Lucky mode

Anything to do with minesweeper...
Cryslon
Posts: 122
Joined: Sun Dec 28, 2008 7:41 pm

Lucky mode

Post by Cryslon » Tue Apr 03, 2012 8:15 pm

50/50s suck. Everybody knows that. And we can do something about it. We can make the game better, reducing the frustration and increasing the fun.

There exist several non-official 50/50-free minesweeper versions such as MinesweeperLive, Tatham's minesweeper and Luckysweeper. It seems the general consensus is these versions are not "real" minesweeper, ie they differ too much from the original game. As Thomas Kolar points out, they lack the risk management aspect of the game.

Two different approaches are suggested below. Unlike MSLive and Tatham methods, these approaches do not affect the variety of generated boards, ie, every board that can be generated by the original, non-lucky minesweeper, can also be generated by a lucky minesweeper. Also, they affect only guesses where no brain work is needed, and don't affect the "risk management". And i think allowing a lucky minesweeper as an official clone is worth considering.

The first approach is a library method. While the number of different guess patterns is infinite, some of them are by far more common and therefore more frustrating. So it is possible to create a library of these common guess patterns. The lucky minesweeper based on library works like this: if you hit a mine, then the program checks if it belongs to one of the guess patterns. If it does, then it shuffles the mines and you continue playing.

The second approach is a solver method. It also begins to work after you blast. When you hit a mine, it finds all the arrangements of mines that do not contradict the opened numbers. If there is a solution without mine in the square you just hit, it checks whether it is possible to eliminate this solution by opening one of the closed squares not containing a mine. If it is not possible, then it is a 50/50 situation, mines are shuffled and you continue playing.

I know only one argument against allowing things like these on the ranking. It goes like this: it is unfair for the old players who didn't get the chance to play lucky version. Or like this: records set on the new version are incomparable with records set on the original version. I find this argument unconvincing. The new gameplay doesn't differ from the old gameplay. If you finish a game on the original, it means the original version behaved exactly like the lucky version (and remember, the ranking is based only on finished games).

Also, I wanna stress that luck does not magically make you faster. It just reduces the frustration you feel when you spend half a minute beautifully solving a board and then realize that you just blasted due to pure randomness.

Both methods are implemented in Freesweeper; you can try them and see that they don't affect the gameplay.
Go IRC! (try mibbit)

EWQMinesweeper
Posts: 410
Joined: Sun Nov 30, 2008 11:50 pm

Re: Lucky mode

Post by EWQMinesweeper » Tue Apr 03, 2012 8:44 pm

if you no longer blasted on 50-50 guesses (what about isolated squares and 33-66?), you would finish more boards, at least get further on boards, thus getting more practice in playing the later parts of a solve, eventually being better at solving the last third or quarter or fifth of a board.

i'm fine with this being available on unofficial versions, but for the reasons you stated, i think it is not compatible with official versions.

in addition to tommy's argument i'd like to mention that when going for the fastest solution (which is what is relevant for official versions/rankings) there is always a best solution to a forced guess, even if it's only a matter of a 3px shorter mousepath. also because of that i don't see why you can let a game lost on a forced guess frustrate you.
„Das perlt jetzt aber richtig über, ma sagn. Mach ma' noch'n Bier! Wie heißt das? Biddä! Bidddää! Biddddäää! Reiner Weltladen!“

aradesh
Posts: 90
Joined: Sat Aug 29, 2009 3:37 pm

Re: Lucky mode

Post by aradesh » Tue Apr 03, 2012 11:56 pm

Some responses:

Cryslon: I think the scenario should be calculated before the player clicks the blast, to avoid latency upon clicking a square. The program could quietly in the background constantly be scanning the board for such 50/50's as the player is playing. If this is done well, it shouldn't lag the game.

Also this needs to be done very carefully, so that the computer can tell that there is no way that there could ever be more information on a guess, so as to not give the player luck unwarrantedly. Also there are tricky situations where the human might be able to see that they never will be able to get more information about the guess which might be hard to program a computer to find. But I talked to cryslon about this on IRC and he seems to understand this already.

[name removed]: Like you I'm also uneasy about it being in an official version. I'm not much bothered about optimal mouse path, as I'm pretty sure I can move the mouse say 3 squares in about the same amount of time it takes me to move 4. Also in another viable tactic is to gain speed, I sometimes aim for the larger target of "2 squares" and don't care which I hit. This way you get the opportunity of being more careless as you're not worried about which one you actually hit and can thus possibly do the motion faster, so your philosophy about mouse path isn't necessarily the only or best one.

Having given it a think, I think I'd be fine with this in an official clone because it does not change the boards at all, it wouldn't change the way we have to play at all - it would just be that every time we get an isolated 50/50, we just happen to be lucky. Indeed. All these 50/50's do is make it so that we solve less boards for no reason. Sure - there are other scenarios that kill us like this too and which won't get solved here - but this scenario is pretty common and easy to deal with.

It would be nice that we'd all be able to complete more boards.

Cryslon
Posts: 122
Joined: Sun Dec 28, 2008 7:41 pm

Re: Lucky mode

Post by Cryslon » Thu Apr 05, 2012 11:49 am

EWQMinesweeper wrote:what about isolated squares and 33-66?
Hidden mines are not affected by my luck functions. Isolated 33-66s are dealt in the same manner as 50-50s.
EWQMinesweeper wrote:if you no longer blasted on 50-50 guesses, you would finish more boards, at least get further on boards, thus getting more practice in playing the later parts of a solve
The whole point of the lucky mode is to allow people to finish more boards, because finishing boards is the main source of fun. You apparently think it's bad to finish boards, but i can't see why.

I don't think more practice on the later parts of solves is unfair. People already can practice these parts with different tools like Clone's or Arbiter's UPK mode.
EWQMinesweeper wrote:also because of that i don't see why you can let a game lost on a forced guess frustrate you.
Maybe that's because you finish thousand exps a day. Surely, a blast due to 50/50 on 40th second is no big deal for you, but it's a large source of frustration for less skilled players like me.

For the sake of argument, let's say there's exactly one 50/50 on each board. And let's say, you correctly solve every second board while i correctly solve every hundredth board. Then you will be able to solve consistently one board in less than 4 minutes, while i'm able to solve one board in maybe 200 minutes. And i don't usually play that long! Therefore i wouldn't even launch minesweeper if i know i'm going to play for less than a hour.

Then, imagine 50/50s are eliminated. Now, you're solving a board each 2 minutes. Sure, not a big improvement for you. But i'm now solving a board in 100 minutes, so it's now worth playing a hour. I hope i expressed my thoughts clearly.
aradesh wrote:I think the scenario should be calculated before the player clicks the blast, to avoid latency upon clicking a square.
This is too complicated. I don't think the lag would be noticeable if the luck function starts on blast.
aradesh wrote:http://a6.sphotos.ak.fbcdn.net/hphotos- ... 1823_n.jpg This is the kind of scenario where Cryslon's 50/50 eliminator should NOT help.
And it doesn't.

