It started as a discussion about removing cheaters from the ranking, and someone posted something about Dan Zhou doing this.
There have been discussions about this before that weren't conclusive, but it looks like we need to revisit this in the light of this discovery.
For those interested, http://www.minesweepergame.com/forum/viewt ... f=12&t=238 is a very relevant topic here.
Here are the relevant parts of the recent discussion here (edited slightly for readability*):
lh says:
[...] Furthermore last time during our private chat you mentioned that Dan Zhou is using keyboard to project 'chording' clicks (finally we made the agreement on it) because there is almost zero difference (never more than 0.01) between interval of left and right clicks when chording behavior is done, based on the detailed analysis of FreeSweeper tool, so one question for valid videos which needs to clarify is: should the minesweeper be played exclusively by one two-buttoned mouse device and a monitor? If not, her record should be removed from world record as someone voted. P.S. For the people mentioned above, I do not meant to offend you, since we should realize the fact that everyone(no matter which country this player belongs to) should be equal in front of rule, no matter how famous you were.
Tommy says:
@lh the chording thing is worrying, can you describe that in more detail?
* fixed typo in lh's first post instead of having the first post and the correction, merged three posts that were made within half an hour, edited chat log to use names instead of numbers.lh says:
@Tommy First you need to have the version of Freesweeper, if you don't, you can download the program file via this link: https://www.sendspace.com/file/9edgod Then go to the World Ranking and download the current time record of Dan Zhou, click 'Game'->'Load video', then open this file. Now lets take a look at the window of 'Video control', look at the green stripes which indicate left clicks and dark red stripes which indicate right clicks, the starting point and ending point in # of 0.01s are indicated in numbers above and below each stripe. As we know, 'chording' is done while we detect that both left button and right button are pressed down, and then either of them is released. So during normal HUMAN operation, we need to press both left button and right button to complete this operation. OK we can find that, for any times that Dan Zhou is making intentional or unintentional 'chording', left click and right click starts and ends at EXACTLY the same number of 0.01s, or the difference is NEVER more than 0.01s. If we looked at some other videos such as her 3bv/s record or some of her other Exp sub40s, we will find this same fact. At this time, both Guo Jin Yang and I don't believe that any HUMAN BEING is capable to press both buttons with two fingers(or one finger) with never more than 0.01s of difference, but, if we consider another possibility that Dan Zhou is using a single 'middle button' or a single 'side button' on the mouse, or using a single key on the keyboard, we tend to believe that zero or 0.01 differences can be managed to make, since a key projection program may cross the boundary of two consecutive 0.01s during its execution. I have no idea whether the latter case that we believed is considered as 'cheating', but we do believe that the game is played as the latter case.
As what I know, 'cheating' is considered when most admins who inspect videos think so, and I am not sure if other players (no mattter what nationalities this player is) are violating the criteria of one mouse and two buttons that I said in previous comments.
Guo Jin Yang 6:56:15 AM : What about scores using both keyboard and mouse clicks
lh 6:57:15 AM : Last time we talked about Dan Zhou's videos, and Jin-Yang Guo said that her video was made by both keyboard and mouse--keyboard for double clicks and mouse for right clicks
Guo Jin Yang 6:57:50 AM : She uses one key to do the same thing as a chord
Guo Jin Yang 6:58:12 AM : What do you think? I think it is legal
Damien Moore 7:01:36 AM : Difficult questions. Many years ago, a player changed his mouse so it made clicks when you pressed down or up. We decided that was illegal. If the original minesweeper game lets you do something, then it should be legal. For example, in the original game you can move using arrow keys on your keyboard instead of a mouse. And you can chord my moving your mouse on a square and pressing SHIFT when you click the left button.
Damien Moore 7:02:29 AM : (I mean, if the original game rules let you do something)
Note: QQ170524603 is Guo Jin Yang, QQ137255795 is me, and QQ1967998590 is Damien Moore, this is part of the group chat record that we discussed about this issue. If the rule on minesweeper.info is really as what Damien said, does anybody know what the rule of the 'original minesweeper game' is? Does anybody know how to reach the developer of this Microsoft game? Maybe we can reach the 'original' minesweeper game developer himself/herself to get the exact opinion.
To see whether Dan Zhou might have used the middle mouse button, I played a game with arbiter where I used the middle mouse button, and then loaded it in freesweeper. That looked differently; the MMB events were recognized as one event.
I agree that this, together with how consistent those clicks are, means that some kind of macro must have been used.
But there is something else that struck me: The length of those double events. Most of them are exactly .10 seconds long, and those that deviate don't do so by much. And this means that I suspect that not only is a macro being used, but I also think that this macro automates both pressing and releasing the mouse button. And if that's the case, I find it highly problematic.
I'm open to the idea of just mapping a key on the keyboard to a mouse button. Press the key, press the mouse button; release the key, release the mouse button. Basically, we can think of a keyboard as a mouse with many keys and no movement. I don't really have a problem with people using a keyboard like this; While you could argue that clicking with a mouse potentially affects your accuracy (as you have to compensate for how your hand moves when you click as you move the mouse), you could also just plug in another mouse, or use a laptop's touchpad.
But doing more than one thing with one input event changes how the game itself works. You have less actions that take time and attention.
Elmar technique, by the way, is also an example of this - both
Code: Select all
LMB down
Code: Select all
LMB up
Code: Select all
(LMB down, LMB up)
So, I would like to propose a baseline criterion:
If a macro is used that maps one user action to more than one game action, that is definitely cheating.
Also, I'll reformulate this for legal clones:
If a clone maps one user action to more than one winmine game action, that clone may not be legal.
If we agree on this, we can then discuss what else might be cheating (or not).