Originally posted by gezza
At least to make the penalty for running away and coming back again a lot of hard work to get your rating back up, if that is what you care about.

Ratings are not a reward, they are a barometer of chess skill. Even if a rating is a trophy, anyone who gets beat by a 1000 rated player who is really rated 1900 is getting shafted.

P-

Originally posted by Crowley
Guys! Guys!

Surely these rating floors will only affect the banded tournaments this person could enter and will have no bearing on actual rating calculations...

I hope they do affect rating calculations. That's the whole point of a rating floor.

Originally posted by gezza
We have a difference of opinion. Cool.

[b]Are banded tourneys all you care about? Do you have an answer for those who are playing clan challenges against 1000-rated players who are really 2000 strength?

I have seen other comments that clan matches are the responsibility of the clan leader. I would expect clan leaders to be irritated enough when som dropped again.

Cheers,
Gezza

ps. Edits, as bold, and inserts got messed up.[/b]

Yes, clan leaders sometimes boot those players who mass-timeout. Then again, they also seem to look the other way when a vastly underrated player joins. They don't seem to mind the extra points they get from this player winning games against players of vastly lesser skill.

In your open-invite example, the players are seeking someone rated much higher than they are, because that allows the open invite to fill automatically. They're not going to cull through the list of 1000-rated players looking for former 2000's.

From my experience with USCF and rating floors, I can assure you that rating floors rarely need to be dropped. The most common case is when a player is getting really old and can't maintain the performance level of their youth. It would not be that much work for RHP site admins [they could even have someone from the game mod team do this for them].

Guess again. My exaggerated example could put 100 times 64 points into the system with a rating floor. The maximum a player leaving can impact is 1200 less whatever his rating ends up at.

Yes, but your exaggerated example is atypical. I'm interested in knowing the actual inflation of rating floors vs. players leaving, not the worst case of each.

Rating floors won't encouragement game-dragging. How desperate do you have to be if you're playing on in hopes that your opponent will leave the site?! A blunder is much more likely. [And don't forget, under the current rules, you may still win a lost game, and get lots of rating points, if the opponent does leave the site].

I would rather have a few players drop 800 rating points, rather than devalue everyone's rating.

Better to spread a slight devaluation out over the whole rating pool, rather than unfairly penalizing the unlucky few who happen to be underrated's opponents.

I doubt that playing strength ever drops so much, but rating can, and should if the player plays in such a way as to merit it.

The point of ratings is to measure playing strength. If the playing strength has not dropped 800 points, then neither should the rating.

But now he has points he does not deserve. That can push his maximum up, and only allow entry to tournament bands which are too strong for him.

Hence the reason I have always proposed to not set the floor until X [say, 30] games are played at a certain rating class.

Really? With the current system, if someone resigns a bunch of games, comes back...

Let's say he never comes back.

...and then does the same again, he gives away points by resigning, takes them back by winning, gives them away be losing - repeat as often as you like with no inflation

There is still individual inflation in favor of those who win games they don't deserve to win.

don't want the rating system devalued. Ratings are actually a measure of likelihood of winning the game.

Exactly! So when you let someone get 800 points too low, it defeats the whole point of ratings. Ratings only mean something in comparison to other ratings. You could add 1000 points to everyone's rating, and they would still serve their purpose.

Historical values of rating give you an indication of strength if you want it

Yes. I want a real indication of strength. All the time. Not just when I decide to go researching.

Care to answer that one? At least with the present system, or a TER impact only, you have to regain all those points you dropped again.

Phlabibit already answered this. Ratings are not about punishment and reward. They are a metric.

Originally posted by SwissGambit
snip ... Ratings are not about punishment and reward. They are a metric.

They are a metric about likelihood of winning a game, not about playing strength. These are not the same thing, though without timeouts etc, they are related. They start to be the same thing when people play without resigning a whole bunch of games, and play at a consistent strength, however they achieve that.

