Posts:
1,032
Registered:
Dec 16, 2001
From:
Powell, OH
Age:
37 Home page
Re: Extend wait limits when opponent is disconnected
Posted:
Feb 27, 2003, 4:00 PM
I must 2nd vitals on agreeing that Scott has a very interesting point on #3. I don't know why I never thought of it that way, and that is exactly why I am glad this issue is being discussed at the forums! Thanks to everyone for their input.
Of course, there are still some questions to figure out. Like emerald said, if a player has say 18 minutes left and returns after 17, what is their timer going to be at now? Should it be (1) left the same, (2) dropped to 1 minute, (3) reduced by 1/2 the time they were gone or something like that?
(1) If we leave the timer at 18 minutes, then vitals point that the opponent can't gripe isn't true, because now the opponent has waited 17 minutes while the player was gone and still has to wait potentially 18 more minutes before the players timer runs out. Just to point out, this is how the current system works in that the timer is not changed between when a player is disconnected and returns. This is fair to the player who got disconnected but potentially unfair to the player who waited (only in that they may have other commitments to go to and can't finish the game). Plus, think about the extreme case. A player is disconnected for 17 minutes and returns only to be again immediately disconnected for another 17 minutes and so on. Obviously that isn't going to happen much but it could. Also, there is something else I've thought of before but seriously doubt anyone does. A player could gain more time by disconnecting for 4 minutes while they study the game on another board. Then return to the game ready to make their next move. Of course the waiting player has the same time addition but still. Basically real-time games over the internet coupled with the possibilities for disconnects can never work exactly the way we want them to.
(2) If we drop the players timer to 1 minute (or possibly it could be 1 second if the player returns at the exact right time) we're not giving the player much of a chance of finishing the game, even with incremental time. This is perhaps overly fair to the player who is waiting but unfair to the player who got disconnected.
(3) Do some sort of time reduction. This one just seems weird to me and is penalizing the player for being disconnected. But it is trying to compensate the player who has waited around. This one tries to be somewhat fair to both players. But I think it would be confusing to people and difficult to implement.
After writing this, I'm not really sure I've made the waters any clearer...this is a very tricky issue and I guess I've been lucky people have stuck with the current system for so long!
Remember, we could always just get rid of cancel/force resignation and instead save games for later play. Listing a players # of "unfinished" games should help discourage players from abusing this policy.
Or, I also sort of liked the idea of 2 separate time limits, one to cancel and the other to force forfeit, again something I just had not thought about.
About Scott's, Vitals' and Dweebo's posts.
Posted:
Feb 27, 2003, 9:48 PM
Scott's post has some excellent points, and although his post did not match the verbosity of my post, in its terseness it is probably equally effective, if not more so. For one, I think the following statement succinctly sums up what I took many paragraphs to say (italics added by me for emphasis):
"However, playing against someone over the internet is simply not the same. The other night I was playing against Arne who lives across the ocean. I'm not sure what he was going through, but we had a mighty thunder storm in Florida that lasted the four hours we played. The lightning popped, I jumped and then it would pop again and the lights would dim then it would pop some more and my cat would jump. Also, I would click my mouse intending to place a stone but had to wait and reclick several times. I understand that exceeding your time limit or misplacing a stone at DSG is not the same as doing so over-the-board. This seems to be a clear distinction that most players intuitively understand.
Winning due to a technical problem unrelated to the actual game does not appeal to me. Losing due to a technical problem is even less appealing."
Well said. Any true gamer would wholeheartedly agree with the italicized statement.
On a related note, some might ask, "how can one tell a misclick from a bad move?" Well, for starters, some are obvious-a stone drops nowhere near the action, or, a stone drops one space away from blocking a forcing threat and lands in a place that obviously serves no purpose. But what about when a misclick happens and it could possibly just be a bad move that the player realized a second too late? Well, if the move truly is a misclick, then the player should request undo immediately. If I were playing under an "undos for misclick only" understanding, then I would trust my opponent to only request if it really were a misclick. Obviously, in a case like that, a player who says "Well, I can’t be sure if that was a misclick or not, so I will deny the undo" can hardly be argued with, although I would hope he has a good reason for essentially accusing the other player of lying. But, I am getting off the main topic here.
