Life In 19x19

topazg wrote:That doesn't answer any part of my question. In reality, 50% of the games I, or anyone else plays, are not escaper games. By deranking me, you remove far more good data than bad, do you not? Doesn't removing the additional amount of good data damage the integrity of the rating system further than the 1 game in the other direction?

If you want to work out the specific special cases for when exactly it becomes mathematically detrimental to the system for you personally, you are free to, however what if the case were - "A player plays 95% of his rated games on the up and up, but every once in a while lets his younger brother play his rated games for him, so he has the chance to play stronger players." As far as the rating system is concerned this works out to be pretty much the same.

I avoided explicitly answering your question because your question has been already answered as far as I'm concerned. I have stated (and stand by the statement):

"At the end of the day, people who are actively engaging in behavior that causes bad data points to be added to the ranking system may be subject to deranking."

or I guess this conversation could have been shortened simply by referring to the KGS Terms of Service:

KGS ToS wrote: To be more specific, these are not allowed:

-Intentionally losing rated games

I guess the question for you to answer would be, do you really feel you derive sufficient additonal value from removing italicized games from your list (additional value that cannot be obtained via other methods previously mentioned of addressing the issue) that it merits going out of your way to introduce incorrectly scored games into the KGS database?

easiest solution:
If a player gets disconnected, a grace period of 5 minutes is given for him to come back.
If he does not come back, he loses the game.

I mean, if you got bad internet connection, no sympathy really..

So which of these steps is illegal?

Escaping a game you are losing.

Playing completely drunk and tired, almost guaranteeing a loss.

Not playing very hard, letting your opponent get away with stuff.

Playing hard, but having bad judgment and resigning a game you were winning.

Resigning a game because you have to leave while not wanting your opponent to be mad at you for escaping.

Throwing a game.

Resigning a game your opponent escaped from.

They all have a negative effect on the ranking system, but it seems only the last one or two are illegal. Escaping a game is actually the rudest item on the list, but it is perfectly fine by the KGS rules as long as you don't do it too often. That's what doesn't make sense to me.

I guess the question for you to answer would be, do you really feel you derive sufficient additonal value from removing italicized games from your list (additional value that cannot be obtained via other methods previously mentioned of addressing the issue) that it merits going out of your way to introduce incorrectly scored games into the KGS database?

Frankly, I don't give a damn how my actions affect a ranking system. I understand why some do, but I bet most people don't care about some how something small they do affects the overall ranking system.

Mef wrote:I guess the question for you to answer would be, do you really feel you derive sufficient additonal value from removing italicized games from your list (additional value that cannot be obtained via other methods previously mentioned of addressing the issue) that it merits going out of your way to introduce incorrectly scored games into the KGS database?

Here the counterquestion for you to answer is, do you really feel that if the answer to the above is yes, the solution is still to invalidate all games by that player to introduce many more incorrectly scored games into the database instead of providing a way to produce that additional value via another method that rewards the escaper less than a resignation?

Kirby wrote:
They key point to take away is: Resigning is always a valid option for either player. The same is true in real life games.

Now here we agree in principle, but not in practice. In fact, I would say it is more a disagreement in definition. The principle is "You should not be required to continue a game you have no desire to finish." In real life, you walk away, it is effectively a resignation, sure. KGS provides that option -- you can just leave the game. If you were to regularly attend tournaments, and resign half your games even though you were crushing your opponents, I would imagine at some point the tournament directors might have a word with you.

No one is talking about deranking someone who mistakenly resigns in a position they think they have lost in. The issue here is intentionally resigning a game whereyou feel you are in a winning position, even though you could otherwise leave that game unfinished.

Mef wrote:...

"At the end of the day, people who are actively engaging in behavior that causes bad data points to be added to the ranking system may be subject to deranking."

or I guess this conversation could have been shortened simply by referring to the KGS Terms of Service:

KGS ToS wrote: To be more specific, these are not allowed:

-Intentionally losing rated games

I guess the question for you to answer would be, do you really feel you derive sufficient additonal value from removing italicized games from your list (additional value that cannot be obtained via other methods previously mentioned of addressing the issue) that it merits going out of your way to introduce incorrectly scored games into the KGS database?

The non-escaper did not do anything wrong. Resigning is a valid option in the game of go. People resign all the time when they could have won, and sometimes they don't even know it.

Do you have a playing account on KGS? What if I were to find a game record that you had where you resigned, but I thought you could have won? Can we derank you because you resigned too soon? Have you ever resigned a game after losing a big group near the early mid-game? Couldn't you have come back? If it were a possibility, then let's derank you for intentionally losing that game.

There is no way to systematically quantify how much someone "intended" to lose a rated game. For example, I could say that I am actually a 9d player, but I have been playing poorly on purpose to have a lower rank.

The person that has committed fault in this scenario is the escaper. But we give them a what - grace period that says it's OK to escape once in awhile when you don't feel like losing?

Mef wrote:... The issue here is intentionally resigning a game whereyou feel you are in a winning position, even though you could otherwise leave that game unfinished.

Did you ask every player that you have deranked if they felt that they were in a winning position?

Furthermore, let's get to the real point:

Why not simply disallow escaping altogether, and provide an option for users to resume games later upon mutual request (with losses after a set amount of time due to connection issues)? Is this type of system inferior in any way?

softbank wrote:easiest solution:
If a player gets disconnected, a grace period of 5 minutes is given for him to come back.
If he does not come back, he loses the game.

I mean, if you got bad internet connection, no sympathy really..

Exactly. I might add an option for a mutually agreed upon postponement of a game. That way there are no benefits that I can see that the current system has.