It can be thought of as punishment and reward. If you want to see it that way, the reward of a realistic rating, is being able to play someone who you get a good game against. Punishment, if you want, is that if your rating tanks due to mass resignation you have to play a lot of games for it to be realistic again.

In the real world, with a limited number of spaces on any team, usually the best (most consistent, highest rated), get in and the rest end up on the second team, or further down.

I don't see the need for actual (as opposed to tournament entry) rating floors in the real world, and I don't see it here.

I'll answer your other points later this week - I have other things to do in the meantime.

Originally posted by gezza
They are a metric about likelihood of winning a game, not about playing strength.

They are not this when your rating goes down 1200 points. Thus a floor, like USCF is a wise decision for any game... not just tournaments.

I still don't see any difference in inflation vs. deflation of ratings based on users tanking 100's of games. Either way, ratings are not going to jump or drop.

When I play a 1900 rated player, I want to know that's what they are. If they are rated 800, that is not who they are regardless of if it's a tourney, clan, siege, or social game.

P-

Originally posted by Phlabibit
They are not this when your rating goes down 1200 points. Thus a floor, like USCF is a wise decision for any game... not just tournaments.

I still don't see any difference in inflation vs. deflation of ratings based on users tanking 100's of games. Either way, ratings are not going to jump or drop.

When I play a 1900 rated player, I want to know t that is not who they are regardless of if it's a tourney, clan, siege, or social game.

P-

But they may do so again. Having a rating of 1900 would be wrong because the use in question is likely to resign the game. A lower rating is appropriate, because the resigner has lost a bunch of games. A 1900 rated player is someone who plays well, and consistently - if they do not, they are not rated 1900.

I am not sure what you are trying to say with your second comment line. Ratings change all the time based on whether you win or lose games.

Originally posted by SwissGambit

snip...
Really? With the current system, if someone resigns a bunch of games, comes back...

Let's say he never comes back.

The principle is the same. If you read the first page of this thread, you will see a comment from Phlabibit concerning V graphs. The concern there is people who leave and come back.

If someone just leaves, with the current system they put a number of points into the system which is equal to the difference between the rating with which they end up, and the starting rating. If they come back, they start again where they left off, and overall the system is intact.

With a floor below which your rating cannot fall, someone leaving puts a number of points into the system proportional to the number of games they resign. They is already an issue if they never come back, as the number of points put into the system is more than with the current system - likelihood of rating inflation is increased. With repeats of this behaviour, each bunch of resignations puts more points into the system. I see this as more of a problem.

Where I disagree with you and Phlabibit, is that you appear only to be concerned with the case where a player with a high ability, but low rating wins games against a higher rated player - causing that player to "unfairly" lose rating points. To me, giving any player, including a lower rated player, a bunch of rating points for nothing by resigning is just as "unfair", and you give them more points with a rating floor.

And as for the case where people "spitefully" resign a bunch of games so as not to give rating points to someone who is winning, I really don't see the issue. You have not yet won the game. Your opponent resigns a bunch of games, with yours amongst them. To say that you "deserve" the rating points without your opponent having actually lost the game is taking things a bit far. I have also yet to see any occurrences of this - care to give some pointers?

Originally posted by gezza
The principle is the same. If you read the first page of this thread, you will see a comment from Phlabibit concerning V graphs. The concern there is people who leave and come back.

If someone just leaves, with the current system they put a number of points into the system which is equal to the difference between the rating with which they end up, and the ...[text shortened]... bit far. I have also yet to see any occurrences of this - care to give some pointers?

To me, giving any player, including a lower rated player, a bunch of rating points for nothing by resigning is just as "unfair", and you give them more points with a rating floor.

It is not "for nothing". Resigning is losing the game, just as getting checkmated is.

Your opponent resigns a bunch of games, with yours amongst them. To say that you "deserve" the rating points without your opponent having actually lost the game is taking things a bit far.

I'm sensing a fundamental misunderstanding of the concept of resignation on your part. If your opponent resigns, he HAS actually lost the game, by definition.

