The Flag Controversy - Intro

   Is flagging or clicking the best speedwise?

   The goal of minesweeper is to open empty squares. The question is what method accomplishes that goal most quickly.

   Most of the best players in the world, such as Lasse Nyholm (44), Sriram Sridharan (49), Damien Moore (52) and Dan Cerveny (52) flag the majority of mines on expert, yet others, like Mike Lowder (53), flag none. I never flag on beginner, I flag about 15-20 mines on int and 70-85 mines on expert.

   The defining factor for speed is efficiency.


The Flag Controversy - Clicks

   An important thing to note with clicking, is that openings occur only if you click on a blank square. If you hit a number, even if that number is touching empty squares, nothing will happen.


The Flag Controversy - Flags

   In minesweeper, when a mine is known, you can mark it by right-clicking on it, causing a flag to appear on that square. If you have marked all the mines touching a number, you can click both buttons, or "double-click" on the number to clear all other spaces touching that number. While for beginners the goal of flagging is to remind them of mine locations, for experts flags are means to the end of opening space quickly via the double-click method.


The Flag Controversy - Flag-Alls 

  Flagging everything is the slowest of the player types. In these examples, flagging is a complete waste of time and unecessary. It is from such cases that people jump to the conclusion that no flags is better.

Flagger Rule #1 Never flag unless double-clicking is useful.



The Flag Controversy - Obvious Clicking

   In each of these cases, one click is most efficient.

                     Here are more instances where clicking is more efficient and faster. For example, the fastest way to clear these corners is one click to the corner square. Remember that you must click on a blank square to get an opening: only two squares works in the left pic and only one in the right pic; the other squares contain numbers.

    An outstanding example, the best I've seen, in favour of clicking is this one. If there are 3 mines, one click is sufficient, if 2 mines, then 2 are needed. Contrast this with the following two scenarios: a hardcore flagger in the case of 3 mines flags once and double-clicks, if 2 must flag twice and double-click twice: or, the efficient clicker who flags once and double-clicks for 3 mines, F&DC's and then clicks once. Clicking definetely comes out better.


The Flag Controversy - Obvious Flagging

   And, there are instances where flagging is faster. In each of these situations, a flag and double-click on the protruding square is faster than three clicks.


The Flag Controversy - Efficient Clicking 

   The goal of minesweeper is to open the empty spaces; and openings don't occur unless a blank square is clicked on. Consequently, an efficient clicker goes for the gold: blanks.

   In this examples, a poor clicker could lose time if he clicks on any of the safe squares with numbers, such as the middle left or middle bottom. If you click there flagging would be far more efficient. Clicking is more efficient here only if the empty square, the corner, is opened; the opening will then expand until it hits numbers in all directions, thus exposing the squares with numbers.

   I had to show you this circus example. Two clicks to the corners! Of course, there are many more situations where clicking should be done efficiently. For instance, and this occurs all the time practically, when you spot a mine don't waste time clicking beside it because you'll always hit numbers. instead, aim for next-to-nearby squares that suggest they're empty. The openings will open the numbered squares for you, except of course in such situations as detailed in "better click."

Clicker Rule #1  Click for open spaces. 



The Flag Controversy - Efficient Flagging

   The following examples are to help you flag more efficiently to increase your speed. Some of these can be done faster with clicking, but I've just pulled some random easy examples to get points across. As I or you come across better cases, they'll be posted.

                In the first example, flagging four times is time-consuming. The fastest way is putting a flag beside the 2 and double-clicking on the one beneath. I call it tail-ending because you flag & double-click at the end of a line of mines, ignoring the main body which takes care of itself. In the second case only the top square needs a flag; as I said, these examples are to get points across, not show the most ideal situations. I would click rather than flag for the second example.

Flagger Rule #2 Ignore bodies, flagging tail-ends. 

                          These examples from the "flags rule" section are another case in point. The first example has two mines but it is only necessary to flag once. The second example has three mines but it is again sufficient to flag the one time. As a beginner I used to flag and double-click every mine, and would even have flagged the third mine though all the surrounding squares had been opened (refer to rule #1 for flaggers and see "flag-all" examples).

                       The fastest way to flag each of these (although clicking these particular examples is faster) is once on the 1 beside the empty square. (In each of these using my flag tip, flagging is only an 1/8th of a sec slower, but still not as good as clicking) Use this idea to cut back flags that you're not using efficiently.


The Flag Controversy - Toss-Ups

   All other cases require an equal number of clicks and so are equally fast, right? Here's an average example. Please send me more.

An efficient flagger would flag the 1 under the 2 and double-click, whereas the clicker could click twice in the two empty squares. Both score 2 motions.


The Flag Controversy - Flag-Click Trick

   In many cases you can either click to open a square or you can  flag & double-click on a nearby mine. The second option usually takes twice as much movement and twice the time, adding up to a 1/3rd of a second in such cases. But, I've found that you can flag and double-click in one motion.

   As you press the flag button, already begin pressing the click button. Aim the flag for the mine and, with the flag button still depressed, slide onto the nearby number letting the click button finish coming down. The result is a flag & double-click without having to hit the flag button twice. Basically, you have eliminated the seperate flag action, and now do a one-stop double-click action. Assuming each extra click takes 1/4th if a second, this saves you about a 1/5th of a second every time you use it. And, assuming you you flag 70-80 mines on expert, it adds up.


The Flag Controversy - Toss-Ups Revisited


   As noted, both the clicker and flagger require 2 clicks. There is no way to cut down the time for the clicker except gradually becoming faster, but the flagger can also become faster. But, with that tip, flagging time is cut by 25-40% thus pulling its time down under the clicker.


The Flag Controversy - Brain Tease

   This takes brain. The most efficient flagging is beside the bottom 2, with a double-click (using my flag-click trick) on the 1. An efficient clicker could argue that the fastest method is a single click to the middle bottom square. Which is it?

   Remember that clicks are only efficient if in a blank square: what happens if there is a mine in any of the lower right 3 vertical squares? In that more likely case the bottom middle square would be a number and up to an additional two clicks are needed to clear the surrounding space that would've been cleared with a double-click.

  You judge: The click method will will be about an 1/8th sec faster in the case where the surrounding squares are empty, but the other half to 2/3rds of the time an extra one or two clicks will be needed at a 1/4 sec approx. With the flag method using my tip, you lose an 1/8th of a sec a third of the time, and you save a 1/4th or more of a sec most of the time.

   Of course, this is just this particular case. I hope it's opened some horizons on the intricacies of which side is better, if any is.


The Flag Controversy - Conclusion

   No technique is right or wrong: the definition of a good technique is one that is efficient. Efficiency is getting something done with the least amount of effort, and as we know, effort takes time. Increasing your efficiency is necessary for good scores.

   As cases in favour of flags show, someone who clicks everything often adds some unecessary time to their game. As the more numerous cases that favour clicking show, flagging everything is obviously not the way to go for those who want to set records. I think anyone would agree that clicking all is better than all flags.

   Yet, the difference between efficient flagging and efficient clicking is narrower. Leaving aside those cases where an efficient flagger wouldn't bother flagging (i.e., where a double-click wouldn't clear anything) most cases take an equal amount of button-pressing whether you flag or click as a general rule. But in most of these circumstances my flag tip will take a fraction of a second off in favour of flagging. So, in all, an efficient flagger enjoys a slight, and I repeat slight, advantage over efficient clickers. (See footnote)

   Yet surely, isn't the most efficient way to take each situation on its own terms and flag or click accordingly? I believe that the ideal is to reject loyalty to either extreme and to take the middle road with discernment. The result will be efficient clicking where clicking is obviously faster, efficient flagging where flagging is obviously faster, and that the rest of the "toss-up" situations are solved quickest with a one-motion F&DC.

  Tip:   The average square is surrounded by 8 others. If the square is a 3 or lower, it is best to flag thrice and double-click. If the square is a 4 or higher it is better to open the remaining 4 squares by clicking, as a general rule. It depends on the context the number is found in.

   [Some object that clicking efficiently must be better because "the very few non-flaggers have some very good times." Although it is an interesting idea, the statement is incorrect because many clickers also have bad scores, such one guy who wrote me recently. The real question is whether clickers over-represent themselves in the rankings; this has yet to be suggested by statistics. As it currently stands, the world's 5 best sweepers flag. (96% in the rankings do).]