Now, on to the other excellent point that Scott made, that for some reason went unsaid until now. Of course, if one has 18 minutes left, under no circumstances should an opponent be able to force a resignation any sooner than 18 minutes! I say the same should hold for canceling, because to some, a cancel is the same thing as a loss (If I am playing someone 600 points below me in rating, and the other player has a clear win, he stands to gain 32 points by winning. If I force resignation, I gain none and he loses none [to be precise, I gain .4 and he loses .4] and if I cancel, same thing).
What do when a player has 18 minutes left and then returns after 17 minutes from a disconnect? This is so rare that I am not sure it is worth the trouble spent on it. If it does happen, giving the player only 1 minute is not fair. As it is now, the clock does not run during a disconnection, and no one seems to mind. Not once have I or any other player that I was aware of complained "Hey, I had to win 5 minutes for him and now he still has his full clock! These are precious minutes of my life that I’ll never get back!" No, no one has said anything of the sort. But what to do when a player legitimately has to go somewhere, and certainly was not accounting for a 20 minute opponent timer AND 17 minutes of disconnected time? Well, in those rare cases, perhaps the player should be docked half the time he was gone. Is that completely fair to either side? NO, but it is the best option, and for the disconnected player, it is a lot better than having resignation forced (or the game cancelled if he is in good shape).
Dweebo’s idea of having suspended games is an interesting idea (and an ambitious one I would guess, it sounds difficult to program, although I don’t know anything about programming except for pseudo code, which I am now learning. But I digress), but I am not crazy about it. Adding a "suspended games" category is undesirable. If this were added at all, it should be in ADDITION to the other 4 options, I.E. the box would say: You may 1) resign 2) force resignation 3) cancel 4) keep waiting or 5) suspend the game, to be completed the next time you and your opponent are at the same table. But, then a player can study the game. This would be an ethical no-no, but at the same time, if a similar situation arises in another DSG or IYT game before it gets to be replayed, what then? So, suspending games is probably not a great idea, and probably not necessary.
Which brings me to my next point. Dweebo said,
"Also, there is something else I've thought of before but seriously doubt anyone does. A player could gain more time by disconnecting for 4 minutes while they study the game on another board. Then return to the game ready to make their next move. Of course the waiting player has the same time addition but still."
To me, this is a non-issue. Could people do this? Sure. Would they? I hope not, and we just have to assume they would not, the same way we assume that people are not using the database or moving stones around on a real board in a game whether disconnected or not! As it is, we trust each other, and with disconnects, we simply have to continue to extend that trust.
One last possible idea, going back to when Dweebo said,
"Do some sort of time reduction. This one just seems weird to me and is penalizing the player for being disconnected. But it is trying to compensate the player who has waited around. This one tries to be somewhat fair to both players. But I think it would be confusing to people and difficult to implement."
I don’t think this is "weird" at all, I think it is creative and clever. I would not view it as "penalizing" the player, but rather, as saving him from a worse fate. I cannot comment on whether it would be difficult to implement, but as far as it being confusing, I would say it is rather clear.
But perhaps the best way to do this would be to give the other player the option? When the player returns, the clock could be frozen as a box comes up saying "your opponent has returned after being disconnected. If you are pressed for time, you may reduce his time remaining by half the amount of time he was disconnected." Once the player makes his decision, the game continues and the timer starts.
Most often, the player would probably not feel the need to enforce this option, but it is there if needed. Obviously certain players who I will not name would choose this option no matter what, but it doesn't bother me because I do not play against those players anyway, and even then, the disconnected player still should have some time left.
That's all I have to say about that.
If I do not accept a game invite right away, it means I will once I have fewer games in progress.
Re: About Scott's, Vitals' and Dweebo's posts.