Originally posted by SwissGambit
To me, giving any player, including a lower rated player, a bunch of rating points for nothing by resigning is just as "unfair", and you give them more points with a rating floor.

It is not "for nothing". Resigning is losing the game, just as getting checkmated is.

Your opponent resigns a bunch of games, with yours amongst them. To say t ion on your part. If your opponent resigns, he HAS actually lost the game, by definition.

So you object to not getting any rating points for winning a game through resignation where you have an advantage, but which you have not yet won, because your opponent's rating drops correctly because he resigns a whole bunch of games before resigning against you, but you have no objection to getting rating points in a game where you are are at a disadvantage but your opponent resigns?

You appear have a fundamental misconception that the rating system is a perfect measure of ability, rather than a metric of approximate likelihood of winning a game.

Sweet. From a difference of opinion, we move on to questioning the other guy's ability to understand basic concepts. Once again I see the futility of trying to present logical arguments to someone who is already convinced in an internet forum. You'll excuse me (or not - see how much I care) if I chose not to reply to your future arguments.


edit: everything was bold.

Originally posted by gezza
So you object to not getting any rating points for winning a game through resignation where you have an advantage, but which you have not yet won, because your opponent's rating drops correctly because he resigns a whole bunch of games before resigning against you, but you have no objection to getting rating points in a game where you are are at a disadvantage uch I care) if I chose not to reply to your future arguments.


edit: everything was bold.

So you object to not getting any rating points for winning a game through resignation where you have an advantage, but which you have not yet won,

No, let the rating system + floor determine this. Even with a rating floor, certain players may get nothing for winning such a game [say if they were already 200+ points higher rated to start and the other guy drops 199 points]. It at least offers some protection for a lower-rated player who scores an upset.

because your opponent's rating drops correctly because he resigns a whole bunch of games before resigning against you,

Is he deliberately resigning games he isn't losing at all to screw me out of rating points? That's sandbagging, not a 'correct' drop in rating.

but you have no objection to getting rating points in a game where you are are at a disadvantage but your opponent resigns?

Of course not. It isn't my fault that my opponent decided to resign.

You appear have a fundamental misconception that the rating system is a perfect measure of ability, rather than a metric of approximate likelihood of winning a game.

My remark earlier appears to have stung a bit. Sorry, but this is easily brushed aside. Rating floors obviously do not effect a perfect measure of ability; they just keep the rating in the same ballpark.

You'll excuse me (or not - see how much I care) if I chose not to reply to your future arguments.

Fine with me - I believe the site owner already decided to implement floors anyway.

Originally posted by SwissGambit
So you object to not getting any rating points for winning a game through resignation where you have an advantage, but which you have not yet won,

No, let the rating system + floor determine this. Even with a rating floor, certain players may get nothing for winning such a game [say if they were already 200+ points higher rated to start and the other ...[text shortened]... s.

Fine with me - I believe the site owner already decided to implement floors anyway.

No, let the rating system + floor determine this. Even with a rating floor, certain players may get nothing for winning such a game [say if they were already 200+ points higher rated to start and the other guy drops 199 points]. It at least offers some protection for a lower-rated player who scores an upset.

Why not let the existing system without a floor decide it? As you say, some players get no points anyway, and a lower rated player can get some points even as the opponent's rating falls. Rating has to fall a long way for you to get nothing at all, 669 points, if Xanthos calculated correctly some time ago.

Is he deliberately resigning games he isn't losing at all to screw me out of rating points? That's sandbagging, not a 'correct' drop in rating.

Let's just say that he does it in alphabetical order, and S happens to be near the end? Is that sandbagging? He has just (according to your resigning == lost definition) "lost" a bunch of games, so his rating (which is likelihood of winning) should fall - why stop it falling artificially?

but you have no objection to getting rating points in a game where you are are at a disadvantage but your opponent resigns?

Of course not. It isn't my fault that my opponent decided to resign.