[name removed]also raised the following question deserving attention on the IRC channel:
EWQMinesweeper wrote:[00:57]< EWQMinesweeper> if lucky mode was added, how would you explain a sweeper why they had to lose on a forced guess that wasn't an isolated 50-50?
I think it is good to get rid of blasts on 3-1 and on 2x2 boxes. Even if the game would still contain some complicated and rare 50-50 situations.
Go IRC! (try mibbit)

dg211
Posts: 12
Joined: Mon Apr 02, 2012 7:19 pm

Re: Lucky mode

Post by dg211 » Thu Apr 05, 2012 12:42 pm

I don't think I would have a problem with this being included in official versions, provided it is completely certain never to allow additional information to be unfairly ascertained. To that end, I would prefer the library based approach, since that can be exhaustively verified, whereas I imagine it would be harder to comprehensively prove the solver free of such problems. I haven't checked out freesweeper yet, but does the replay record the fact that a mine has been moved and highlight it when playing back? That would be useful for others to confirm that no unfair advantage had been gained. Tommy mentions that he likes the risk management aspect of more complicated guesses and I agree that this is part of the game, but I would expect that any guess complicated enough to add interest to the game would also be complicated enough to potentially allow additional information to be obtained in a lucky mode and so would be excluded anyway.

Like Cryslon, the only real argument I can see against it is that is somewhat unfair to players in the past. However, if you follow this argument you are basically ruling out making any progress ever. It's like arguing that giving women the vote was unfair to all the women in the past who didn't have that right. And you could equally argue that the introduction of clones was unfair to previous players because they included much more sophisticated tools for analysing games and hence improving your play.

aradesh
Posts: 90
Joined: Sat Aug 29, 2009 3:37 pm

Re: Lucky mode

Post by aradesh » Thu Apr 05, 2012 2:36 pm

You know I think maybe this mode would be nice for a tournament setting. If you're at a tournament and you blast a really nice time right near the end on a 50/50, and you've not put up decent times, it can be really heartbreaking. You didn't do anything wrong. At home you can just keep playing for days and days until you do it again, but in the tournament setting you don't have this option. Just a thought.

I've been thinking about it more though and I don't think it would be good as an official clone. We'd have to give up all the other clones, or upgrade them (a big inconvenience). Also we're all used to playing minesweeper like this and we get by and complete games... Our records would likely be similar on this clone, it just might take slightly less time on average to put up a similar record (I think rather slightly, as losing a fast game in the last say 10 3bv for me is pretty rare. For example, I have 16 sub50's and I've only 50/50 LC'd a sub50 once) There's the possibility of a secondary ranking though.

dg211
Posts: 12
Joined: Mon Apr 02, 2012 7:19 pm

Re: Lucky mode

Post by dg211 » Thu Apr 05, 2012 3:13 pm

The world ranking should be a measure of skill, not the amount of free time you have to devote to playing. Finishing a game with lucky mode requires no less skill than almost finishing it and happening to blast on a 50-50. If you can get a good time this way, you could still have gotten it before, it just would have taken more time, not more skill.

If this were official, there would be an incentive to use clones that feature it and for other clones to upgrade, but no-one would be forced to switch. I don't think another ranking would be a good idea, we already have loads.

dg211
Posts: 12
Joined: Mon Apr 02, 2012 7:19 pm

Re: Lucky mode

Post by dg211 » Thu Apr 05, 2012 4:47 pm

Just did some very quick Monte Carlo (which may be dodgy) on 1k exp boards to count just isolated 50-50s

The results were:
0 - 798
1 - 169
2 - 31
3 - 1
4 - 1
which would mean if you played those boards perfectly and there were no other guesses, the expectation for completed boards would be 890.4
or you could expect to complete 12.3% more boards simply by removing isolated 50-50s

dg211
Posts: 12
Joined: Mon Apr 02, 2012 7:19 pm

Re: Lucky mode

Post by dg211 » Thu Apr 05, 2012 5:19 pm

Those figures just include boards where switching a mine with a horizontal or vertical neighbouring safe square wouldn't affect any other numbers on the board, i.e. only one specific type of 50-50. Also including 2x2 blocks containing 2 mines which could be on either diagonal gives the following results for numbers of forced guesses on 1000 exp boards:

0 - 576
1 - 315
2 - 74
3 - 28
4 - 4
5 - 3

so that with perfect play and ignoring other types of guesses you could expect to finish 755.8 boards, or 32.3% more if you removed just those two basic types of guess.

Note there may be a small amount of double counting on 2x2 blocks where each pair of mines could be switched independently.

qqwref
Posts: 121
Joined: Thu Sep 23, 2010 4:17 pm

Re: Lucky mode

Post by qqwref » Thu Apr 05, 2012 6:49 pm

Hang on, I already have a program to analyze this. Mine has three types of 50-50 patterns stored:
- 50/50 on square (0,0) and (0,1), and also the same pattern vertically
- the 2x2 square
- 50/50 on square (0,0) and (0,1), and (0,3) and (0,4), and also the same pattern vertically

Out of 100k exp boards:
0: 47607
1: 34105
2: 12951
3: 3901
4: 1010
5: 318
6: 84
7: 19
8: 3
9: 1
10: 1

In rare cases (50-50s nearly on top of each other, with very dense mine patterns) it might count twice, but it should be a good approximation anyway. Using these results you could expect to finish at best 68459.4 games, so let's say about 68.5%. Thus you'd get about 46% more games completed if you didn't have to worry about 50-50s.
NF player. Best scores 1-10-39.

EWQMinesweeper
Posts: 410
Joined: Sun Nov 30, 2008 11:50 pm

Re: Lucky mode

Post by EWQMinesweeper » Thu Apr 05, 2012 8:07 pm

these percentages only apply to the ratio of games you lose of exactly that kind of forced guesses. assuming that mistakes made by the sweeper account for the most blasts, i reckon the effect becomes very small.

also, by assuming that a better sweeper makes fewer mistakes, the effect will be bigger for sweepers who already finish a lot of games.
„Das perlt jetzt aber richtig über, ma sagn. Mach ma' noch'n Bier! Wie heißt das? Biddä! Bidddää! Biddddäää! Reiner Weltladen!“

aradesh
Posts: 90
Joined: Sat Aug 29, 2009 3:37 pm

Re: Lucky mode

Post by aradesh » Fri Apr 06, 2012 2:24 pm

Yeah I'm not sure it's fair to say "you will complete approximately x% more games". I mean maybe it is, I just haven't worked it out. Is this correct?:

"Out of those games you happen to not make errors in, and obtain luck in the situations which aren't covered in these cases, 48% of the time you will be tripped up by an isolated guess of these types."

I think the best way to test this would be for players who know roughly how many games they complete an hour, to play on this clone, and then see how many games the tend to beat per hour :P I'd be willing to say play for 10 hours on this clone if it was made to test this. I know that I complete about 5 games per hour on average.

User avatar
Tommy
Posts: 223
Joined: Mon Dec 01, 2008 9:22 pm
Location: Vienna

Re: Lucky mode

Post by Tommy » Sat Apr 14, 2012 2:08 pm

There was quite a long discussion about this topic on IRC a couple of days ago, and I would like to summarize my opinion.

I do not think that changing the game in such a way will make it suitable for the world ranking and current tournament systems.

And for a couple of reasons:
- Comparability with the current world ranking and old scores
- Simplicity of the rules of the game
- Changes in long-term risk management


Quite a bit of what I say will rest on the following assumption:
If the rules for a competitive sport are changed, there is a fundamental difference between making it easier and making it harder.

Take minesweeper. If you beat someone's highscore, but the game is easier now, it means nothing - you will have beaten the score, but under easier circumstances. But if the game is harder now, you have beaten the score despite the game being harder now - which is even better than beating it on an equal footing. And I therefore think that all measures to competitive minesweeper that are taken need to make the game harder, unless it is absolutely necessary (the lucky beginner score rule is borderline here, as it actually potentially reverted some old players to 1.00, but at the same time took otherwise unbeatable, lucky .10 records off the ranking).

