Calculation Techniques: Discovered Checks

The following position is from a critical last round game in a recent Oregon scholastic tournament:

White played 1. Ne5+. This move checks the black king and also creates a discovered attack against black's queen. Black played the only move to get out of check and defend the queen, 1...Ke6. White certainly predicted this move, and probably became excited at the prospect of winning, since black's king is moving into a line with white's rook. White played 2. Ng4+!, a discovered check from the rook. This move wins a queen since white's knight is immune from capture and any black king move hangs the queen to Qxf5. Except that it doesn't! Black played 2...Qxe4+ and white resigned. Oops.

Hopefully you weren't tricked by my flawed analysis! Regardless, I think it illustrates what probably happened in the mind of the player with the white pieces that led to them blundering their rook with check! If you're thinking "I would never do that," consider that both players in this game were rated ~1950 USCF and it was a classical time control.

What happened above is a very common type of blunder in positions with discovered checks: we forget that the checking piece can be taken. I think this happens with discovered checks because we are used to being able to put the piece we are moving anywhere, even on heavily guarded squares. Consider the very first discovered check example most of us ever saw:

White puts the knight on the improbable c6 square where it can be captured 3 different ways. Of course it doesn't matter, since it's check from the white queen. Seeing so many examples like this causes us to feel a kind of invulnerability when we have discovered check, and to focus too much on the piece we're moving. I believe this because in the student/scholastic games I look at, it's incredibly common to blunder the checking piece, as in the first example. This is just one way that discovered checks mess with many players' intuitions.

Below are a couple calculation tips and techniques to help with discovered checks. Recall the most basic calculation technique: to look at sequences of forcing moves, like checks and captures. In most discovered check positions, it's not really possible to look at every sequence of checks, since there are usually a lot of them. Instead, you can use the tips and techniques below, which focus on which checks are typically the most important, and how to deduce which check is the strongest without having to "brute force" calculate all of them. I think the information here is probably useful for most players in the UCSF 1300-1700 range. There will be a quiz at the end where you can practice the techniques in some puzzles.

Calculation Tip: Look at the Double Checks first

In the following position, which I found in the lichess puzzles, there are 9 possible discovered checks. Trying to calculate and compare all of them, the way I'm sure Stockfish does, would quickly become overwhelming. But thankfully we don't have to! Generally good advice when calculating is to look at the most forcing moves first. In discovered check situations, double checks are the most forcing, since the opponent must move their king, so we should look at those first. Black to play:

This pattern is very common in the lichess Mate in X puzzles. In the classic book How to Beat Your Dad at Chess by Murray Chandler this pattern is called Pillsbury's Mate. If we don't recognize the pattern, we can find the mate very quickly by looking at the double checks first. Starting with 1...Rxh2++ is clearly a dead end after 2. Kxh2 since there are no follow up checks. Instead, the other double check 1...Rg1++ is better, after 2. Kxg1 Rg8+ is practically mate - white can only delay the inevitable by sacrificing both bishops.

Shockingly, this puzzle is rated 2191 at the time of writing! Perhaps even more shocking is that in the lichess rapid game it's from, the black player, rated 2040, did not find this mate even with 11 minutes on their clock. If they had followed the advice and looked at the double checks first, they would have found the mate in just a few seconds.

Calculation Technique: Using a "Blank" Move

In discovered check positions you may have a bunch of similar looking checks, and you need to figure out which one is the strongest, or if it even matters at all. When I'm calculating a position like this, I often say to myself "I am going to play one of the checks," but I don't mentally commit to a single one. I think of this as a "blank" move, as if I'm leaving a blank space in my variation that I'm going to come back to and fill in later. I try to figure out what the opponent will do and often find I can actually calculate several moves ahead without the precise placement of the initial discovered check mattering much. At some point in the variations you calculate after the blank move, it may suddenly become obvious which discovered check is best. Then you can go back and "fill in the blank!"

I don't know the real name of this technique, or if there is one. I think most really strong players just do this intuitively and probably don't consider it a "technique" worth describing - but maybe they haven't seen as many low-rated students struggle calculating discovered checks as I have!

Part One: The Blank Move stops a Defensive Resource

Consider the follow simplified example, white to play:

White has 7 bishops moves and all of them are check. Most of the moves are along the b1-h7 diagonal. The one radically different check is 1. Bg8+, which seems silly as black can capture the bishop with the king.

What about the other checks, is there an important difference between them? In this example, if black responds to any of my bishop checks with 1...Kg8 then I can play 2. Qh7#. But I notice that black can also respond to my generic check with 1...Qh6!, which would prevent 2. Qh7# Once I have seen this defensive resource, I can consider whether or not any of my starting checks address it in some way. Here the only move that leads to mate is of course 1. Bg6+!, putting the bishop in the way of black's queen so that it cannot come to h6.

This was quite a simple example, and many readers probably saw the answer immediately by simply recognizing the common pattern. But hopefully it clearly demonstrated the thinking technique, how it works, and how it can be useful!

The generic, blank move can serve many purposes. In the above example, we had to recognize the opponent's defensive resource and use the blank move to stop it. You can also use the blank move to think about where your piece wants to go next.