Does this not strike you as even a little inconsistent? If my opponent has an advantage - I may be playing on because I do not yet judge it to be a winning advantage - and then resigns, he gives me rating points I do not want, because they now mean my rating does not reflect my ability. I try very hard to avoid losing to timeouts, etc. in the hope that the conditions I mentioned in previous posts for rating to reflect ability still hold. Someone dumping points on me just devalues those points.

My remark earlier appears to have stung a bit. Sorry, but this is easily brushed aside. Rating floors obviously do not effect a perfect measure of ability; they just keep the rating in the same ballpark.

I just find it somewhat a shame when internet debates degenerate into insults. Which is precisely why I used the same terms and followed up with the comment I did. I am thicker skinned than you may think, but equally I will not carry on indefinitely.

Fine with me - I believe the site owner already decided to implement floors anyway.

With me too - It should have been a few days ago, so perhaps our debate has shown it to be a contentious issue, which should not be implemented so hastily. I do not recall him stating whether TER only, or actual rating would have been floored.

Originally posted by gezza
[b]No, let the rating system + floor determine this. Even with a rating floor, certain players may get nothing for winning such a game [say if they were already 200+ points higher rated to start and the other guy drops 199 points]. It at least offers some protection for a lower-rated player who scores an upset.

Why not let the existing system without a ...[text shortened]... I do not recall him stating whether TER only, or actual rating would have been floored.[/b]

Why not let the existing system without a floor decide it?

Already answered.

Rating has to fall a long way for you to get nothing at all, 669 points, if Xanthos calculated correctly some time ago.

In a traditional ELO system, it takes a 400 point difference before the winner does not get points for winning the game. I don't know if RHP is using the old ELO, or a modern version like Glicko.

Let's just say that he does it in alphabetical order, and S happens to be near the end? Is that sandbagging?

It could be, if the player later uses his lower rating to unfairly win banded tournaments and clan match games.

He has just (according to your resigning == lost definition) "lost" a bunch of games, so his rating (which is likelihood of winning) should fall - why stop it falling artificially?

My resigning = lost definition? I almost hate to tell you this, but this is consensus in the chess world. [Then again, many on RHP tend to go their own way and ignore the way things are in the rest of the chess world...]

Again, your question has already been answered.

Does this not strike you as even a little inconsistent?

No.

If my opponent has an advantage - I may be playing on because I do not yet judge it to be a winning advantage - and then resigns, he gives me rating points I do not want, because they now mean my rating does not reflect my ability.

Let's say in the same position, instead of just resigning, your opponent makes a crude blunder which instantly loses [and he resigns without playing another move]. How is that substantively different? After all, it wasn't your ability that got you the win. Would you still not want the points?

I try very hard to avoid losing to timeouts, etc. in the hope that the conditions I mentioned in previous posts for rating to reflect ability still hold.

Time is part of the game, and time management is a necessary ability in serious chess.

I just find it somewhat a shame when internet debates degenerate into insults.

Sorry, but you clearly made two remarks back-to-back that indicated that you had a radically different view of resignation than most serious chess players. Your most recent post indicates that you've stuck to your position. The concept of what resignation means is critical to the debate at hand. My comments were not intended as an insult.

With me too - It should have been a few days ago, so perhaps our debate has shown it to be a contentious issue, which should not be implemented so hastily.

The US Chess Federation has used rating floors for years. They are already a proven deterrent to sandbagging, and do not affect the rating of legitimate players much.

I do not recall him stating whether TER only, or actual rating would have been floored.

I have sent feedback in hopes of learning what has actually changed, or will be changed.

Originally posted by SwissGambit
Why not let the existing system without a floor decide it?

snip...

I have sent feedback in hopes of learning what has actually changed, or will be changed.[/b]

Why not let the existing system without a floor decide it?
Already answered.

I must have blinked. You said "Even with a rating floor, certain players may get nothing for winning such a game". I pointed out that your rating has to fall a very long way for your opponent to get nothing at all. Therefore I see little benefit in introducing a system which inflates the ratings of the site just so that you get a few more points when your opponent resigns. What is the benefit of not using the existing system?