Comparability with the current world ranking and old scores

This is an obvious one that I basically explained already.

Simplicity of the rules of the game

At the moment, I can explain the rules of the game itself to a newbie with ease:

This is a grid, some squares contain a mine, some don't, this is determined randomly after the game starts. You need to open all the non-mine squares to win, but if you hit a mine, you die and need to start a new game. Every square contains a number indicating the number of mines in the 8 adjacent squares; zeroes are blank and will recursively open surrounding squares as well if opened. You open a square by left-clicking on it, and flag a square as containing a mine by right-clicking on it. And if you release one mouse button while the other one is pressed on a square with a number, and there are as many flags as the number indicates surrounding the square, all other surrounding squares are opened. The middle mouse button does the same thing. The first square you open is never a mine. And finally, the timer indicates how long you have played already, the mine counter starts at the number of mines in the grid, and counts down for every flag you place. If you win, the time it took you from the first action that opened a square to the action that opened the last square is your score - the lower, the better.

That is all the rules of the game itself at the moment, if I didn't forget something. Most of them are obvious from just watching someone play. I made everything bold that isn't self-explanatory in my eyes. And question marks are useless ;)

Now, this isn't a problem with an approach that eliminates luck completely - you could just say "If you really need to guess, and guess wrongly, the game will make it so you guessed right after all".

However, I'm not entirely sure, but this may well be equivalent to the problem that is well-known to be NP-complete. So any practical implementation will need to be imperfect. And that means that the rules will have to state exactly what patterns will be recognized and what patterns won't. To explain this rule to a newbie, you need to go past the basic rules; You will need to explain details about the game that are best learnt through practice.

Changes in long-term risk management

Every professional minesweeper player trades consistency for speed. We all want to finish games, so we obviously don't take this to an extreme. However, it is also obvious that in lucky mode, there is additional free consistency. This means that more other risks can be taken to get the same consistency as before lucky mode - but with better times. This shows that lucky mode would not only make the game easier, it would make it easier to get good times, and this is highly relevant to the ranking, and nothing that can be ignored.

The argument that gameplay would be the same is invalid, because it wouldn't be.

But what shall be done?

The answer, in my opinion, is twofold:

Implement 3bv and/or ZiNi limits

http://minesweeper.info/forum/viewtopic.php?f=15&t=119
http://minesweeper.info/forum/viewtopic.php?f=12&t=84

Especially Ronny's article is very noteworthy. If every sweeper has a chance to get the easiest boards even fairly often, they will have multiple attempts to solve such an easy board, and forced guesses will not matter anymore. Even if there will still be frustrating blasts, everyone will know that they get another chance sooner or later.

By contrast, imagine Ian blasting his current expert highscore on a 50:50 near the end. That would be even more frustrating than trying to get a new highscore now would be futile.

Everyone, learn how to finish games!

I am being serious - it is not that hard. The level where most sweepers are frustrated is expert - int is too easy to win. So I will concentrate on exp.

Let me quote [name removed]for a moment (omitting stuff other people said in-between):
[19:55] tkolar EWQMinesweeper, how often do you win games?
[19:56] EWQMinesweeper tk: it depends on my mood, daily shape and style of play
[19:56] tkolar min/max?
[19:56] EWQMinesweeper some times i don't finish a board for 30min, some times 5 games in 10min
Now, [name removed]is one of the best players in the world, has by far the most sub50es, and really wins a lot of games. So saying that everyone should manage to complete games that quickly would be pretty elitist.
However, what I will say is that any competent sweeper should manage to win a game every twenty minutes on average. We all know that 5 games in 3 hours are not usually a problem for someone in the world ranking, and I am not even talking about extra-careful games.

If minesweeper is frustrating for you because you don't win expert often, work on your consistency. Forced guesses account for a good number of blasts, but they are only a part of it, and a smaller part than you may like to think.
Don't anthropomorphize computers - they don't like it.

Cryslon
Posts: 122
Joined: Sun Dec 28, 2008 7:41 pm

Re: Lucky mode

Post by Cryslon » Mon Apr 16, 2012 12:23 pm

Tommy wrote:If the rules for a competitive sport are changed, there is a fundamental difference between making it easier and making it harder.

Take minesweeper. [...]I therefore think that all measures to competitive minesweeper that are taken need to make the game harder.
I find this very unconvincing. You can't say what is "easier" and what is "harder". These are undefined terms. I'm unable to see how lucky mode makes anything easier.

Also, i think the pleasure players get from playing should have a higher priority than the elusive concept of a fair ranking. More so, the ranking is not fair now, and it has never been.
At the moment, I can explain the rules of the game itself to a newbie with ease: [...] Now, this isn't a problem with an approach that eliminates luck completely - you could just say "If you really need to guess, and guess wrongly, the game will make it so you guessed right after all".
As i've said above, we don't want to eliminate all guesses, because in some situations there is some skill in guessing: one guess gives more chance to win than another. We want to eliminate isolated stupid patterns like 3-1 or 2x2 box.
However, I'm not entirely sure, but this may well be equivalent to the problem that is well-known to be NP-complete.
Yes, it's NP-complete. But we have finite boards and we're able to find all the guess patterns and make a complete library. So we can write an efficient implementation.
And that means that the rules will have to state exactly what patterns will be recognized and what patterns won't. To explain this rule to a newbie, you need to go past the basic rules; You will need to explain details about the game that are best learnt through practice.
I don't see this as a big problem. Eg, we have 3BV limits; and you suggest ZiNi limits. These rules are much harder to explain, but you don't see this as a problem at all.
Every professional minesweeper player trades consistency for speed. We all want to finish games, so we obviously don't take this to an extreme. However, it is also obvious that in lucky mode, there is additional free consistency. This means that more other risks can be taken to get the same consistency as before lucky mode - but with better times.

The argument that gameplay would be the same is invalid, because it wouldn't be.
These sentences are in a direct contradiction with the last part of your post where you brag about how easy it is to win exp. If "professionals" you refer to are able to win exp every two minutes, then why these "professionals" aren't taking more risks to play faster? Certainly it wouldn't hurt to take three times as many risks and win exp every six minutes? Why they don't do it? I think that's because it's pretty much impossible for them to take additional risks to solve faster; they're already going at full speed.
Implement 3bv and/or ZiNi limits

If every sweeper has a chance to get the easiest boards even fairly often, they will have multiple attempts to solve such an easy board, and forced guesses will not matter anymore.
Huh? First of all, this is pretty much irrelevant to the topic. Secondly, tougher limits => less easy boards => rankings are effectively frozen. This is a solution for what?

And again, you seem to be concerned only with rankings representing some sort of idealized pure skill.
By contrast, imagine Ian blasting his current expert highscore on a 50:50 near the end. That would be even more frustrating than trying to get a new highscore now would be futile.
For me these two sentences don't make sense.
Go IRC! (try mibbit)

EWQMinesweeper
Posts: 410
Joined: Sun Nov 30, 2008 11:50 pm

Re: Lucky mode

Post by EWQMinesweeper » Mon Apr 16, 2012 3:02 pm

Cryslon wrote:
Tommy wrote:If the rules for a competitive sport are changed, there is a fundamental difference between making it easier and making it harder.