Posted:
Feb 28, 2003, 6:25 AM
Hello,
Regarding Scott's point about returning after being disconnected, if a player gets disconnected with 12 minutes remaining and it takes 8 mins to return, should there not still be 12 minutes remaining? The time limit of a game is the time it takes to actually play the game. Players shouldn't be penalized if they are disconnected and return.
The issues seem to be (to me at least), how much time is reasonable to allow the remaining player to wait before he gets a chance to decide the outcome of a game, and how to restrict the remaining player from taking advantage of the situation. These seem to be two distinct problems.
If it is possible to programmically (sp) determine the probable winner of a match at the point of a player being disconnected, then it may not too restrictive to not allow the remaining player to force a resignation if, say, the probability of the player winning the match is less than 75% or so. Is it not the case that for a given board, it can be determined statistically who the probable winner is? So for example, if it was determined that the remaining player had only a 40% chance to win, his only choice would be to cancel the game (even though, like Scott says, he would be a loser no matter what his rating but at least he wouldn’t win); if it’s 75% or higher, he may force a resignation or cancel.
It may be a nightmare to code for that (if so, my apologies, Dweebo), but I think something like that may remove some of the disagreement about what a player should do in this situation. Plus, it would make it more consistent.
You know, one of the reasons I find pente to be such an enjoyable game to play at Dweebo’s site is because of the sportsmanship that has always been present. By and large, the people who play here are some of the nicest people to meet online. Perhaps the best way to avoid the disagreements regarding this subject is to know your opponent as best you can.
Re: Extend wait limits when opponent is disconnected
Posted:
Mar 1, 2003, 10:53 AM
Dweebo -
Here's the vote so far: 4 for 10 mins. 3 for at least 7 mins. 1 for 5 mins.
Then there's Scott's excellent suggestion which might take a little longer programmtically, but may be the best suggestion of all. It would also be 7 mins. or longer under most conditions.
While mulling all of this over, why don't you go ahead and make a quick change and increase the time to 7 mins. since we've got 8 of 9 (89%) in favor of at least that increase so far.
Posts:
1,032
Registered:
Dec 16, 2001
From:
Powell, OH
Age:
37 Home page
Re: Extend wait limits when opponent is disconnected
Posted:
Mar 1, 2003, 4:58 PM
The system has been this way for over 3 years, I don't see a huge need to rush in a special change for this. I would much rather make a change to 7 minutes along with other related changes in a release.
It may be a quick change, but what that really means is a few hours of work for me (changing the code, checking everywhere to verify the change will work, testing the change, checking in the code, building the code, restarting the servers, etc.). That is why I normally make "releases" of groups of fixes/enhancements except for cases where there is a major bug. Or if there is something I feel like adding for fun, I'll put up a small release.
Also, there are currently something like 55 open issues that I have to work on, see them at http://sf.net/projects/dsg. Some of them have been waiting for over a year because I don't have the time for them. I have tried in the past to get other open source developers to help but that hasn't worked so well yet. So I try to work on issues that have the most importance to all players, and this does appear to be an important issue. As I stated before, I will be working on whatever solution we "agree" on in the next few weeks.
Thanks again for everyone who has posted comments. -dweebo
Posts:
39
Registered:
Dec 16, 2001
From:
~THE BACK WOODS~
Re: Extend wait limits when opponent is disconnected
Posted:
Mar 1, 2003, 7:32 PM
Here are my thoughts on Dweebo’s reply to Scott’s post.
-Dweebo: I must 2nd vitals on agreeing that Scott has a very interesting point on #3. I don't know why I never thought of it that way, and that is exactly why I am glad this issue is being discussed at the forums! Thanks to everyone for their input. Of course, there are still some questions to figure out. Like emerald said, if a player has say 18 minutes left and returns after 17, what is their timer going to be at now? Should it be (1) left the same, (2) dropped to 1 minute, (3) reduced by 1/2 the time they were gone or something like that?………………
I think that if we made it a 3-minute minimum for time left to finish a game, then this would not be too unfair to either player. If the player returns and has over 3 minutes left on the game timer, then the player should be left to finish out the game with that amount of time, if the game timer is under 3, then the player would be given up to a minimum of 3 minutes to finish the game. This, to me, would be a better option than having a resignation forced. 5 minutes would be better than 3, but I would be fine with at least having 3. However, this system would not work to well for speed games. We need to look at still having the regular system that we have now in place for games that had a starting timer under the amount of time given to return from a disconnect. All this seams like it would be very complicated to do.
