@Dennis: i hope Denmark wins...but i think the result will be 2-2

Dennis, "just beat Sweden". I didn't really understand.
Anyway if you Danes are lucky to tie the game then you can still go to the quarters.
Happy sweeping and happy soccer watching.
BTW: Zlatan rules big time

oups.. of course, Sweden-Latvia is still possible, but I can't imagine this to happen.

@Dennis: If Ibrahimovic can keep his performence you shouldn't think of beating Sweden. But.. I'd like to see a 3:3.
Either Sweden-Netherlands or Sweden-Germany would be a great quaterfinal.

2.53 3BV/s ---> 2.57 3BV/s

Nothing that special, only moderately exciting.

Congratulations to all record breakers/those tieing records!

Good luck all!

MS 13,86 --> Clone 13,30

I got 13,30 on clone after first "normal" game of russian team. I was pleased that russian team won at least one match, i got my best int result 5 minutes after match ended. 3BV 35, 3BV/s 2,85.

http://www.gammel.narod.ru/13-2.mvf - video.

P.S. Congratulations to Oli! First place in AR is magnificent result. Congrats to Sorin too, maybe you'll repeat Oli's achievement.

49x2

Nice to hear from you Owen! 6 hours in 3 years . I remember the times when we used to compete in the AR...good old days!

I got my second 49 today (with videos this time). Nothing fancy, just a 124 3BV, but I'm happy for it anyway.

Congratulations to Oliver for achieving first place in the AR!

YEAH!!! Portugal won over Spain! 1-0 Now I would have liked to see BOTH Spain and Portugal in the quarter final but since I'm a Portugese fan I was simply hoping Portugal could beat those Spanish guys! I still don't get why Morientes didn't get more time but hey, better for Portugal and for me! Now Denmark just needs to beat Sweden!

Oh yeah, and congratz to Otto for being the first (maybe only) German qualified for the quarters.

LOL, I like that scenario, but actually I'd prefer:
10-0
1-0

@Roelof: I hope so, but I'm not really sure. Maybe one of the Czech bundesliga players (like Koller, Rosicky, Tyce, Lokvenc) wants to see Germany in the quarter final and helps us by making some goal... My favorite scenario would be - 3-2 and - 0-0. Would be ridiculuous to become 2nd with such a performance...

I made another 67 on expert....when will i ever beat my record? (which is still 65)

52

12

2

Hey everyone...Havent post here in a few months now. Just thought I'd drop by to see how things are. I see Lasse still holds the world record (as I suspected). Only he can sub 40 I feel.

I met another of Ireland best sweepers recently - Daniel Lynch. I'm sure you're all familiar with him posting here. He's a sound chap. Had a game of chess or two with him

As for Minesweeper, I've played about 6 hours in last 2-3 years. I played an hour the other day and got 2-16-69 which I might keep for my active ranking scores....

Good luck all

Owen

56

16

1

Hey Stephan,

Do you think Germany is able to beat the Czech reserves?

Regards,

Roelof

LOL Stephan! You really had me laughing out loud over that one! Oh, and good luck becoming the next German on top!

Congrats to Oliver Scheer for being the first German to get on the top of the AR! At least one ranking where Germany is in top position (this might be due to the lack of Czech, Latvian, and Dutch people in the AR...).

Active Ranking has been updated.

57x4- > 57x5

Great, and 20min before that I got a 58....!
Btw the 57 was my 30th sub60.....

71->64

16

2

Perhaps my total is now worth mentioning. 82 sounds quite good to be honest.
1-14-58 should be reached this year!

oops, I forgot to send in my scores (17-70).. any chance of them still being put in since I mentioned them here just over a week ago?

63

16

AR times i think today is the deadline (i don't recall seeing it announced though)

3.32 --> 3.43 3bv/s

123 (augh)

33

5

I would work more on my scores, but I can't escape the nagging feeling that I'm - well, wasting my time. How will this help me in life again? Oh well, it's fun - and it takes time away from studying for SATs. (and how will that help in life, either?) Anyway, still far away from my goal of beating JK Rowling in expert (she's 101) and far away from anything special. Love this game.

probably not... other than the obvious "stop the timer, screenshot that, move the bits around to where they were the first time"

67

16

2

My computer crashed while pasting the screenshot of my new beginner best time onto MS Photo Editor. Is there any way to retrieve that screnshot?