Take minesweeper. [...]I therefore think that all measures to competitive minesweeper that are taken need to make the game harder.
I find this very unconvincing. You can't say what is "easier" and what is "harder". These are undefined terms. I'm unable to see how lucky mode makes anything easier.

Also, i think the pleasure players get from playing should have a higher priority than the elusive concept of a fair ranking. More so, the ranking is not fair now, and it has never been.
artificially increasing the chances of surviving a forced guess isclearly making it 'easier'. the consistency of the rankings should not be touched, just to make things easier. a changing of the rules, as drastic as lucky mode, would ultimately have to result in a new ranking.

talking about the pleasure players get from playing: your reasoning, if i understand it correctly, is that losing on forced guesses is unpleasant. what i cannot understand is why there is any need for a density ranking then. also, doesn't this mean we would have to eliminate all forced guesses? (or all guesses in which the odds are worse than/equal to 50-50 and let the player win all other guesses if they take the best guess).

Cryslon wrote:
At the moment, I can explain the rules of the game itself to a newbie with ease: [...] Now, this isn't a problem with an approach that eliminates luck completely - you could just say "If you really need to guess, and guess wrongly, the game will make it so you guessed right after all".
As i've said above, we don't want to eliminate all guesses, because in some situations there is some skill in guessing: one guess gives more chance to win than another. We want to eliminate isolated stupid patterns like 3-1 or 2x2 box.
that contradicts with the idea of comparable record times. in a 2x2 box in a corner you can either count the remaining mines (slow, but best option to complete the board) or you could assume what every fast sweeper would assume and take a good guess (not going to explain how to solve it as i expect everyone here to know it already). i for example lost a sub40 this way. it's the perfect example for why lucky mode contradicts with the rankings/rules. (blast is on youtube: biggest minesweeper fails)

Cryslon wrote:
However, I'm not entirely sure, but this may well be equivalent to the problem that is well-known to be NP-complete.
Yes, it's NP-complete. But we have finite boards and we're able to find all the guess patterns and make a complete library. So we can write an efficient implementation.
do it. if you had forgotten a pattern and a sweeper blasted on exactly pattern, you better have a good excuse. as of now it's just bad luck. with lucky mode it would be YOUR FAULT.

Cryslon wrote:
And that means that the rules will have to state exactly what patterns will be recognized and what patterns won't. To explain this rule to a newbie, you need to go past the basic rules; You will need to explain details about the game that are best learnt through practice.
I don't see this as a big problem. Eg, we have 3BV limits; and you suggest ZiNi limits. These rules are much harder to explain, but you don't see this as a problem at all.
3bv, 3bv limits and zini is easy to explain. it shouldn't take more than 5min to explain it to a sweeper who has completed expert 5 times.

Cryslon wrote:
Every professional minesweeper player trades consistency for speed. We all want to finish games, so we obviously don't take this to an extreme. However, it is also obvious that in lucky mode, there is additional free consistency. This means that more other risks can be taken to get the same consistency as before lucky mode - but with better times.

The argument that gameplay would be the same is invalid, because it wouldn't be.
These sentences are in a direct contradiction with the last part of your post where you brag about how easy it is to win exp. If "professionals" you refer to are able to win exp every two minutes, then why these "professionals" aren't taking more risks to play faster? Certainly it wouldn't hurt to take three times as many risks and win exp every six minutes? Why they don't do it? I think that's because it's pretty much impossible for them to take additional risks to solve faster; they're already going at full speed.
some of these people tk is referring to are: me, manu, lasse, dion, tk, tjips, even kamil. tk also did not brag, as you wrote. to clarify the point he made: when avoiding all risks quite a few players are able to win expert every 2 or 3 minutes. of course they hardly ever play like this. read his quote of my words on irc: these sweepers are able to vary their playing style. i for example am by far not going at full speed. at full speed i might win 5 boards per week and probably could get much better records. i want to win a lot of boards though and see my normal playing as practice for tournaments.

Cryslon wrote:
Implement 3bv and/or ZiNi limits

If every sweeper has a chance to get the easiest boards even fairly often, they will have multiple attempts to solve such an easy board, and forced guesses will not matter anymore.
Huh? First of all, this is pretty much irrelevant to the topic. Secondly, tougher limits => less easy boards => rankings are effectively frozen. This is a solution for what?

And again, you seem to be concerned only with rankings representing some sort of idealized pure skill.

i completely agree with tk. it is not irrelevant - it follows his argument that despite forced guesses it is still fairly easy to complete expert. the rankings would not be frozen either. it just would not become easier to beat records. making it easier would make current records meaningless.
the question is whether lucky mode is something that can be applied to the rankings or not. being "concerned only with rankings representing some sort of idealized pure skill" is absolutely fine in this context.

Cryslon wrote:
By contrast, imagine Ian blasting his current expert highscore on a 50:50 near the end. That would be even more frustrating than trying to get a new highscore now would be futile.
For me these two sentences don't make sense.
same for me. ian finished his 33. so what? ian was/is on the same level as manu, dion, me and maybe tk and pavel. single-digit-ranked players seemingly don't care so much for their position on the ranking anymore. at least i am yet to hear a top10 player complain about ian's 33. ian did not stop playing because of his 33.
„Das perlt jetzt aber richtig über, ma sagn. Mach ma' noch'n Bier! Wie heißt das? Biddä! Bidddää! Biddddäää! Reiner Weltladen!“

Cryslon
Posts: 122
Joined: Sun Dec 28, 2008 7:41 pm

Re: Lucky mode

Post by Cryslon » Mon Apr 16, 2012 6:06 pm

EWQMinesweeper wrote:artificially increasing the chances of surviving a forced guess isclearly making it 'easier'.
Define "easier". My definition of "easier" is "requiring less skill". Obviously, lucky mode is not easier by this definition.
the consistency of the rankings should not be touched [...] a changing of the rules, as drastic as lucky mode, would ultimately have to result in a new ranking.
You haven't convinced me that allowing lucky mode makes rankings inconsistent. What kind of apocalypse do you expect if we simple start to put the new records on lucky mode into the old ranking table?
your reasoning, if i understand it correctly, is that losing on forced guesses is unpleasant.
Yes. Would you argue it is pleasant?
what i cannot understand is why there is any need for a density ranking then.
Your reasoning here is unclear. I guess you mean that since there are lots of forced guesses in density, i should think nobody wants to play it, and therefore i'm completely wrong. But the guesses in density are different from the kind of guesses i want to eliminate. Density is all about counting probabilities, and making good choices that will in case of success allow you to solve further. On the other hand, 3-1 and 2x2 is just pure stupid randomness.
also, doesn't this mean we would have to eliminate all forced guesses? (or all guesses in which the odds are worse than/equal to 50-50 and let the player win all other guesses if they take the best guess).
I'm fine with eliminating 3-1, 2x2 and derivative long tubes. I think this would be better than to leave these patterns as is.
that contradicts with the idea of comparable record times. in a 2x2 box in a corner you can either count the remaining mines (slow, but best option to complete the board) or you could assume what every fast sweeper would assume and take a good guess (not going to explain how to solve it as i expect everyone here to know it already). i for example lost a sub40 this way. it's the perfect example for why lucky mode contradicts with the rankings/rules.
How the fact that there are situations where lucky mode doesn't help show that lucky mode "contradicts with the rankings/rules"? What does the quoted phrase mean at all?
3bv, 3bv limits and zini is easy to explain. it shouldn't take more than 5min to explain it to a sweeper who has completed expert 5 times.
Thomas spoke about complete newbies. And it is easy to explain the concept of lucky mode to a not-so-complete newbie.
read his quote of my words on irc: these sweepers are able to vary their playing style. i for example am by far not going at full speed. at full speed i might win 5 boards per week and probably could get much better records. i want to win a lot of boards though and see my normal playing as practice for tournaments.
OK. Great. Would you take part in an experiment then? Play freesweeper with lucky mode enabled, on faster speed than you usually play, to see would your avg be significantly better than on a classic clone.
the rankings would not be frozen either. it just would not become easier to beat records. making it easier would make current records meaningless.
He discusses board limits in the lucky mode thread. This is pretty off-topic, isn't it? We don't discuss here any problem that is answered by introducing tougher board limits (which would make current records impossible to beat, frozen and therefore pretty meaningless).
the question is whether lucky mode is something that can be applied to the rankings or not. being "concerned only with rankings representing some sort of idealized pure skill" is absolutely fine in this context.
I will repeat: while imho the impact of the lucky mode on the rankings would be negligible (and i doubt it would exist at all), the benefit of reducing frustration is huge and obvious.
Go IRC! (try mibbit)

