vital piece of information could mean losing an entire battle. The Challenger tragedy. Mission-critical computer software. Professions such as surgery, race car driving hinge on detail. Competitive sprinting and hurdling.
Carlsen was really lost
Gates could have maintained a winning advantage by playing 9 Re1. This provides an escape route for his King in case Black came down to h2 with the Queen by first trading off White’s defensive Knight on f3. White’s King can escape to f1, e2, and then d1 as needed. Retaining a piece advantage, objectively White should win the game.
Of course, in practice, even if Gates had played correctly, I doubt he would have won. Carlsen would probably have castled Queen side and continued playing tricky moves while waiting for blunders, likely because of the blitz time format. It would have definitely been entertaining hustler-style blitz chess by Carlsen in any case.
For the benefit of those who want to see how quickly White can win, after Re1, if Black insists on trying to check White’s King to oblivion (which Carlsen would not have done), I have provided a sample forcing continuation by White.
That was an entertaining game.
But I felt I needed to point out to the wider public that in fact this was the diametric opposite of the brutally logical play that won him the World Champion title just months ago! He won here in nine moves not because he played the logical, correct moves, but because he resorted to every element of hustler psychology to try to create a quick decisive outcome. I hope that this article helps explain what happened.