Thanks.

78=>77

23=>22

2

Hi
congrats to all recent record-breakers
I am 101 in total now, with exactly the same scores as Michael Reption before he got his 18 on int
Maybe I'll go the same way, but I think improvement on expert is easier for me
Cya

77 --> 77 x 3

Hurray indeed! Although I am a bit annoyed... having got three more sub-80s in the last two days, bringing my total to eight, I think I'm owed a sub-75 round about now.

Hurray, the guestbook's back.

2.64 3BV/s ---> 2.89 3BV/s

Wooow!! That one was really an impressive jump!

32.89 on a 92 3BV board! And again, playing NF!

2,70 --> 2,71 3bv/sec

i made it on a 237 3bv board.
a little step..

And what's really ironic about it is that the first personal best time I ever made was also an 18.

That was on Beginner, of course.

77

22 --> 18 (total 101 --> 97)

2

I DID IT I DID IT I DID IT I DID IT!!!!!

http://users.ox.ac.uk/~ball1418/minesweep18.jpg

Guys, I'd REALLY like to be added to the Best Ever Rankings now. It is still sub-100 to qualify, right? And also... Detrusor's site has the Best Evers grouped by country, and if I got this right, I am actually the FIRST African player (South African, to be precise) to make sub-100... a truly historic moment....

I shall now go away and be happy for a VERY long time.

If you need my Expert screenshot, change the "18" in the above URL to "77".

3bv/s 2.71 -> 2.86

On a 107 board

If this were Quake, you'd be dead now.

Oh my god, Elmar and I just posted at the EXACT same time. Cool.

@Lloyed: The bell curve you're talking about is the normal distribution Roman refers to.
@Roman: Which distribution gives the best fit depends on the database you assume, logic. If you base your calculations on all games, that database would be asymetric, correct. However if you want to make predictions of a player's current probability to beat his record you shouldn't base it on his total database. IMO the current distribution should always be symetric.

Congrats Gustaf and Jan! Nice work!

19 -> 18

18.81 in NF (I always use NF for the moment)
3BV of 45 -> 2.53 3BV/s
My second sub20
Simply happy!

59 -> 54

WOW, big jump, isn't it?
The video will be on my page later today.
Very happy sweeping!

@Lloyd: I know rule about 3 standard deviations, but for your stats it means probability 0,997 for interval (13.49;52,25) - it's useless information. As you write average of distribution migrates to lower times, this fact means that we have asymmetrical distribution and normal distribution is inapplicable.

To finish off my last thoughts about changing times, my average on the clone stats page is 32.87 on int, but that takes into account times from over a month ago. For the last 550 games or so, my average was less than 30, meaning a curve plotted with ALL of my games would not be an accurate depiction of my current playing. It would be more accurate to use the last week or last few hundred games for such a plot.

Another problem that would arise from that suggestion of using the last few weeks, or last few hundred games is that such graphs become more accurate with more data points (i.e., games finished) but the more games you finish, the better you should be getting, meaning your average should be going down as you play more and more games. Still, it should be able to be used for a very rough approximation, which would be more accurate for the most recent data.

Those are just some of my thoughts on the matter. Feel free to agree or disagree with any or all of my points. Later tonight, I'll try to make a very short post showing the formula to calculate probability of a certain time based upon average time and standard deviation.

Hey, I saw my name mentioned! It appears my many many games of intermediate are being of some use to someone at least. I thought I'd also post on my thoughts of probability.