EWQMinesweeper
Posts: 410
Joined: Sun Nov 30, 2008 11:50 pm

Re: Lucky mode

Post by EWQMinesweeper » Mon Apr 16, 2012 7:35 pm

Cryslon wrote:
EWQMinesweeper wrote:artificially increasing the chances of surviving a forced guess isclearly making it 'easier'.
Define "easier". My definition of "easier" is "requiring less skill". Obviously, lucky mode is not easier by this definition.
by your definition giving everyone 2-30-100 3bv boards wouldn't make it easier. surviving guess and thereby games you would not always have won before means that on average you need half the time to complete a board, thus requiring less playing time.

Cryslon wrote:
the consistency of the rankings should not be touched [...] a changing of the rules, as drastic as lucky mode, would ultimately have to result in a new ranking.
You haven't convinced me that allowing lucky mode makes rankings inconsistent. What kind of apocalypse do you expect if we simple start to put the new records on lucky mode into the old ranking table?
lucky mode and non-lucky mode scores are simply not comparable. i would not use lucky mode and would consider every record played with it as invalid.

Cryslon wrote:
your reasoning, if i understand it correctly, is that losing on forced guesses is unpleasant.
Yes. Would you argue it is pleasant?
hm...i'm pretty indifferent about blasts on forced guesses, as they are not my fault.

Cryslon wrote:
what i cannot understand is why there is any need for a density ranking then.
Your reasoning here is unclear. I guess you mean that since there are lots of forced guesses in density, i should think nobody wants to play it, and therefore i'm completely wrong. But the guesses in density are different from the kind of guesses i want to eliminate. Density is all about counting probabilities, and making good choices that will in case of success allow you to solve further. On the other hand, 3-1 and 2x2 is just pure stupid randomness.
there are plenty of forced guesses in density mode in which you have no information other than that there are x safe squares in y unopened squares. it is 100% the same as in normal mode. there is as much randomness in density mode as in normal mode. just today i had an 8x8 board with 26 mines in which i was left with 2 50-50s.

Cryslon wrote:
also, doesn't this mean we would have to eliminate all forced guesses? (or all guesses in which the odds are worse than/equal to 50-50 and let the player win all other guesses if they take the best guess).
I'm fine with eliminating 3-1, 2x2 and derivative long tubes. I think this would be better than to leave these patterns as is.
it would be utterly inconsistent and you, cryslon, (and the others in favour of lucky mode) would be the one(s) to blame in case of blasts on forced guesses you did not include. do you really want that?

Cryslon wrote:
that contradicts with the idea of comparable record times. in a 2x2 box in a corner you can either count the remaining mines (slow, but best option to complete the board) or you could assume what every fast sweeper would assume and take a good guess (not going to explain how to solve it as i expect everyone here to know it already). i for example lost a sub40 this way. it's the perfect example for why lucky mode contradicts with the rankings/rules.
How the fact that there are situations where lucky mode doesn't help show that lucky mode "contradicts with the rankings/rules"? What does the quoted phrase mean at all?
lucky mode's solution is based on the number of mines in that pattern. the solution you would choose when trying to finish in a good time is different from that. the ranking lists sweepers by their fastest times, not by how many games they complete.

Cryslon wrote:
3bv, 3bv limits and zini is easy to explain. it shouldn't take more than 5min to explain it to a sweeper who has completed expert 5 times.
Thomas spoke about complete newbies. And it is easy to explain the concept of lucky mode to a not-so-complete newbie.
to explain lucky mode you needed a list of which patterns are included and which not. and you needed to explain why. 3bv and zini and very short algorithms.

Cryslon wrote:
read his quote of my words on irc: these sweepers are able to vary their playing style. i for example am by far not going at full speed. at full speed i might win 5 boards per week and probably could get much better records. i want to win a lot of boards though and see my normal playing as practice for tournaments.
OK. Great. Would you take part in an experiment then? Play freesweeper with lucky mode enabled, on faster speed than you usually play, to see would your avg be significantly better than on a classic clone.
no. for that i would have to spend a significant amount of time playing, risking that potential records would be invalid. playing for a week or two would not be long enough to show any significant differences. my minesweeping priorities are the current ranking, the active ranking, tournaments, density ranking.
instead, i suggest that you play a few hours on winmine, using sorin's recorder. it records every played game, including blasts. it would show us how many times you would actually need lucky mode and whether that number is high enough to account for a significant enough ratio of your blasts.
Cryslon wrote:
the rankings would not be frozen either. it just would not become easier to beat records. making it easier would make current records meaningless.
He discusses board limits in the lucky mode thread. This is pretty off-topic, isn't it? We don't discuss here any problem that is answered by introducing tougher board limits (which would make current records impossible to beat, frozen and therefore pretty meaningless).
players would have to get better to beat records. that's absolutely fine. the gaps among the top players are far too big. we have hardly any professionals (minesweeper has seen less than 10 such professionals so far). all players have much room for improvement. after all nobody is perfect. kamil is pretty close to perfection though.

Cryslon wrote:
the question is whether lucky mode is something that can be applied to the rankings or not. being "concerned only with rankings representing some sort of idealized pure skill" is absolutely fine in this context.
I will repeat: while imho the impact of the lucky mode on the rankings would be negligible (and i doubt it would exist at all), the benefit of reducing frustration is huge and obvious.
[/quote]

nobody is forced to play minesweeper. or forced to play using a specific version. the mentioned consistency of the ranking and comparability of records is much more important than the frustration very few players might take from a certain kind of forced guesses, which imo probably account for an insignificant amount of blasts.
„Das perlt jetzt aber richtig über, ma sagn. Mach ma' noch'n Bier! Wie heißt das? Biddä! Bidddää! Biddddäää! Reiner Weltladen!“

dg211
Posts: 12
Joined: Mon Apr 02, 2012 7:19 pm

Re: Lucky mode

Post by dg211 » Mon Apr 30, 2012 10:57 am

Hi, I meant to post a reply sooner, but I've been very busy lately with moving house, so sorry for taking so long.

