Website archived by minesweepergame.com

Strategy

From MinesweeperWiki
Jump to navigationJump to search

Minesweeper strategy is the art of solving games. Techniques include learning patterns and where to click first, using guessing tactics and developing efficient clicking and mouse movement.

Patterns

A pattern is a common arrangement of numbers that has only one solution. If you memorise a pattern it will reduce the amount of time you waste thinking.

Before you start learning patterns, you should learn the basics. If a number is touching the same number of squares, then the squares are all mines. You can solve most Beginner games this way. Here are some examples:

  • The 1 on the corner touches only 1 square, so it must be a mine.

  • The 2 touches 2 squares, so they must both be mines.

  • The 3 touches 3 squares, so they must all be mines.

  • The 4 touches 4 squares, so they must all be mines.

  • The 5 touches 5 squares, so they must all be mines.

  • The 6 touches 6 squares, so they must all be mines.

  • The 7 touches 7 squares, so they must all be mines.

  • The 8 touches 8 squares, so they must all be mines.


There are two basic patterns which combine to make all other patterns. The first is 1-1 and the second is 1-2. Whenever you see a 1-1 pattern starting from an edge (or where an open square functions as an edge) the 3rd square over is empty. This makes sense because the first 1 touches two squares, which must contain the mine, while the second 1 also touches a third square, which must be empty. Whenever you see a 1-2 pattern the 3rd square over is always a mine. This makes sense because the first 1 touches two squares, which must contain the mine, while the 2 also touches a third square, which must contain the second mine. Here are some examples:

  • There is 1 mine in the first two squares, and 1 mine in the first three squares. The 3rd square over must be empty.

  • There is 1 mine in the first two squares, and 2 mines in the first three squares. The 3rd square over must be a mine.

  • A 1-1 pattern starting from an opened square.

  • A more complicated version of the 1-2 pattern.


The two most famous patterns are 1-2-1 and 1-2-2-1. These are so common new players should memorise them immediately. If you look carefully they are just combinations of the 1-2 pattern.

  • The 1-2-1 pattern has one solution.

  • Apply the 1-2 pattern from the left.

  • Apply the 1-2 pattern from the right.

  • The 1-2-2-1 pattern has one solution.

  • Apply the 1-2 pattern from the left.

  • Apply the 1-2 pattern from the right.


At first it seems like there are many patterns. If you study them, they are actually 1-2-1 and 1-2-2-1 patterns (or combinations of patterns). These in turn are variations of the basic 1-1 and 1-2 patterns. Each set of numbers reduces when you subtract known mines. Here are some final examples:

  • 242 reduces to 121‎

  • 345 reduces to 121

  • 1222 reduces to 1221

  • 2331 reduces to 1221

  • 222 reduces to 121

  • 2331 reduces to 1221

  • 13231 reduces to 12121

  • Reduces to 122121

Guessing

Sometimes in Minesweeper you need to guess. A typical case is a 50/50 situation where one mine is hidden in two squares. Guess quickly and move on. Thinking does not improve your chance of guessing correctly, it only wastes time. Waiting to see if you guessed right also wastes time, so assume you survived and try to keep playing. Do not delay taking forced guesses - solving the rest of the board first is a waste of time if you end up guessing the wrong square.

Many players are impatient and guess instead of solving. Do not guess unless it is necessary. The fastest way to solve 'Example A' is to click the unopened squares in a row. But if you click fast there is no time to react, so you will lose if the middle square is a mine. You have just guessed for no reason! A smart player will click the outer two squares first, which allows enough time to react to the initial click and decide if there is a mine.

Opening safe squares is as important as finding mines. If you can prove a square is safe, open it instead of guessing where the mine is. In 'Example B' there is a mine in the two yellow squares. Instead of guessing, open the safe 3rd square. This can allow you to open even more squares (marked blue) which may help you solve the original guess.