From what little I remember (or learned) about probability in college thus far, I would take the minesweeper stats to be more of a bell curve (I hope that's the right one). There will be a central point where the curve lies on, and the curve to both sides of that would be fairly symetrical. That high point would most likely be the average time for that individual person. The rest of the curve would represent the probability that person has of getting that time, based on standard deviations or so. In my case, 19 sec would fall at the very lowest end of the curve, while 40 or so would be at the high end, with my average being right around 30 or so (give or take, but currently it is just under 30 for the games I played last week). If you had all my times, or at least a significant number of my times, you could calculate the average time, and the standard deviation. The formula for calculating the probability of a specific time, based on the average and the standard deviations is relatively simple, but I can't remember it right now (I'll try to post it later when, when I have access to a calculator).

If my memory serve me correctly, a time which is 3 standard deviations or further away from the center has a 5% probability of occuring. 4 or more standard deviations represents about 1% chance, while 2 deviations represents about an 83% chance, if I remember correctly.

So basically, I think minesweeper times would follow that kind of a curve (which looks a lot like the graph of my int time details). The only problem with the idea is that, in theory, most of us are getting better the more we play, so as time goes by, the averages will migrate towards lower times, and the average itself would not accurately represent the "current" average. (more to come...)

2.62 3BV/s ---> 2.64 3BV/s

27.85 seconds on a 71 3BV board... 0.02 3BV/s of improvement, but it is nice, considering it was playing NF!

By the way, what are the fastest speeds achieved on intermediate by the great non-flaggers here (Stephan, Martin, etc) ? Do you have the videos? I'd like to take a look. Send me if you can.

@Roman - I must admit that I've never studied the gamma distribution, but I'll have a look at it later

Sorry I got your name wrong before, btw

@Dave: normal distribution gives more accurate interpolation near average of distribution, but for "unstable" players (big dispersion) probability to get results far from average is still high. Example (based on Lloyd Rhoads stats - it's just biggest stats we have) (probability to get sub13)x(number of games)=1,3, but his best is 19,45. Gamma distribution gives this value approximately 0,0001.

@ Rod:

Whew! I'm glad that even someone from Brazil doesn't know what it is. Doesn't make me seem so bad all of a sudden.

I must say, though, that I was out playing pool last night, and they had some of the Russia/Spain game on the tv there. Of course, it had already finished hours ago, but they were playing it anyway. I guess there must be enough interest in the game to put it on tv, but I think it was on a specialty channel.

Well, maybe soccer is more popular here than I thought, but that still doesn't mean that I have to like it or pay attention to it!

Bye everyone, have fun watching soccer!

Romman - I'm not convinced that that would work, since you'd hope that the mean of the distribution would decrease with time

I'd have thought a normal distribution would at least be a decent first approximation though (lots of times clustered around an ok sort of time, with less and less scores as you go further away from that)

Hey, Dave and Manu: don't worry for not knowing what the Euro2004 is! I also asked what it is on the IRC channel and received some slaps there, because of it!

P.S.: I'm from Brazil (!!! )

Sorry Dave for ignoring that you are canadian. I was maybe too sure that only an american could say something like that !
Thing is that even someone who is not interested in soccer at all MUST know what is the Euro : the higher level of competition that we can expect for soccer.

For non-mathematicians: with good interpolation you can find probability to get better result, so you can count approximate time you need to beat your record

time less than 10s is 0 in such interpolation). This way shows good results, but maybe somebody knows better method. Join to irc-channel, we could discuss this problem there.

P.S. I'm upset, because our team (Russia) lost it's first match.

There is a question to players who knows theory of probability: what probability distribution should we use to get better interpolation of distribution of game time on int?
At first I counted average of distribution and dispersion, after I used gamma distribution with the same average and dispersion (but with dislodged beginning to 10 seconds, so probability to get

79

19

1

Hop Suisse Hop Suisse!


I'm happy with only 1 match win!

Speak for yourself, Dave... Soccer took over the number one most popular sport in Canada a few years ago, over hockey. There are tons of Canadians (and lots of Americans as well) that like soccer a lot

I meant to say that very few North Americans LIKE soccoer, not "don't like".

@ Manu:

I'm actually Canadian, as Katy said, and I think you missed an option......

4. Maybe I don't like soccer. In fact, very few North Americans don't like soccer. I personally don't like it, but maybe I don't know what I'm missing!?

so, us americans have some sort of liscense to be ignorant about the rest of the world, and apparantly the rest of the world has a liscense to be ignorant about canadians. poor canadians.