Mef wrote:1: Escaping potentially excludes a valid data point from the ranking data set, whilst resigning an escaped game assuredly includes a false data point. Further as others have stated, resigning the escaped game not only includes a false data point, but it increases the likelihood that more valid data points will not be included (because it extends the duration prior to have escaped games count as forfeits).

if escaped games are eventually forfeited, then no it doesn't exclude a valid data point

hyperpape wrote: Escapers want to avoid losses--this lets them do that for at least one game.

No, escapers aren't trying to avoid loses.

In every other server, closing the game window is the same as accepting a loss -- and its fewer button presses.

In my experience, most escapers are either new players, kids, or non-english speakers who don't know there's an extra button to push to resign a game.

Kirby wrote:The non-escaper did not do anything wrong. Resigning is a valid option in the game of go. People resign all the time when they could have won, and sometimes they don't even know it.

They did not do anything wrong...until they resigned. Once again the KGS Terms of Service explicitly states:

KGS ToS wrote:To be more specific, these are not allowed:

Intentionally losing rated games

Now, you have also brought up the point "How can you be sure they think they are in a winning position." On its own this is a valid argument, however, if you are assuming the other person has escaped that would imply the escaper is in a losing position. If they are not in a losing position, why are you considering them an escaper? In that case, the person who has left the game has done nothing wrong, perhaps they have a network issue, or an emergency.

Kirby wrote:Do you have a playing account on KGS? What if I were to find a game record that you had where you resigned, but I thought you could have won? Can we derank you because you resigned too soon? Have you ever resigned a game after losing a big group near the early mid-game? Couldn't you have come back? If it were a possibility, then let's derank you for intentionally losing that game.

For the record my account is deranked (=

However...

Mef wrote:No one is talking about deranking someone who mistakenly resigns in a position they think they have lost in. The issue here is intentionally resigning a game whereyou feel you are in a winning position, even though you could otherwise leave that game unfinished.

xed_over wrote:...
if escaped games are eventually forfeited, then no it doesn't exclude a valid data point

Not all escaped games are eventually forfeited. You don't get a loss if you only forfeit once in awhile. Your rank is also calculated based on time. Games that happen further back in your game history affect your current rank less.

So getting an immediate win for an escaped game would provide more weight toward your rank, because if the win is delayed by a couple of months, for example, the win holds less weight.

In an extreme case, let's say that all of my "should-be" wins were games against different escapers. If those escapers don't play often, my rank goes down because I only have losses recorded, when really my "real" rank may be stable at the current rank.

If escapes were counted as immediate losses, then this situation would not occur.

Mef wrote:

Kirby wrote:The non-escaper did not do anything wrong. Resigning is a valid option in the game of go. People resign all the time when they could have won, and sometimes they don't even know it.

They did not do anything wrong...until they resigned. Once again the KGS Terms of Service explicitly states...

Yes, that's what the KGS policy currently is. I think the discussion here is what it should be.

So, again, can you answer this question:

Kirby wrote:Furthermore, let's get to the real point:

Why not simply disallow escaping altogether, and provide an option for users to resume games later upon mutual request (with losses after a set amount of time due to connection issues)? Is this type of system inferior in any way?

Mef wrote:Now, you have also brought up the point "How can you be sure they think they are in a winning position." On its own this is a valid argument, however, if you are assuming the other person has escaped that would imply the escaper is in a losing position. If they are not in a losing position, why are you considering them an escaper? In that case, the person who has left the game has done nothing wrong, perhaps they have a network issue, or an emergency.

Maybe they are an escaper because they are "intentionally losing games". This is what you guys don't like, right? It appears that the policy is more harsh toward resignations than it is toward escaping, so if I want to sandbag, escaping is a good way to do it.

Mef wrote:or I guess this conversation could have been shortened simply by referring to the KGS Terms of Service:

KGS ToS wrote: To be more specific, these are not allowed:

-Intentionally losing rated games

That should probably be reworded. According to that, resigning is never allowed under any circumstances

Mef wrote:I guess the question for you to answer would be, do you really feel you derive sufficient additonal value from removing italicized games from your list (additional value that cannot be obtained via other methods previously mentioned of addressing the issue) that it merits going out of your way to introduce incorrectly scored games into the KGS database?

I derive a lot of pleasure from finishing my games on the day I play them. I can promise you I don't feel like I'm going out of my way at all. I believe any impact the occasional occurrence of it may have is inconsequential with regards of its impact on the system as a whole, and I know with some degree of certainty that deranking me will have a greater negative impact on the ranking system as a whole. As a result, the short answer to your question is "yes, I do", as I perceive the impact small and the benefit reasonably high, and with the way the current system works, I don't feel that value can be gained in other ways. As it stands, I find myself checking each game under the "Resume->" tab just in case they've come back on and sneakily moved every time I log on, and I don't like feeling like I have to do that very much at all. I would rather be able to just void the game, but that's not an option, so I'm left with resigning if I want it finished.

Kirby wrote:

xed_over wrote:...
if escaped games are eventually forfeited, then no it doesn't exclude a valid data point

Not all escaped games are eventually forfeited. You don't get a loss if you only forfeit once in awhile. Your rank is also calculated based on time. Games that happen further back in your game history affect your current rank less.

So getting an immediate win for an escaped game would provide more weight toward your rank, because if the win is delayed by a couple of months, for example, the win holds less weight.

This is partly true...it is true that the loss holds less weight over time, however once the forfeit is awarded it is used in calculating your rank no differently than if it had always been a loss. The only difference would be if there was a transient effect in how you selected the opponents you played during the period where the forfeit was not yet recorded. If you played identical opponents and got identical results in the meantime your rank would be identical whether the loss was immediate or the forfeit was delayed.