The strongest argument against removing 50-50s that I have heard so far is that it allows players to alter the risk/completion balance and therefore makes it "easier" to get good times, and therefore could not be comparable to previous records. Whilst I agree that it would allow players to be more effective, like Cryslon I am not convinced that that is the same as "easier". In my mind there is an analogy here between blasts due to 50-50s etc. and injuries in sport. From time to time, an athlete may not be able to compete for weeks or even a whole season due to an injury which is fairly random and largely not their fault. In the last few decades there have been a lot of improvements in things like equipment, immediate treatment and physiotherapy which reduce the likelihood and severity of injuries. Therefore athletes can have more opportunities to get good results in competition, and presumably can alter the balance between the intensity of their training and the risk of injury to be more effective. But does that mean it is "easier" to get records now? Should we ignore Usain Bolt's times because he has some advantages in terms of equipment etc. over Jesse Owens? As [name removed]is fond of saying, the 100m is still 100m. And a lucky mode exp game is still an exp game, with all the skill-based requirements completely intact.

Tommy argues that it would reduce the simplicity and elegance of the rules, but there already are some pretty subtle minutiae to the rules and since I don't believe removing 50-50s should affect the way you play a particular board, I don't think it is a rule that needs to be explained in detail to new players, and can come under the heading of advanced rules, like 3bv limits (I think I had sub50 before I really bothered to understand 3bv). It is true that the list of forced guess to be removed might be somewhat arbitrary, but I don't believe it should be necessary to memorize a list of scenarios in order to play the game, again because removing the types of guesses that I suggest shouldn't affect the way you play a particular board.

Another counterargument is that the subset of forced guesses that would be removed is arbitrary and that other types of forced guess which are not affected would become even more frustrating. I don't really buy this, you appear to be arguing that you'd rather be kicked in the nuts 10 times completely randomly than kicked in the nuts 6 or 7 times with some of the random occurrences being arbitrarily withheld. [name removed]has also said that if some library of guesses is removed, then the author of the library would be to blame for blasts on guesses not including in the library, which to me is completely nonsensical. You might as well argue that people who stubbornly refuse to countenance the removal of 50-50s are to blame for every 50-50 blast.

One other thing that occurs to me - we have recently been talking a lot about trying to attract more new players to the community. Which of the following statements do you think will be more appealing to new players? (1) If you've been playing for a while, you've probably noticed that sometimes there is not enough information to uniquely determine whether a particular square is safe, and you have no alternative but to guess. Obviously this has nothing to do with your skill, and there are clones which remove some of these non skill related blasts. OR (2) You've probably noticed sometimes you have to guess. No matter how much you practise or how skillful you become, more than 1/3 of expert boards will screw you through no fault of your own. We have ways to remove some of these non skill related blasts, but we decided not to implement them because they're just part of the game we love, so suck it up!

User avatar
Tommy
Posts: 223
Joined: Mon Dec 01, 2008 9:22 pm
Location: Vienna

Re: Lucky mode

Post by Tommy » Mon Apr 30, 2012 10:45 pm

dg211 wrote: The strongest argument against removing 50-50s that I have heard so far is that it allows players to alter the risk/completion balance and therefore makes it "easier" to get good times, and therefore could not be comparable to previous records. Whilst I agree that it would allow players to be more effective, like Cryslon I am not convinced that that is the same as "easier". In my mind there is an analogy here between blasts due to 50-50s etc. and injuries in sport. From time to time, an athlete may not be able to compete for weeks or even a whole season due to an injury which is fairly random and largely not their fault. In the last few decades there have been a lot of improvements in things like equipment, immediate treatment and physiotherapy which reduce the likelihood and severity of injuries. Therefore athletes can have more opportunities to get good results in competition, and presumably can alter the balance between the intensity of their training and the risk of injury to be more effective. But does that mean it is "easier" to get records now? Should we ignore Usain Bolt's times because he has some advantages in terms of equipment etc. over Jesse Owens? As [name removed]is fond of saying, the 100m is still 100m. And a lucky mode exp game is still an exp game, with all the skill-based requirements completely intact.
We have an equipment advantage nowadays, we have great mice/mousepads. But I personally expect equipment changes to only positively affect the degree of perfection with which actions are transformed into results. For instance, with mouse precision, pointer position corresponds better to mouse position. Better running shoes may be better at transforming all the kinetic energy from muscles into forward motion.

Anyway, the two are not related at all. Usain Bolt has better equipment than Jesse Owens, and that is OK. But as far as I know, Usain Bolt does not risk his life more or less because injuries can be treated better nowadays. He doesn't say "OK, I can abuse my body 30% more than Jesse Owens, because I'll have approximately the same chance of a fatality or serious injury over the next 3 years that way, but better stats". There is an important difference between metagame changes and equipment improvement, and they are distinct even in the example you give. The reason why metagame changes are acceptable in this case is that availability of medical treatment is more important than running stats, it can be crucial to survival in real life. Also, people are expected not to abuse their bodies to be competitive - this is why doping is not allowed. So, availability of medical treatment does not really play a role in the running metagame anyway.
Tommy argues that it would reduce the simplicity and elegance of the rules, but there already are some pretty subtle minutiae to the rules and since I don't believe removing 50-50s should affect the way you play a particular board, I don't think it is a rule that needs to be explained in detail to new players, and can come under the heading of advanced rules, like 3bv limits (I think I had sub50 before I really bothered to understand 3bv). It is true that the list of forced guess to be removed might be somewhat arbitrary, but I don't believe it should be necessary to memorize a list of scenarios in order to play the game, again because removing the types of guesses that I suggest shouldn't affect the way you play a particular board.

Another counterargument is that the subset of forced guesses that would be removed is arbitrary and that other types of forced guess which are not affected would become even more frustrating. I don't really buy this, you appear to be arguing that you'd rather be kicked in the nuts 10 times completely randomly than kicked in the nuts 6 or 7 times with some of the random occurrences being arbitrarily withheld. [name removed]has also said that if some library of guesses is removed, then the author of the library would be to blame for blasts on guesses not including in the library, which to me is completely nonsensical. You might as well argue that people who stubbornly refuse to countenance the removal of 50-50s are to blame for every 50-50 blast.
The rules at the moment are a paragraph of intuitive text, not an arbitrary list that one would have to learn by heart. Yes, I want to know the precise rules of the game I play, and not only that, I want them to be in my blood. This is the case now, and it will never be if I have to remember which patterns are part of the game now, and which aren't. Especially if these patterns are subject to change.
One other thing that occurs to me - we have recently been talking a lot about trying to attract more new players to the community. Which of the following statements do you think will be more appealing to new players? (1) If you've been playing for a while, you've probably noticed that sometimes there is not enough information to uniquely determine whether a particular square is safe, and you have no alternative but to guess. Obviously this has nothing to do with your skill, and there are clones which remove some of these non skill related blasts. OR (2) You've probably noticed sometimes you have to guess. No matter how much you practise or how skillful you become, more than 1/3 of expert boards will screw you through no fault of your own. We have ways to remove some of these non skill related blasts, but we decided not to implement them because they're just part of the game we love, so suck it up!
Not "suck it up", "change your mindset". It isn't single games that matter in minesweeper, even if the ranking seemingly says so. It's your ability to perform that will ultimately result in highscores. Bearing that in mind - I recently blasted a pretty epic exp game:
http://minesweeper.info/forum/viewtopic ... =182#p1263
The obvious reaction is, of course, to go into "fuck this shit" mode and stop sweeping for a week. A much better way to think of it is: "Hey. I almost got a fucking 35. And it is not my fault that I didn't get it; I now have definitive proof that I have the ability to get one. So, since I'm not going to stop playing, I'm probably going to get one at some point - maybe even something better". In other words, don't live in the moment; Always bear in mind that luck evens out over time. Of course, nobody gets rid of the natural frustration completely. But it has stopped affecting me, and it is possible for other people to get into that mindset, as well.