I'm afraid he isn't kidding. In another forum an american bet that brasil would win.

Dave > there are very few possibilities :
1) You just got a terrible accident and forgot everything.
2) You're trying to make a joke.
3) You're american.

What is Euro2004? Is it a soccer thing?

6x23!!!

Hi
I can guarantee that Germany won't win!!
My favorites are France, Spain and Italy.
Happy sweeping

I think the order of which you put the flags in your first post, Elmar, shows which positions the countries will get (from left to right of course )
The order of the second post isn't too bad either though

Sorry, forgot the Czech Republique....

@Detrusor: Bravo! You really have to slow down, I remember you braking into the sub100 total not long ago.

And now to something completley different:
Who will win the EURO2004 starting today?

Of course I hope Germany will get far. I know they suck at the moment, but then again, they always do.....

I do hope France makes it to the quarters because then my town will set up a giant screen on the square close to where I live.

My vote would go for Portugal. With their team and the support of the hosting audience they should have a good chance. We'll know more as of tonight when they play against Greece

Anyways, enjoy the EURO2004 everyone!

62.12

17.56 -> 15.81 !!!

2.24

My dear friends, i announce you a BIG record !!! I'm very very happy....
I was discussing with Elmar in msn, he has sent me a 29 3BV board on which i trained (i acheived 9 secs)..Then i turn the "legal game" on, i played some games... and wooowww a beautiful "enter your name" poped up ! i made a 15.81 seconds on a 34 3BV board... My previous record was 17.56 sec so i'm very very happy... I enter the sub80 in the total, so it's a very good day !

You can see the video by clicking on the video link there : http://www.planete-demineur.com/french/sweepers.php?mode=viewprofile&id=1

Congrats Rod! Soon enough you'll be sub-20!!!

Also, congrats to Oli!!! Great work!

20 ----> 20 x 2!!

Oh my God, what is happening to me?!?!?!?
I sincerely thought my previous 20 was pure luck. The 3BV was 34, etc. But I couldn't believe in my eyes when I saw ANOTHER 20 in the timer this time!! And in a 41 3BV board! Yet better! And I even didn't play intermediate for the last 3 days, I was concentrated in expert!! So cool!

@Michael: hum... that's right. In the case you mention, it IS a case of UPK, but my definition doesn't cover this situation, in wich the final guess is on green squares, not in red. Maybe I have to think a little bit more about this!

@Jon: try to e-mail me your history file to rodrigocamargo@escelsa.com.br. I can remove the buggy game from it to you. And convert your history file, too.

Congrats Oli
Seems like some of you guys are passing me... I really must be doing something about that soon!

Hey, I just said d a m n , it's not like I really swore! Oh well.

Nice James! Keep at 'er and you'll be sub-90 this period I bet! I'm sitting at 87/24 so far this period, but I'd like to lower my exp by about 3 or 4, and I'd like to lower my int by 1 or 2. I have to, to keep up with Rod this period!!!

**** you Rod!!!

Have fun!

Clone 98 ---> 94

That's for you Dave. First sub-100 in 2 weeks, but I have been learning and getting more consistent. I am pretty sure I can improve it before the period is over.

OKay, let me ammend that last statement. When I say that if the KNOW a mine is in a certain location (regardless of how they know it), that is not entirely correct. If they know that it will be in a certain location by looking at the opened squares, that is obviously okay. Just in case anyone called me on that.

Christoph: I tend to agree. In my opinion, "Unfair Prior Knowledge" is just that......prior knowledge of any unopened squares. If someone KNOWS that a mine will be in a certain location (regardless of how they know it) then it is not a fair game. In my opinion, it is that simple.

I don't think that Rodrigos definition of UPK is complete. What do think about that:
UPK is every knowloge about the board that cannot be achived by analysing the opened squares. That includes locations of mines, locations of free squares and knowloge of the best mousepath.

Congrats, Oli! Well done!!