If you need to guess and there are more empty squares than mines involved, it is always better to guess an empty square instead of guessing a mine. Flaggers often make the mistake of guessing the mine because they love to chord.

Sometimes you can improve the chance of guessing right. There might be an arrangement of numbers with more than one solution, and the solutions require different amounts of mines. Instead of guessing, you can solve it by flagging the rest of the board and seeing how many mines are left. You can solve 'Example D' if there is 1 mine or 3 mines, but you must guess if there are 2 mines left. If you decide to save time and guess immediately, think about the mine density of the level you are playing. For example, the solution with more mines is more likely on Expert than Intermediate. Keep in mind though that the density of each level is pretty low, so less dense solutions are more common overall.

  • Example A: Do not take unnecessary guesses. Open the two outer squares first. While clicking on the 3rd square you can look at the result of the 1st square.

  • Example B: Open a safe square instead of guessing where the mine is. In this case you can now also open the blue squares.

  • Example C: You can not find any mines without guessing, but you can open the safe blue squares. If a blue square is a 1 you can then open the orange squares, and so on, maybe until the board is solved.

  • Example D: Is there 1, 2 or 3 mines in the corner? Find out before you guess.

Perhaps you have solved part of a board and need to guess in order to reach the rest of the board. You can improve your chance of winning by clicking randomly! The average chance of hitting a mine is 0.206 on Expert and 0.156 on Intermediate and Beginner. These odds are much better than a 50/50 guess. Remember you are more likely to get openings by clicking on edges. Your bravery is often rewarded by finding that the original 'guess' becomes solveable when approached from a different direction.

Another important thing to remember is usefulness. If two solutions are equally likely, choose the one that will help most if it is correct. Sometimes one solution eliminates another guess, or gives an easier arrangement of mines. A common mistake is turning a 33/66 guess into a 50/50 guess instead of solving it. For example, if you know there is one mine in three squares, do not open the middle square!

Always choose the most likely solution. This can be very difficult to calculate! Sean Barrett has written Minesweeper Advanced Tactics as a guide. Local probability is easy to calculate but is usually wrong. For example, in the image below some squares are both 50/50 and 66/33 guesses! When all unsolved areas are considered, a simple 50/50 guess often has one square much more likely to contain the mine. A general rule of thumb is that if one square in a 50/50 situation touches a high number, it is more likely to be a mine than the other square.

A special case of probability is when guessing involves the top left corner. Minesweeper makes the 1st click safe, so if you click a mine it is moved to the top left corner (or the nearest empty square on its right). If you have a 50/50 guess and one square is the top left corner, the corner is always more likely to be the mine. When starting an Expert game the chance of a mine somewhere is 0.206 but the top left corner nearly doubles to 0.370 after the 1st click.

The following example illustrates many of the above points. It looks like there are three unavoidable 50/50 guesses, and two unavoidable 66/33 guesses. One strategy is to guess quickly and hope for the best. This option will give the best score if you survive. A second strategy is to click a random square that does not touch any numbers. This usually has better odds of being safe and often helps solve the game. A third strategy is to determine the number of mines remaining by flagging the rest of the board. This reduces the number of solutions. In this example there are 79 possible solutions but only 2 of them have 4 mines. A fourth strategy is to guess in the most useful place. Clicking square I has the potential to eliminate all the other guesses! For example, if it is a 4 or 7 the game can be solved no matter how many mines remain. A fifth strategy is to guess the most likely solution. A mine is more likely in L than K and more likely in H than D. A final strategy is to calculate the exact probability of each square taking the entire game into consideration. This is the hardest but most accurate method. Results for this example are available.

The guessing strategy you choose depends on whether you want to win more games or make time records.

  • Three 50/50 situations

  • Two 66/33 situations

  • Guessing1c.png

First Click

The first click in Minesweeper is always safe, but where is the best place to start? It depends whether you want quantity or quality.