The thing to say is:
"You've probably noticed sometimes you have to guess. This is especially frustrating in the beginning - as you get better, you finish more games, simply by virtue of being faster, and individual losses become unimportant. And at some point, you just stop caring about blasts that are not your fault."

It may be the case that many good sweepers don't think that way. Let me be honest: You're doing it wrong. Care about personal failure, about blasts due to misclicks. But if the game gives you a situation where you need to guess, and you lose, there is no reason - whatsoever - to blame yourself. Just stop doing it and play the next game.
Don't anthropomorphize computers - they don't like it.

qqwref
Posts: 121
Joined: Thu Sep 23, 2010 4:17 pm

Re: Lucky mode

Post by qqwref » Tue May 01, 2012 2:00 am

Tommy wrote:But as far as I know, Usain Bolt does not risk his life more or less because injuries can be treated better nowadays. He doesn't say "OK, I can abuse my body 30% more than Jesse Owens, because I'll have approximately the same chance of a fatality or serious injury over the next 3 years that way, but better stats".
Maybe not in this case... but I've heard that American Football players take more hard hits on the head than they used to, because of the helmets causing a feeling of safety. From here: "these helmets ... also created a sense of invulnerability that encouraged players to collide more forcefully and more often." So that would be the case of exactly this happening, players risking their life more because they know/think they are safer.
Tommy wrote:In other words, don't live in the moment; Always bear in mind that luck evens out over time. Of course, nobody gets rid of the natural frustration completely. But it has stopped affecting me, and it is possible for other people to get into that mindset, as well.
I do agree that luck is part of MS and always will be, but I think having less randomness that the player can't affect is a good thing. It's not all or nothing - and for someone who plays rarely, that 50-50 might be the difference between getting that first sub60 in 2012 or in 2013 (or never). "Don't worry about it" is a great mindset for an individual, but it shouldn't stop the community from fixing a problem that bothers some (but not all) players.
Tommy wrote:if the game gives you a situation where you need to guess, and you lose, there is no reason - whatsoever - to blame yourself
Who ever said they blame themselves for a missed 50-50? I don't think anyone who understands the pattern will look at that and go "that was my fault, I suck". And yet, it's still frustrating... I'd say it's frustrating precisely because it ISN'T your fault, and is instead just the program screwing you out of an otherwise perfectly good game.
NF player. Best scores 1-10-39.

User avatar
Tommy
Posts: 223
Joined: Mon Dec 01, 2008 9:22 pm
Location: Vienna

Re: Lucky mode

Post by Tommy » Tue May 01, 2012 6:40 am

qqwref wrote:
Tommy wrote:But as far as I know, Usain Bolt does not risk his life more or less because injuries can be treated better nowadays. He doesn't say "OK, I can abuse my body 30% more than Jesse Owens, because I'll have approximately the same chance of a fatality or serious injury over the next 3 years that way, but better stats".
Maybe not in this case... but I've heard that American Football players take more hard hits on the head than they used to, because of the helmets causing a feeling of safety. From here: "these helmets ... also created a sense of invulnerability that encouraged players to collide more forcefully and more often." So that would be the case of exactly this happening, players risking their life more because they know/think they are safer.
But football is a sport that pits teams against each other in a match, as opposed to individuals (or teams) getting scores that are compared later. So changes hurt much less, because you don't have a definitive score to compare after a match anyway - performance depends on the other team as much as anything else. RTS games are patched all the time, and it doesn't matter, because there is nothing to compare but the players participating in a match right then.
Tommy wrote:In other words, don't live in the moment; Always bear in mind that luck evens out over time. Of course, nobody gets rid of the natural frustration completely. But it has stopped affecting me, and it is possible for other people to get into that mindset, as well.
I do agree that luck is part of MS and always will be, but I think having less randomness that the player can't affect is a good thing. It's not all or nothing - and for someone who plays rarely, that 50-50 might be the difference between getting that first sub60 in 2012 or in 2013 (or never). "Don't worry about it" is a great mindset for an individual, but it shouldn't stop the community from fixing a problem that bothers some (but not all) players.
But a solution that isn't all-or-nothing is very likely to be ugly. If it involves lists of patterns, it is very ugly indeed. Yes, of course less randomness that the player can't affect is a good thing. But no, I don't think that the price is remotely adequate here. And we are not designing a game from scratch here, we have something to work with that we can't significantly alter.
As stated before - stricter limits would solve this problem anyway.
Tommy wrote:if the game gives you a situation where you need to guess, and you lose, there is no reason - whatsoever - to blame yourself
Who ever said they blame themselves for a missed 50-50? I don't think anyone who understands the pattern will look at that and go "that was my fault, I suck". And yet, it's still frustrating... I'd say it's frustrating precisely because it ISN'T your fault, and is instead just the program screwing you out of an otherwise perfectly good game.
In any case - I think that that is one of the fundamentals of the game. A single game is worth the time you invest to play it. Even if you get far in the board, if you only play for an hour and get sub120es consistently, and then you blast at, say, 110 seconds - you have only lost a very small fraction of your playing time to that game - about 3%. If you invest a disproportionate amount of emotions into that game, it's your fault.

I will go so far as to say that it is incorrect to have this mindset from a professional point of view (and disadvantageous from a casual point of view). It is emotionally draining and will make the game less enjoyable.
Don't anthropomorphize computers - they don't like it.

qqwref
Posts: 121
Joined: Thu Sep 23, 2010 4:17 pm

Re: Lucky mode

Post by qqwref » Tue May 01, 2012 5:37 pm

Tommy wrote:But football is a sport that pits teams against each other in a match, as opposed to individuals (or teams) getting scores that are compared later. So changes hurt much less, because you don't have a definitive score to compare after a match anyway
That's true, but it's still an instance of people taking more risks because they are more safe. Arguably, another example is cubing. In the 1980s the physical technology of cubes meant that people had to make turns by applying relatively large amounts of force with their wrist, and many people developed injuries by doing this (look up "Cuber's thumb" or "Rubik's wrist"). With modern cubes there is much less of a risk of this because of smoother turning, so as a result people are turning much faster and spending more time solving than they did in the 80s.

Anyway, as far as MS goes, I don't think spending less time to achieve a good time is a problem if it takes the same amount of skill to achieve that time. Playing a lot is not a skill, unless we're talking about playing at a high level for many hours at a stretch. It seems like people are suggesting that players somehow tailor the amount of risk they take so they can complete the desired amount of games, which would mean that less risk from the game would mean they can introduce more risk into their style. However, I think it's more likely that players add a certain amount of risk to the game depending on how fast they want their games to be, whether it is to complete a lot of games or to go for personal records, or anywhere in between (e.g. tournaments). For some very experienced players (EWQ?) there may be no difference in risk between the two; nevertheless, I would say that with 50-50s mostly removed the vast majority of players will play the same as they did before and just notice that they finish more games.
Tommy wrote:If you invest a disproportionate amount of emotions into that game, it's your fault.
So it's wrong to be really unhappy about a potential record, but really happy about an actual record? That doesn't seem very consistent.
Tommy wrote:I will go so far as to say that it is incorrect to have this mindset from a professional point of view (and disadvantageous from a casual point of view). It is emotionally draining and will make the game less enjoyable.
Again, this isn't an argument about how an individual should react to playing MS. I agree with you that it's no fun to be frustrated. But as a community we must recognize that some people WILL be frustrated by certain parts of the game, and that that is a valid argument for trying to improve those parts.
NF player. Best scores 1-10-39.