@Rodrigo: I'm still having trouble sending you mail. I get the following message:

-----------------------------------------------------------
This is the Postfix program at host shanks7.bol.com.br.

I'm sorry to have to inform you that the message returned
below could not be delivered to one or more destinations.

For further assistance, please send mail to

If you do so, please include this problem report. You can
delete your own text from the message returned below.

The Postfix program

: host sidney11.bol.com.br[200.221.24.152] said:
552 RCPT TO: Mailbox disk quota exceeded (in
reply to RCPT TO command)

------------------------------------------------------------
Doesn't this mean that there is not enough free space in the mailbox? If you intend to keep it that way, I would like to know if you accept other delivery methods. I could perhaps upload the files to my home site, so you can download them from there instead?

BTW: Is there a possibility to personalize the history-files, so that noone can "steal" your file and publish it as their own?

If you want, you can see his video by clicking on the video link on this page : http://www.planete-demineur.com/french/sweepers.php?mode=viewprofile&id=6

@Jon
Benoit Duffez got this 1.00 on a version prior to 0.92, and he got the error message when trying to open his history.
You can see on my site that his 3BV/s is 0, because it has been manually set as 0...

BTW: I see that Benoit Duffez has a 1.00 beginner game, and his statistics seems to be OK. I don't know if he got it on the 0.92 version, though. I should add that besides a time of 1.00, I THINK my 3BV counter showed 0.

@Rodrigo: I discovered a bug in the 0.92 version of the clone yesterday. I had just unzipped it, and looked through some of the options (history, auto-recording etc.) Then I started playing on beginner. The first board was not an easy one, but it seemed to me that the timer did not start counting. When I finished the board after several seconds, the pop-up box appeared with a time of 1.00 seconds. This only happened on my first game (maybe it could be because I had not hit F2 yet?). If you need more information to pin down the error, I could try to reconstruct the situation. I have a video of the game, but the counter works fine there. After that game I got an error message that said something about "division by zero" when I tried viewing the history file. I don't know if that problem could occur when someone finish a board in one click, too. It would be unfortunate if someone get their history file messed up because they get a 1.00 time.

@James
You say that 87.5% of the mines can be fairly revealed... People who practiced the DB will necessarily and exclusively make the best choice for their double clicks. I think that even a good player, but who didn't practice on it, can't make during ALL the game, the best decision, especially for the alternance double clicks / NF. So, like Gergely said, there are on videos some signs of obvious UPK, like 1. only click on safe zones that make openings, 2. ALWAYS make the best choice for flag-double click/NF.
There are for sure, many other points 3. 4. etc... that could make a right addition to the definition of UPK

@James: I think there are some signs which show on a video whether the board was practiced or not. Even if it is the dreamboard: people need some hundreths of second to apprehend what's on the screen, then to react and the movement of the mouse can easily give it out. So if one plays int on the original MS version RANDOMLY, and did not memorized the dreamboard, then it can be seen with the help of difference between conscious and unconscious moves of the mouse. In Rodrigo's example with the DB, the most logical next move should be starting to solve the board - but if you have upk, you should just make a CONSCIOUS new opening, leading the mouse straight over safe places instead of hesitating a bit. Of course you can avoid recognition but that adds some seconds to your result!/And maybe it's a bit hard to play that fast and that conscious./

I am not saying anything about practicing the board or about board cycling. I am also not saying that ALL scores on the dreamboard are legit. I am simply stating that just because you recognize that board, it doesn't automatically constitute as UPK.

@Rodrigo (and everyone): I see a small problem with your example, especially when demonstrated with the dreamboard. In this case, it doesn't matter whether or not you have prior knowledge of the board. Why? In the case of your example the first two openings, which could be achieved easily by dumb luck alone, the EXACT location of 29 mines are revealed by the numbers, many of which would only be marked by a true novice player. Through the course of normal play, one or two flags + Double-clicks would reveal the EXACT location of at least 5 more mines. I will admit that it is a very easy board to recognize and play, but the fact of the matter is that approximately 87.5% of the mines are revealed in a truly fair manner. My conclusions are these:

Either UPK is an overrated concept...

or...