Rating has to fall a long way for you to get nothing at all, 669 points, if Xanthos calculated correctly some time ago.

In a traditional ELO system, it takes a 400 point difference before the winner does not get points for winning the game. I don't know if RHP is using the old ELO, or a modern version like Glicko.

You argue that the RHP rating system should be changed without understanding what it currently does? How amusing!


It could be, if the player later uses his lower rating to unfairly win banded tournaments and clan match games.

Putting a floor on the TER would solve the first, and I have already given you a solution to the second. Neither of these causes inflation of the rating system. And if it is just someone leaving for a while, that is hardly sandbagging.


Me:
He has just (according to your resigning == lost definition) "lost" a bunch of games, so his rating (which is likelihood of winning) should fall - why stop it falling artificially?

You:
My resigning = lost definition? I almost hate to tell you this, but this is consensus in the chess world. [Then again, many on RHP tend to go their own way and ignore the way things are in the rest of the chess world...]

You appear to be arguing for a special case for mass resignations, to be treated differently to losing one game after another. I can cope with "resign=lost". I assigned ownership to you to emphasise that you at least appear to agree with it. I argue that it should be treated the same as a loss - one should lose rating points, always. You are the one arguing for something else, but you do not justify it, even though you appear to agree that in principle a resignation is the same as a loss.

Again, your question has already been answered.

I had not noticed an answer to why rating points should not fall for mass resignations, but ought to if you start playing differently so that your apparent ability changes.

Does this not strike you as even a little inconsistent?

No.

Now there's a surprise. Funnily enough, it does appear inconsistent to me. I say why later. I bet that surprises you too.

Let's say in the same position, instead of just resigning, your opponent makes a crude blunder which instantly loses [and he resigns without playing another move]. How is that substantively different? After all, it wasn't your ability that got you the win. Would you still not want the points?

If I put him under enough pressure that he makes a blunder, there is a significant difference. Good players rarely blunder - that is part of the definition of being good. But it does happen, and yes, I want the points, as I have kept the game complicated enough while at a disadvantage for the blunder to happen. Keeping the game complicated is a reflection of ability.

Time is part of the game, and time management is a necessary ability in serious chess.

Agreed. So what has that got to do with someone who doesn't care about rating and goes on holiday and gets timed out, or decides to resign a few hundred games? What has that got to do with me saying I manage time to the best of my ability so that my rating means something?

Sorry, but you clearly made two remarks back-to-back that indicated that you had a radically different view of resignation than most serious chess players.

BS. My remarks treat resignation (including mass resignation) as just another loss. I argue that it should be treated no differently. I emphasise that you agree, except when it causes you to get very few rating points for a win against a resignation.

Your most recent post indicates that you've stuck to your position. The concept of what resignation means is critical to the debate at hand. My comments were not intended as an insult.

Picking up on the details of language use is a common way to argue, especially on the internet. Which is why I only bother to argue on the internet when I am in a masochistic mood.


I've noticed that:
1) You want rating points if your opponent resigns against you.
2) You don't want your opponent to lose rating points by resigning, because that would mean you get less rating points for your win against their later resignation.

And yes, in point 2 I assume that your opponent has already resigned enough games for him to reach the rating floor you want to implement.

I find the above two points inconsistent, and that they lead to rating inflation for the whole system. Try and argue otherwise.

The US Chess Federation has used rating floors for years. They are already a proven deterrent to sandbagging, and do not affect the rating of legitimate players much.

Is this just a US thing? From: http://www.exeterchessclub.org.uk/bcftable.html
"USCF and ELO have the same form but I believe USCF ratings run a little higher than ELO: say, about 100 points on average"

I have seen the same sort of comment elsewhere. Looks like rating inflation to me. And if it puts everyone's rating up by 100 points, I'd say that they do affect the rating of legitimate players.

I have sent feedback in hopes of learning what has actually changed, or will be changed.