EWQMinesweeper
Posts: 410
Joined: Sun Nov 30, 2008 11:50 pm

Re: Lucky mode

Post by EWQMinesweeper » Tue May 01, 2012 6:29 pm

qqwref wrote:Arguably, another example is cubing.
you have read most articles from the wiki, like every dedicated sweeper should have done, have you? carpal tunnel syndrome and other injuries to the hand, wrist and arm were and still are not uncommon among the elite sweepers.
qqwref wrote:Anyway, as far as MS goes, I don't think spending less time to achieve a good time is a problem if it takes the same amount of skill to achieve that time.
skill is very loosely defined here. being sort of omni-present in the msinfo community i suppose i would be able to come up with a sorted list of 100-200 sweepers, based on their ms skills. but we just do not have any objective criteria for skill.
qqwref wrote:Playing a lot is not a skill
it is! the more you play, the more you practice, the better you should be in the end. it's a world ranking, not a kindergarten. everybody is free to play as much or as little as they want.
qqwref wrote:t seems like people are suggesting that players somehow tailor the amount of risk they take so they can complete the desired amount of games, which would mean that less risk from the game would mean they can introduce more risk into their style.
and this is part of the problem: the elite players would profit a lot, as they blast more often on forced guesses than on misclicks.
qqwref wrote:However, I think it's more likely that players add a certain amount of risk to the game depending on how fast they want their games to be, whether it is to complete a lot of games or to go for personal records, or anywhere in between (e.g. tournaments).
normal sweepers wouldn't even notice it as they make too many mistakes or do not know enough patterns.
qqwref wrote:
Tommy wrote:If you invest a disproportionate amount of emotions into that game, it's your fault.
So it's wrong to be really unhappy about a potential record, but really happy about an actual record? That doesn't seem very consistent.
playing this game is supposed to be fun. taking frustration from forced guesses seems irrational to me. when i lose a forced guess i'm not frustrated, just like when i win a game in which i made it through forced guesses i am indifferent about the guesses, but happy that i solved the rest.
qqwref wrote:Again, this isn't an argument about how an individual should react to playing MS. I agree with you that it's no fun to be frustrated. But as a community we must recognize that some people WILL be frustrated by certain parts of the game, and that that is a valid argument for trying to improve those parts.
"119 3bv boards for everyone!" (if you don't get this, you are not familiar enough with msinfo.) some people will always be crying about irrational and subjective parts. there is no objective evidence that blasts on isolated 50-50s account for a significant enough ration of blasts in a normal sweeper's game. again: only very good swepers will actually profit from lucky mode, whereas there won't be any difference for the rest.
„Das perlt jetzt aber richtig über, ma sagn. Mach ma' noch'n Bier! Wie heißt das? Biddä! Bidddää! Biddddäää! Reiner Weltladen!“

qqwref
Posts: 121
Joined: Thu Sep 23, 2010 4:17 pm

Re: Lucky mode

Post by qqwref » Tue May 01, 2012 8:21 pm

EWQMinesweeper wrote:carpal tunnel syndrome and other injuries to the hand, wrist and arm were and still are not uncommon among the elite sweepers.
Yeah, and? I was talking about a situation where making something safer did, in fact, lead to people pushing themselves harder than they used to. Just because MS players have gotten repetitive stress injuries doesn't mean there is a meaningful analogy to be made.
EWQMinesweeper wrote:
qqwref wrote:Playing a lot is not a skill
it is! the more you play, the more you practice, the better you should be in the end. it's a world ranking, not a kindergarten. everybody is free to play as much or as little as they want.
You're taking my words out of context and changing the topic.
EWQMinesweeper wrote:the elite players would profit a lot, as they blast more often on forced guesses than on misclicks.
Getting better times is GOOD if you deserve them. I don't know why this keeps getting suggested as a problem. If you have the skill to get a 33 on a board of a given difficulty, and then you get a board of that difficulty and achieve a 33, I see nothing wrong there.
EWQMinesweeper wrote:there is no objective evidence that blasts on isolated 50-50s account for a significant enough ration of blasts in a normal sweeper's game.
Excuse me? We've already shown, mathematically, that a player will blast on a common 50-50 pattern on 30-something percent of boards. This is objective evidence, and you can mathematically show that out of every 100 boards you might potentially complete, you will only actually complete 70 or so because of bad 50-50 guesses. Finishing about 3/2 as many games is definitely significant for a normal player.
EWQMinesweeper wrote:only very good swepers will actually profit from lucky mode, whereas there won't be any difference for the rest.
So? Only very good sweepers profit from plenty of clone features: 3BV calculation, cl/s displays, mouse path, correctness analysis, slowed-down replay playback, daily average times, etc. This is clearly not a valid argument against including a feature.
NF player. Best scores 1-10-39.

EWQMinesweeper
Posts: 410
Joined: Sun Nov 30, 2008 11:50 pm

Re: Lucky mode

Post by EWQMinesweeper » Tue May 01, 2012 8:44 pm

qqwref wrote:Yeah, and? I was talking about a situation where making something safer did, in fact, lead to people pushing themselves harder than they used to. Just because MS players have gotten repetitive stress injuries doesn't mean there is a meaningful analogy to be made.
better mice allow current players to play much more, yet we still play more than we should.
qqwref wrote:You're taking my words out of context and changing the topic.
some people seeming keep forgetting it. that's why i repeat it every now and then.
qqwref wrote:Getting better times is GOOD if you deserve them. I don't know why this keeps getting suggested as a problem. If you have the skill to get a 33 on a board of a given difficulty, and then you get a board of that difficulty and achieve a 33, I see nothing wrong there.
please, please do not talk about whether you think a certain score is deserved or not. telling whether a certain record is deserved or not is so vague that even manu, tommy and i would struggle in every other case.
qqwref wrote:Excuse me? We've already shown, mathematically, that a player will blast on a common 50-50 pattern on 30-something percent of boards. This is objective evidence, and you can mathematically show that out of every 100 boards you might potentially complete, you will only actually complete 70 or so because of bad 50-50 guesses. Finishing about 3/2 as many games is definitely significant for a normal player.
only looking at all boards isn't enough here. you willingly ignore that depending on where you start you might blast, that sweepers make mistakes and do not see patterns. this probably accounts for the vast majority of blasts.
qqwref wrote:So? Only very good sweepers profit from plenty of clone features: 3BV calculation, cl/s displays, mouse path, correctness analysis, slowed-down replay playback, daily average times, etc. This is clearly not a valid argument against including a feature.
pretty wrong. 3bv calculation and mousepath are easy to understand, cl's displays is something that afaik only i really use (in the sense that i can play many different solving styles at different speeds), correctness is easy as well, slowed-down replays is nothing that top sweepers should need and is often used by slower sweepers. daily averages are pointless.
„Das perlt jetzt aber richtig über, ma sagn. Mach ma' noch'n Bier! Wie heißt das? Biddä! Bidddää! Biddddäää! Reiner Weltladen!“

Post Reply