Your definition of it needs to be refined (by the whole community, maybe)....

or...

The dreamboard is not truly subject to the rules defined by UPK...

or...

All of the above.

congrats oli! that's quite an achivement

@rod: i think you did a really great job defining and explaining UPK. i never understood the situation fully before and this really helps me

47

Hopsing was so kind to upload my video on his site.

Congrats, Oliver!

Some weeks ago someone asked if there's a program which traces the road of the mouse: here it is: mof.aklabs.com
I just found it by accident. Btw, I don't take any responsibility of how it works as I haven't tried it out

47!?!? Holy Cow! First German to go sub50, I guess we can call this race over.
What was the 3bv, do you have a video?
If so, her damit!

5x23!!!! must be a bug in the game

I am ear-witness!! He was screaming, and while playing I was even talking with him

52 -> 47!

I never thought i would make a sub50 this soon, but a 47! i was like OMG ask Hopsing after a year finally a new record

You often can't know for sure if someone had prior knowledge of the board, and there's the potentially for people subconsciously knowing the board and not realising

Michael - nope, I hate apple with a passion, sorry... I seem to remember there being something called virtual PC for windows stuff though. And when you say the other one has a slow mouse, you could always get another mouse..

I just don't know how you can live without minesweeper

@Rodrigo: I almost agree. One quibble, though: suppose someone plays your first example board, leaving the guess in the bottom right corner to last. Only when he has these squares left unopened, he realises he has seen the board before. What happens then?

@Dave: I have two computers at home; one's a Mac (and you've got to admit, EXCEPT for Minesweeper, Macs rule) and the other has a really slow mouse. I have only once got sub-120 on that machine.

1 --> 1x2

I got 1,87 on clone, 3BV 3.

www.gammel.narod.ru/1-2.mvf

Well, guys, I have been thinking about what truly defines an UPK. I think that there is always a way to determine it exactly. I wrote my definition on a webpage with examples and pictures. Something short and simple. It also says about patterns and cycling. The address is http://www.geocities.com/minesweeperclone/upkdefinition.htm .

It would be nice to know if you agree with my point of view. Just let me know about it.

That's probably because nobody plays beg anymore since it was removed from the AR. Personally I tried it only one day. Take a look at my stats and you'll see.
6,60 does sound pretty good though....

I just looked at the 3BV/s ranking on Detrusers site and couldn't belive the values for beginner. There's only one >5 !!!
My personaly best on the clone is 5.80 (3BV=17 score=3.93). On MS version I managed a 6.60 (3BV=7 score=2.06) unfortunatly I have no vid of the second one.
Is there anyone with compareable values?

@Jon: No, I didn't receive any of your e-mails. Could you please send again to minesweeperclone@bol.com.br? Ah, by the way, thanks for telling me that, because I didn't notice that my mailbox was full.

79

19

1

Hello everybody!
I'm crazy about density and today I solve the intermediate (16x16) with 88 mines. I know one guy who solve with 86 and 2 with 85. Do you know anyone who do better than me? I think I'm perhaps the world recorder with this score...
If you know something about density, I'm always interest.
Best regards from Switzerland!

what kind of a computer doesn't have minesweeper on it?

"If you go with the winner of that poll, I would accept it, however, you would still have the dilemma of deciding what constitutes as "Unfair Prior Knowledge." Not only that, you would still be stuck with having to take someone's word for the questionable records that are on the list, and we would be right back to where we started."

It's a good point, I admit. But we wouldn't really be where we started, because (1) we have the decision, and that allows to move forward; (2) the next poll could be to determine what constitutes UPK; (3) although it depends on what exactly we decide for 2, there are some records that we pretty much know are UPK, and we're going to have the "problem" of having to take people's word whatever, unless we decide that only clone records are eligible, which you'll notice a huge majority of voters voted AGAINST .

I'm not joining the AR because of my difficult position: I've just finished my Finals, so (1) I don't like my times, but haven't had much chance to work on improving them, which I hope to do very soon; (2) in three weeks I won't have access to a computer with Minesweeper, possibly for a long time.