lichess.org
Donate

Calculation Techniques: Switching the Move Order

ChessTacticsAnalysis
This post is about the calculation technique of "switching the move order" when you spot a tactical idea.

In my opinion switching the move order is the single most important calculation technique to remember (beyond the basic skill of "looking at forcing moves"). I use this technique multiple times in every tournament game I play. In the chess lessons I give I am amazed at how often reminding students of this technique causes them to immediately solve a puzzle they haven't quite gotten.

The Technique

Often while playing a game or solving a puzzle you will find a tactical idea or variation that doesn't quite work. You should ask yourself what happens if you play the same moves but in a different order. This is the basic concept of switching the move order.

Like many calculation techniques, switching the move order is often done by the slow, deliberate, logical part of your brain, especially when you are first learning to do it. It's a way to get the thinking part of your brain engaged with a position when your intuition isn't getting the job done, or in other words, when you don't immediately "see" a winning tactic. Perhaps I look at a position, I intuitively see a tactical idea, I see that it doesn't work, then I remind myself "I am supposed to look at different move orders," and I force my brain to process the moves in different orders, even if they seem pointless at first.

Most master-level players have practiced this technique a lot and it has become a natural part of their thinking process. I think I use it on essentially every tactical idea I see during a slow game. Even if the tactic seems to simply be winning, I often find myself looking at different move orders, because maybe one of them is even more easily winning!

Basic Examples

The following basic examples all contain a winning tactic. You can read the first paragraph below each position, in which I discuss the right idea in the wrong move order, and then try to "fix" the solution by changing the move order. If you do this it will be somewhat similar to playing a game, seeing the correct idea, then needing to explore different move orders to figure out which one works.

Black to play and win:

https://lichess.org/study/GzhGMO8d/QlYBAjXf

Let's say I look at the move 1...Qa5+. This looks promising if I notice that my queen is supporting the move pawn to e5 next, forking white's bishop and knight! Unfortunately, this version of the tactic does not work at all, for many reasons! First of all, white could respond to ...Qa5+ with Bd2, attacking the queen and getting out of the fork. Qd2 would be similarly fine for white since black needs to deal with the threat to their own queen. Lastly, white can even fall for the tactic by meeting ...Qa5+ with c3. Then after ...e5 white has Nb3! saving one of the forked pieces and creating a counter-attack against black's queen, allowing them to save the bishop on the next turn.

So should I abandon my idea? No! I should ask what happens if I play the same moves in a different order. In this case, there were only two moves in my tactical idea, ...Qa5+ and ...e5, so switching the move order is easy. What if I start with 1...e5? This immediately attacks white's bishop and knight, white must play Bxe5 to avoid losing a piece. Then we go 2...Qa5+ and this is a basic fork of white's king and bishop. In this example I had the right idea but I initially put the moves in the wrong order.

Here is another example, it's white to play. We're up a bishop in an endgame, but we need to make sure black doesn't capture our last pawn. This position was part of the homework for one of my online classes and I was really shocked how many players rated around 1200 USCF got it wrong.

https://lichess.org/study/GzhGMO8d/mUFZ0xpx

Black is threatening to play ...Kg3 when it will be impossible for us to save our h3 pawn. This caused most of the kids in my class to play 1. Bf6, preparing to meet ...Kg3 with 2. h4 when the bishop defends the pawn and white will win easily by bringing up the king next. Try to figure out what is wrong with this order of moves!

In reality, starting with 1. Bf6 is a terrible blunder because black will play ...h4! Then, no matter what, black's next moves will be ...Kg2 and ...Kxh3. We simply don't have time to capture the h4 pawn with the bishop, get the bishop out of the way, and then play h4 ourselves.

The correct solution is of course the same two moves, just in the opposite order: 1. h4. Now white can safely meet ...Kg3 or ...Kg4 with 2. Bf6 and we've reached our desired defensive piece configuration.

Switching the move order is also relevant in the next example, albeit in a more abstract way. White to play and win:

https://lichess.org/study/GzhGMO8d/rRCXT2yf

Most of the same 1200-rated students I showed this position to quickly suggested the move 1. Qd5, threatening mate on f7 and also the undefended bishop on c5. Does this work? No! Black has only one defensive move but it's a good one: ...Qe7, simply defending against both threats and getting ready to kick back white's queen with ...Nf6 next.

In this example our original, flawed solution involved the white move Qd5 and then the possible follow up moves Qxc5, and Qxf7. There's no way to switch the move order of these since our queen begins the puzzle on d1. It cannot take on f7 or c5 first! However, we can think about what white was trying to do more generally, and try to switch the order of the ideas instead of the moves. Of course the specific move Qxf7 would be nice to play as it's checkmate, but more abstractly our idea involved a capture on f7. Which is in fact possible in the starting position.

So the correct solution is 1. Bxf7+!. If black plays ...Kxf7 only then do we place our queen on d5, forking the king on f7 and the bishop on c5. At the end of this white has won a pawn and black's king position has been weakened.