Your best chance of finding an opening is in a corner, then on an edge, then in the middle. Emmanuel Brunelliere (France) calculated the theoretical odds as follows:

Beginner Intermediate Expert
Corner 59.54% 59.94% 49.94%
Edge 42.14% 42.61% 31.42%
Middle 25.09% 25.54% 15.69%

Tim Kostka then used his knowledge of Board Cycles to find the actual chance of finding openings on Windows Minesweeper. The first click is always safe because any mine is moved to the top left corner or nearest empty square to its right. This means the top left corner gives fewer openings than the other corners. It also means fewer openings result from the edge and middle squares touching the top left corner. Exact values for each square are on his website. Most of the variation is due to low outlier values near the top left corner.

Beginner Intermediate Expert
Corner 50 - 60 % 50 - 60 % 40 - 50 %
Edge 34 - 42 % 36 - 43 % 25 - 32 %
Middle 19 - 24 % 21 - 26 % 12 - 16 %

Your best chance of getting a large opening is in the middle, then on an edge, then in a corner. So far no one has calculated the theoretical advantage, but Tim collected actual results from Windows Minesweeper. The biggest openings occur in the very center of the board and decrease as you approach edges. The biggest openings from clicking on an edge are in the middle and decrease as you approach corners. This chart shows the variations in the average number of squares for each opening:

Beginner Intermediate Expert
Corner 18 27 16
Edge 20 - 24 31 - 42 19 - 26
Middle 23 - 32 35 - 66 23 - 41

In summary, the best place to start depends on your preference for size or frequency. Large openings are more helpful but you will lose more games trying to find them. Small openings can be difficult but you will start more games. It is possible the benefits of either method cancel each other.

The Windows Vista version of minesweeper always gives an opening on the 1st click. In this case you should always start in the middle to get the biggest opening on average. (This version is not accepted for the World Ranking.)

  • Probability of openings on Beginner.

  • Average size of openings on Beginner.

Efficiency

The fewer clicks you take, the faster you will finish. Learn to be efficient.

The game ends when all safe squares are open, not when all mines are flagged. Beginners often waste time flagging every mine. The only good reason to flag is to clear more squares by chording. So before you place a flag, decide if it is useful.

  • There is no reason to flag this mine except to make it look pretty!

  • Flagging these 8 mines is a waste of time because you can not chord on the 8.

  • There is no reason to flag the pink squares, because you can not use them to chord.

Some players never flag because time spent flagging can be better used to open more squares. This style is called No Flags, or NF. Flaggers argue that flags allow you to chord and clear multiple squares at the same time. It is generally agreed that NF is more efficient near high numbers (5,6,7,8) while Flagging is more efficient near low numbers (1,2,3,4). Near a high number like 7 a NF player needs only one click to open the safe square, but a Flagger needs seven flags and a chord. Near a low number like 1 a Flagger would place one flag and chord, but a NF player would need as many as seven clicks to open the safe squares. It is also generally agreed that NF is more efficient on low 3BV boards while Flagging is more efficient on high 3BV boards. For example, an Intermediate game with 3BV 40 has an average of one number touching each mine, while a 3BV 120 game has an average of three numbers. A perfect NF player would need 40 and 120 clicks. A very inefficient and unlucky Flagger would need 80 clicks (40 flags, 40 chords) for both games. These examples are extreme cases but show the general reasoning. In reality NF players are not perfect and waste clicks at full speed, while Flaggers never need to flag all mines or chord on every number. If a player uses only NF or Flagging, there is probably no advantage to either method. The advantage comes when both techniques are combined and the player uses the most efficient solution for each situation.

If you flag you can save time by using the 1.5 Click technique. The normal way to chord is moving the right button down and up to flag, and move both buttons down and up to chord. The 1.5 Click trick moves the right button down to flag, presses the left button down, and releases both up to chord. This eliminates one movement from every flag and chord combination. As long as the right button starts going down before the left button, the flag will get placed. The shorter the gap, the more time you save. You can nearly double your flagging speed with this method.