Part Two: The Blank Move gets a piece to a useful position

Here's another type of position where the blank move technique is helpful. White to play and win:

We have the same seven discovered checks as in the last example! Once again, 1. Bg8+ allows 1...Kxg8, so realistically we have six discovered checks from b1 to g6 to consider. There is no way to make any meaningful observation about the starting position (without calculating) to determine which check is strongest, as they all have certain potential advantages, such as targeting the f7 pawn, the e6 knight, or the a2 pawn. Instead let's calculate using a blank starting move.

On move one we play one of the discovered checks along the b1-h7 diagonal. Black must play 1...Kg8. We can follow a sequence of obvious forcing moves: 2. Qh7+ Kf8 3. Qh8+ Ke7 4. Nf5+ Kd7. Now there is no check from our knight and our queen checks on c8 d8 and e8 are all pointless, dead-end sacrifices. This is a good time to return and try to "fill in the blank" first move. In the final position we visualized, where do we want our bishop to be? We could try to get our first move to set up a bishop check once black's king is on d7. This line of reasoning leads us to the answer. With the king on d7 we could set up a bishop check on c6, b5, or a4. Bc6+ is no good since it's unguarded and black can capture it with the king. Ba4+ would get captured by black's rook. But b5 would be a safe check, and in fact it would be mate. So our first move needs to be 1. Bd3+. Then the rest of the variations goes 1...Kg8 2. Qh7+ Kf8 3. Qh8+ Ke7 4. Nf5+ Kd7 5. Bb5#.

I believe for most strong players, thinking this way is intuitive. They are used to making adjustments to variations as they calculate, in this example reaching the position with black's king on d7 and then realizing that the starting check may make a huge difference. But I've seen a lot of students struggle in these positions by getting "locked in" on a specific variation without realizing they can calculate and look ahead while maintaining some mental flexibility about certain moves.

Final Thought: the "Windmill" and how it can cause you to BLUNDER!

An important pattern involving discovered checks is the tactic called a Windmill. The most famous example of this tactic is probably from the game Torre - Lasker, Moscow 1925:

Torre sacrifices his queen in order to set up the windmill - the white rook going back and forth, alternating between a check on g7, which forces the black king to h8, and a discovered check. Torre uses these discovered checks to take two of black's pieces on the 7th rank before returning to capture black's queen. Note that after the moves shown above, black can play 32...Kg6 and regain some of the lost material. After 33. Rh3 Kxf6 34. Rxh6+, which occurred in the game, Torre was up three pawns and went on to win.

The windmill is a cool, flashy tactic, and that makes it easy to remember. I still remember getting excited seeing it for the first time in How to Beat Your Dad at Chess (where it's called a "see saw"). But it's important to recognize that this pattern is very rare. For it to work, lots of pieces need to be in very specific positions. In the Torre - Lasker game, the black knight on f8 is essential. If f8 was occupied by any other piece, or was empty, then the tactic wouldn't work. You can figure out why. Another important thing to realize is that recognizing a tactical pattern is often the last thing that happens right before you blunder. Using your brain to slowly, deliberately calculate and visualize chess moves is hard work. But recognizing a pattern is effortless. And when we recognize a familiar pattern, we are tempted to just play our move right away, without doing any of the hard work, which is usually required to appreciate any hidden or subtle details in the position. Our brains want chess to be easy even though we know it's not! Let's revisit this position from earlier:

In the game, black played the move ...Rxf2+ instantly, missing a very easy mate in 4, despite being rated 2191 and having 11 minutes on their clock. Black probably recognized the windmill pattern and assumed the game would go Kg1 Rg2+ Kh1 Rxe2+ Kg1 Rg2+ Kh1 Rxd2+ etc. Instead after ...Rxf2+ white replied with the best move Bf3, remained comfortably up a piece, and won in a few moves. If black plays ...Rxd2, white plays Bxb7. If black plays ...Bxf3+, Nxf3 saves the hanging d2 knight. This wasn't an example of blundering the checking piece, but it was similar: black blundered because they missed a clever way for white to get out of the starting check and disrupt the whole pattern before it could begin.

Many of the lichess discovered check puzzles have, in my opinion, very inflated ratings. And the games they come from show that these tactics are often missed by very strong players! I think this is because of our natural desire to apply pre-existing patterns like the windmill to these positions, rather than to calculate them carefully and deliberately.

Quiz

Below are 14 puzzles involving discovered checks. I tried to create many different tactical ideas from the same basic type of starting position. All puzzles are white to play and win. The next section contains the puzzle solutions. Here are a few final tips for solving:

• The puzzles start easier and get harder.
• Look at the double checks first. Try to set up double checks.
• Notice if the checking piece (bishop on b2/c3) can get captured.
• Some positions lead to mate, others lead to winning material, others lead to forced draws.
• If you see a way to win material, confirm that white has actually reached a winning position at the end. In many of the puzzles white begins down a queen or more.
• Some of the positions are very similar but have very different answers. You'll need to figure out how the small differences change both sides' options!

Puzzle 1:

Puzzle 2:

Puzzle 3:

Puzzle 4:

Puzzle 5:

Puzzle 6:

Puzzle 7:

Puzzle 8:

Puzzle 9:

Puzzle 10:

Puzzle 11:

Puzzle 12:

Puzzle 13:

Puzzle 14:

Quiz Solutions!

Puzzle 1:

Puzzle 2:

Puzzle 3:

Puzzle 4:

Puzzle 5:

Puzzle 6:

Puzzle 7:

Puzzle 8:

Puzzle 9:

Puzzle 10:

Puzzle 11:

Puzzle 12:

Puzzle 13:

Puzzle 14: