Minesweeper pisses me off. Shooting brakes don't.

"Garrett Davis" (GarrettDavis) 04/01/2015 at 17:15 • Filed to: None

1

12

I play a lot when I'm on conference calls when I don't need 100% focus, and I find myself running into walls of guessing way too often. I could be missing something from the above, but I spent 10 minutes staring at it to no avail. I used to think of Minesweeper as the ultimate logic puzzle where there is always an answer, but the more I play it, I run into these scenarios more often than not. Guessing should not be a part of this game!

Between situations like this and accidentally clicking the wrong tile, I very rarely actually finish a game anymore.

/rant

Shooting brake Shelby porn for your time:

Rainbow > Garrett Davis 04/01/2015 at 17:21

1

The one next to the middle three is safe. There must be one in either of the two below it in order to satisfy both the 2 and the bottom 3, so if there was a mine here, that would mean the bottom 3 is touching 4 mines.

So there's that, at least.

Dusty Ventures > Garrett Davis 04/01/2015 at 17:21

0

Watching my dad play minesweeper would crush your soul. Like watching Mats Valk with a Rubik's Cube

duurtlang > Garrett Davis 04/01/2015 at 17:23

1

You have a lot to learn. Let me help you out. Stars are safe, arrows are mines. No guessing involved here. Can you figure out why?

Garrett Davis > Dusty Ventures 04/01/2015 at 17:26

1

Maybe I need to figure out some more advanced strategies here. I know there are people like your dad who tear through them, I just wonder what I'm missing to have to guess so often. Or maybe I just need to realize that there's a lot more guessing than I thought. Hmmm...

Garrett Davis > Rainbow 04/01/2015 at 17:29

1

Ahh, you are correct. I need to start thinking more here.

Garrett Davis > duurtlang 04/01/2015 at 17:31

0

I guess I do. I see where some of these would be safe, but maybe I need to figure out the logic on some of the others.

Garrett Davis > duurtlang 04/01/2015 at 17:34

2

I trusted you!

duurtlang > Garrett Davis 04/01/2015 at 17:40

1

Ha! You're right, I used the wrong icon. Sorry about that :)

Rainbow > duurtlang 04/01/2015 at 17:48

0

I see a problem at the top, actually. If that arrow is a mine, then the next two over to the left are safe. (as that 2 will have both of its mines there) In that case, the star would need to be a mine in order to satisfy the middle 2. Either the star or the arrow is wrong.

duurtlang > Rainbow 04/01/2015 at 17:49

1

You're right. I used the wrong icon. He found out the hard way, see his reply. Now I actually feel bad.

RamblinRover Luxury-Yacht > Garrett Davis 04/01/2015 at 18:04

0

Probabilities help some of the time, but it's all really just trained intuition at that point. Actually puzzling out the probability overlaps is a *huge* hassle - see idiocy below.

It's more likely that the far left-hand side has a mine over the left mine than over the 2, just due to the combined probabilities, I think - in short, if the remaining mine count is low enough you're a little more likely to have a mine between a 2 and a 4 than two separate mines for separate answers. You have to compare the likelihood of a square being a mine overall to being a mine in the specific, though, and if it's more likely any given square *is* a mine than *isn't*, it makes multiple solutions the preferred expectation instead. The tiles above the 2 have a 50% chance each of being a mine, obvs. The remaining 4 unknown tiles around the 4 also have a 50% chance each of being a mine (sort of). Further, the spaces next to the 2 next to the 4 each only have a 1/3 chance each *based on that square only*... but the combined likelihood of there being a single mine over the 2 that satisfies both the 4 and the 2 to the right is slightly higher than some other possibilities. If that's the case (safer to test), then the other squares next to the 4 go to 1/2 probability each specifically instead of generally (two squares, one is a mine) If the one up and to the right of the 2 far to the left is a mine (50% unlikely) and another mine is known, then all is known: 50% unlikely. If that square is not a mine (50% unlikely), then there's also only one remaining possibility for the (other) mine to be. I *think* both actually work out to the same base probability, which then makes it hinge on the likelihood of mines overall. In your example, two mines are still less likely than one.

tl;dr I confused myself like 8 times there, and I'm still not sure I have it right. My basic point, though, is that it would be possible to build a probability analyst to train for this. You'd be guessing a lot, but less than random, and it's possible to approach that with brute intuition. You'll still lose a lot.

RamblinRover Luxury-Yacht > Garrett Davis 04/01/2015 at 18:09

1

Hmm. In this, I can see that one of my specific conjectures on the far left was wrong, but the general conjecture (mines in an L between the 2 and 4) was correct.