Here are some examples of efficient flagging:

  • You could flag all 3 mines and chord on a 2, but it is more efficient to flag as shown and chord on either 1.

  • You could flag both mines, but it is more efficient to flag the outer mine as shown and chord on the 1.

  • Flag as shown and chord on the 1 above. There is no need to touch the other squares.

  • There is no reason to flag the pink squares. Flag as shown and chord on the 1 beside it.

Here are some examples of efficient NF:

  • Click once on the yellow square.

  • Click once in the corner.

  • If the yellow square is an opening, the blue squares will open. NF players look for openings.

  • If the yellow square is an opening, the blue squares will open.

It is not always easy to tell if NF or Flagging is more efficient. In these next four examples NF takes fewer clicks if there is an opening, but it takes more clicks if there is no opening. name removed has made an excellent slideshow of an Intermediate game being solved efficiently, with detailed explanations.

  • Example A: An efficient Flagger would place one flag and chord on the two pink squares. Then a click would be made on the yellow square. Total of 4 clicks. (If you flagged all the mines you would need 8 clicks.)

  • Example B: A NF player would click on the yellow square and hope an opening clears the blue squares. Then a click would be made on the two purple squares. Total of 3 clicks with an opening or 6 clicks with no opening.

  • Example C: NF players will click the yellow square. If it is an opening, at a minimum the blue squares will open. If it is not an opening, the player needs to click all coloured squares. Total clicks vary from 1 to 6.

  • Example D: Flaggers will put one flag and chord on the 1. If there is an opening, the blue and green squares all open. If there is no opening, the blue squares open and the green squares need to be clicked. Total clicks vary from 2 to 5.

An important way to increase solving speed is to make fewer mouse movements. It takes time to move your mouse. New players follow their eyes with the mouse instead of only moving it intentionally toward a target. The next stage in reducing movement is learning to 'see' the solved board. This often allows you to solve at your current mouse location. For instance, if your mouse is near the 2 in 'Example A' you can flag the red square and chord instantly. This is obvious to a professional player because they have solved the adjacent squares in their head. A new player would have to move elsewhere and come back later. The red square in 'Example B' can be similarly solved. Less movement equals better scores.

  • Example A: The red square is easily marked as a mine because the player is thinking several moves ahead.

  • Example B: The red square is easily marked as a mine because the player is thinking several moves ahead.

Efficiency is measured by your Index of Efficiency, or IOE. This compares the number of clicks taken to the 3BV of the board. An IOE of 1.00 means you solved a 3BV 50 board in 50 clicks. It is possible to solve a game in fewer clicks than its 3BV by combining Flag and NF techniques. Both Clone and Arbiter save IOE highscores as an incentive for improvement. Arbiter further breaks IOE into Correctness (clicks that changed the board) and Throughput (the potential IOE if all clicks had been correct). It also has a Path statistic that measures mouse movement in pixels.

The best way to improve efficiency is to play slowly. Find the most efficient solution and path to each problem before pressing any buttons. You will soon see improvements while playing at full speed.

More Tips

  • Do not use Questionmarks.
  • Press 'F2' to start new games. Keep one finger on this button, it is faster than using the mouse.
  • Avoid moving the mouse without a reason. New players often waste time moving the mouse everywhere their eyes look.
  • Ignore the clock. Looking at the clock during a game wastes time, and will make you nervous if you are going fast.
  • Many players listen to music while they play. This distracts them and lets them play on autopilot without nerves.
  • Play in a warm room or heat your hands in hot water before you play. This increases blood flow and reaction time.
  • Take short exercise breaks to increase blood flow and stimulate your brain.
  • After a long playing session, it can help to change the version you are playing. This helps focus your eyes.
  • If you accidentally click down on a mine, slide onto a different square before releasing the mouse button.
  • Use the 1.5 Click.

Links

Retrieved from "https://minesweepergame.com/website/authoritative-minesweeper/wiki/index.php?title=Strategy&oldid=3082"