-Dweebo: (1) If we leave the timer at 18 minutes, then vitals point that the opponent can't gripe isn't true, because now the opponent has waited 17 minutes while the player was gone and still has to wait potentially 18 more minutes before the players timer runs out. Just to point out, this is how the current system works in that the timer is not changed between when a player is disconnected and returns…………….
I agree that if that is how the system was, my NO gripe point is a bad one. That is way to long to wait. I was not thinking the player should get all the time back that was left on the game timer at the time of disconnect.
-Dweebo: After writing this, I'm not really sure I've made the waters any clearer...this is a very tricky issue and I guess I've been lucky people have stuck with the current system for so long!
If the system never changes from the way it is today, I would still love this site as much as I always have. If the worse thing that happens to me here is having to deal with some one forcing my resignation a few times per year, then I would have to say WOOOOOOOOOOOOHOOOOOOOOOOOOOO and would always keep coming back. When I first started playing at this site, the problems with disconnect and having forced resignation slapped on you was WAY more common. This site has improved so much in just a short period of time and we owe it all to you Dweebo, Thanks again for all you do!
Peace Out Vitals
P.S. Dweebo, Get your homework done and come play some Pente!!!!!!
Posts:
55
Registered:
Jan 21, 2002
From:
ohio
Age:
65
Re: Extend wait limits when opponent is disconnected
Posted:
Mar 2, 2003, 11:59 AM
vitals point of having minimum 3 min left would work with the exception of say after the 5th move vitals gets booted , i wait 17 min and he comes back, now to finish game w/ 3min on his clockto my , say 18. this of course wouldn't be fair to vitals. so, i would offer cancellationas much as i hate to give him anything!!!!! LOL yuk yuk yuk
Posts:
2
Registered:
Dec 16, 2001
From:
Kansas
Age:
19
Re: Extend wait limits when opponent is disconnected
Posted:
Mar 14, 2003, 7:21 AM
I was disconnected tonight and I thought of a situation where keeping someone's time going while they are gone would not work. In a speed game, this would be an automatic loss pretty much. I realize that most games are not speed pente, but I thought I would bring it up. Didn't know if anyone had thought of this situation before or not. But I think a speed game would need a different set of time rules, making things more confusing, so I don't even know why I brought it up. Oh well.
Re: Extend wait limits when opponent is disconnected
Posted:
Mar 14, 2003, 10:34 PM
i think that an extention would be fine. whether its 7 10 or 12 is fine with me and my old pc. ill go along with any time limit that the site master rules. dufus
Posts:
1,032
Registered:
Dec 16, 2001
From:
Powell, OH
Age:
37 Home page
Re: Extend wait limits when opponent is disconnected
Posted:
May 15, 2003, 6:58 PM
Just to reiterate the news posted in the News forum, the timeout has been extended to 7 minutes.
Also, the popup box now appears immediately when the opponent leaves the table with a timer counting down to 0. It also has a big paragraph explaining a players options and has a new button to resign at any moment. The buttons to cancel and force resignation are only active after the 7 minutes are up.
Posts:
542
Registered:
May 9, 2002
From:
Northeast USA
Re: Extend wait limits when opponent is disconnected
Posted:
May 16, 2003, 5:26 AM
Ok, here's my two cents on this topic:
First of all, I want to add my own point which I don't think has been mentioned yet and in my opinion is probably the most important. Perhaps this is already implemented but I doubt it... It is possible in internet based gaming sites like this one to actually determine accurately whether someone has disconnected voluntarily (for example, closing out the gaming window in the middle of a game) or has been involuntarily disconnected (for example, because of a network error or a computer crash). I know that I have a high speed always on cable internet connection and when I get booted it's because my computer CRASHES and it crashes HARD. Incidentally, this type of crash does not occur on my computer while doing any other activity -- at least not with nearly the same frequency -- this may be a bug on the server side that should be looked into, but I digress... The point is, I believe that this type of detection of voluntary/involuntary disconnection is important and should be detected and additionally, I believe there should be a separate chain of events that occur in each case.