Examples From Mate in 4 Puzzles

Here are three examples I found in lichess Mate in 4 puzzles. Once again you can try to solve the puzzles on your own, or read the first paragraph below the puzzle to see the wrong order, then try to fix it.

https://lichess.org/study/GzhGMO8d/wv8XgEdR

This is a classic case of two possible move orders. For many players the most intuitive order is incorrect. At first it may look like we can easily back-rank mate black by playing 1. Nd7+ Ka8 2. Rc8+ Rxc8 3. Rxc8+ Bb8 4. Rxb8#. Unfortunately this move order allows a defense: 1. Nd7+ Ka8 2. Rc8+ Bb8! and black does not get mated, since the h8-rook defends the b8 bishop through white's c8 rook (an x-ray).

The correct solution involves the exact same four moves, just in a different order. 1. Rc8+! We start with this move so that black must capture our rook. 1...Rxc8 2. Nd7+ Ka8 3. Rxc8+ Bb8 4. Rxb8# For many people starting with 1. Rc8+ is a lot less intuitive as we are placing our rook on a square where it is attacked twice and only defended once, but of course this does not matter!

https://lichess.org/study/GzhGMO8d/p42FTnxP

When I solved this puzzle the first move I looked at was 1. g4+. Taking the pawn seems very risky since black exposes themselves to a discovered check from white's bishop so I focused on 1...Kh4. I then saw the line 2. Rh6+ gxh6 3. Qxh6 Kg3 4. Qh2#. Try to figure out what is wrong with this line!

The problem is that after 1. g4+ Kh4 2. Rh6+ black does not have to capture the rook and allow white's queen into the attack, but can instead play 2...Kg3 immediately. There are some more checks to examine but white's attack inevitably ends, for example 3. Rh3+ Kxh3 4. g5+ Kg3 and it's over, if 5. Rg1+ Qxg1+! 6. Kxg1 Bxa6 and white can resign.

Similar to the last puzzle, we can put the moves in a better order and begin with 1. Rh6+. Playing this on move 1 is much more forcing as black must play 1...gxh6. Now we can continue as before: 2. g4+ Kh4 (this move order also eliminates the need to calculate ...Kxg4 since in this position the g4 pawn is already guarded by the d7 bishop) 3. Qxh6+ Kg3 4. Qh2 mate. We've reached the same final position as in my original line without giving black the option to escape.

https://lichess.org/study/GzhGMO8d/CJfjAaJp

When I solved this puzzle, I began by looking at the one sequence of clearly safe checks: 1. Qa8+ Kd7 2. Rg7+ Ke8. Here I think I briefly scanned most of the follow up checks and decided that this variation was a dead end, since they are mostly pointless sacrifices (Rg8+??, Re7+??, Qxd8+??, etc.), and my initial impression was that I actually chased black's king to a safer area of the board.

My first thought was to switch the move order. White should begin with 1. Rg7!, threatening to mate next with Qa8. The puzzle isn't over as black can try the defensive move 1...Qb6!. Now white must find 2. Qa8+ Qb8 3. Rxc7+ Kxc7 4. Qc6#.

What happened in this puzzle is very common - by switching the order of moves in an attacking sequence, we often turn a check into a mate threat or vice-versa. Personally, I need to look at moves that threaten mate more diligently. Often I catch myself calculating a long series of checks, chasing the opponent's king across the board, and I have to remind myself that I should look for a non-check move to restrict the king's movement and set-up a checkmate threat. For my students in the 1000-1500 range I notice the opposite is usually true - they often play mate threats that are easily stopped, for example Qh5 threatening Qxh7, when the opposite move order - sacrificing on h7 first, then bringing in the queen with check - is stronger. I think this is because they are still overcoming the typical beginner thought pattern of "I will threaten h7, and if they don't see it, I will checkmate them on the next move," instead of looking at sequences of forcing moves.

Examples From My Games

Coming soon?

One Final Position

The following position is a great chance to work on the switching the move order technique. I'm not sure where I found this position but my guess is I saw it on that one other chess website and that it's a composed endgame study (don't worry, it's not as hard as most endgame studies). White is down two minor pieces but has a lot of passed pawns. White's candidate moves and potential resources are very obvious: bxa6 creates a new passed pawn on the queenside, the h-pawn can be turned into a queen with check immediately, we have a standard forcing check with g5+, and finally the additional passed pawn on g6 is likely pretty important!

To solve the position you just have to put the moves in the correct order! With four moves as part of the tactic, and all of them being legal starting from move 1, there are lots of possible orders! A good first step towards solving the puzzle is to establish which first moves obviously do not work. Be aware that when you feel like you have solved the position there may be one final tricky defensive move for black, which actually introduces a second, shorter layer of tactics with an interesting but easier move-order issue. Below the diagram is a short conclusion to this article and below that you can find the solution to this puzzle.

White to play and win:

https://lichess.org/study/GzhGMO8d/cEKxAtZO

Conclusion

Switching the move order is a standard calculation technique and is a specific way to spot different versions of a singular underlying idea (the subject of a future article?). The main time you want to use this technique is when you see a tactic that doesn't quite work. Quite often it actually does work, the moves just need to be in a different order. Another time to use it is when you've been looking at a position for a while and feel stuck. Go through the ideas you've already looked at and switch the order of moves. One reason strong players are strong is that they have really internalized this technique and apply it to the majority of the ideas they see during a game!

Final Puzzle Solution Below

https://lichess.org/study/GzhGMO8d/ZLCbizUh#0