For those that just read this and say "you don't know what you're talking about, that would be impossible to detect accurately" -- I play online poker for real money, and believe me when I tell you that probably the MOST important aspect of the software for many players is: what happens when I have REAL money in the pot and I am booted? And likewise, what happens if my opponent is voluntarily disconnecting in an attempt to not be forced to put MORE REAL money in the pot, but not actually fold either? If this was not addressed extremely accurately and effectively, probably 99% of serious players would not play online.
Now, let me sum up what seems to be the common opinion of myself and others about what should happen in the event of an INVOLUNTARY disconnection: 1) Game is paused 2) Dialog appears immediately -- A) Option to RESIGN NOW is available -- B) Timer should count down from 7 min 3) Once 7 min expires, game should be RESUMED -- A) IF it's remaining player's turn -- he takes his turn, THEN (or if it's booted player's turn)... 4) Game is PAUSED 5) Dialog appears again -- A) Option to RESIGN NOW still available -- B) Option to CANCEL GAME becomes available -- C) Option to CONTINUE WAITING becomes available -- D) Option to SAVE GAME (which would imply there is some way to LOAD GAME) becomes available (maybe) 6) If CONTINUE WAITING selected: Game is RESUMED 7) After every 7 minutes elapsed time: return to step 4 and repeat. 8) When the booted player's timer reaches 3 min: execute steps 4 and 5 -- if CONTINUE WAITING selected here: Game remains PAUSED 9) Dialog with a 3 minute timer appears and counts down to 0 10) Dialog appears again -- A) Option to RESIGN GAME still available -- B) Option to CANCEL GAME still available -- C) Option to SAVE GAME still available (maybe) -- D) Option to FORCE RESIGNATION becomes available -- E) Option to CONTINUE WAITING still available 11) If CONTINUE WAITING selected: return to step 9 and repeat.
Obviously this can be tweaked but I think that's what the majority of people were getting at for the most part. HOWEVER, I believe this is vastly different than the sequence that should occur in the case of a VOLUNTARY disconnection:
1) Game is NOT PAUSED (or is TEMPORARILY paused if it's easier to pause the game when popping up a dialog) 2) Dialog appears immediately -- A) Option to RESIGN NOW becomes available -- B) Option to CANCEL GAME becomes available -- C) Option to SAVE GAME becomes available (maybe) -- D) Option to CONTINUE WAITING becomes available 3) If CONTINUE WAITING selected: game is RESUMED (or continues running if never paused) (remaining player takes his turn if required) 4) After every 7 minutes elapsed time: return to step 1 and repeat using the following modified dialog: -- A) Option to RESIGN NOW still available -- B) Option to CANCEL GAME still available -- C) Option to SAVE GAME still available (maybe) -- D) Option to CONTINUE WAITING still available -- E) Option to FORCE RESIGNATION becomes available 5) In addition, a maximum of 2 VOLUNTARY disconnections should be allowed as above (in case one "accidentally" closes the window, etc). On the third such occurance, the Modified dialog as in step 4 shall be used to allow the option to FORCE RESIGNATION, even if not much time has elapsed between disconnecting and reconnecting each time -- since this many occurances may likely be intentional malicious behavior.
Posts:
1,032
Registered:
Dec 16, 2001
From:
Powell, OH
Age:
37 Home page
Re: Extend wait limits when opponent is disconnected
Posted:
May 16, 2003, 3:18 PM
> For those that just read this and say "you don't know > what you're talking about, that would be impossible > to detect accurately"
Right, it is impossible to detect accurately. Therefore at DSG I always give the person who left the benefit of the doubt that they were disconnected "involuntarily".
I was not thinking the player should get all the time back that was left on the game timer at the time of disconnect.