It will be good to know. I am sure it will all happen in good time. Here's my vote for a TER floor.

Originally posted by gezza
[b]Why not let the existing system without a floor decide it?
Already answered.

I must have blinked. You said "Even with a rating floor, certain players may get nothing for winning such a game". I pointed out that your rating has to fall a very long way for your opponent to get nothing at all. Therefore I see little benefit in introducing a sy I am sure it will all happen in good time. Here's my vote for a TER floor.[/b]

I must have blinked.

Hint: Discouraging spite-resignations isn't the main point of rating floors. It is only an small side benefit. There is a more important consideration in play.

You argue that the RHP rating system should be changed without understanding what it currently does? How amusing!

This from the guy who quotes the calculations of others on ratings matters? 🙄 I know exactly how the traditional ELO system works. The Glicko is like the ELO, but it takes activity into account. But talk is cheap. Let's see you explain why 669 points is the magic number.

Putting a floor on the TER would solve the first, and I have already given you a solution to the second.

The only 'solution' you have offered is that the clan leader should boot the 'offender' in the case of a mass-resignation. But this is flawed on many levels. First, a mass-resigner is not necessarily an 'offender' - life takes priority over chess sometimes. Second, it traps the clan leader in a no-win situation when such a player wants to come back. They either must refuse the application, or share the blame for sandbagging. Third, the player themself may be refused entry back into their clan because of the concern over sandbagging.

And if it is just someone leaving for a while, that is hardly sandbagging.

Read it again - it is not that they leave, it is that they come back and unfairly win clan challenges and banded tournaments.

You appear to be arguing for a special case for mass resignations, to be treated differently to losing one game after another.

No. In both cases, a rating floor should apply.

I argue that it should be treated the same as a loss - one should lose rating points, always.

Which is a flawed argument - the obvious counterexample is a 1000 player losing to a 2000 player. 1000 does not lose any points. This is an important precedent: A loss in a game does not always justify a loss of rating points.

You are the one arguing for something else, but you do not justify it, even though you appear to agree that in principle a resignation is the same as a loss.

Sandbagging, and stopping the skewing of rating data due to mass-timeouts, or mass-resignations [which do NOT reflect chess ability at all] are perfectly good justifications for floors. I have repeated this over and over both here and in other threads.

I had not noticed an answer to why rating points should not fall for mass resignations, but ought to if you start playing differently so that your apparent ability changes.

First, let's clear up your wording a bit. In the event of a mass-resignation, the player's rating should faIl until it hits the rating floor. Second, I stated earlier that a) it is extremely rare for someone's playing strength to actually drop hundreds of points and b) in the rare event that the drop in strength is real, a player can petition a volunteer, like the game mods, to lower their floor.

Good players rarely blunder - that is part of the definition of being good. But it does happen, and yes, I want the points, as I have kept the game complicated enough while at a disadvantage for the blunder to happen. Keeping the game complicated is a reflection of ability.

And yet some blunders happen without any real pressure, in an uncomplicated game. Even good players are mortal. The point remains that such a blunder is just as much an undeserved gift as a win by resignation due to the opponent leaving the site. It is inconsistent for you to accept the former, yet eschew the latter.

Agreed. So what has that got to do with someone who doesn't care about rating and goes on holiday and gets timed out, or decides to resign a few hundred games? What has that got to do with me saying I manage time to the best of my ability so that my rating means something?

You said, "I try very hard to avoid losing to timeouts, etc. in the hope that the conditions I mentioned in previous posts for rating to reflect ability still hold.", which implies that if you DID lose on time, it would mean that rating did not reflect ability as well as it should. I dispute this - time management is part of the game.

BS. My remarks treat resignation (including mass resignation) as just another loss. I argue that it should be treated no differently. I emphasise that you agree, except when it causes you to get very few rating points for a win against a resignation.

Nice try at weaseling out, but you earlier said,

"And as for the case where people "spitefully" resign a bunch of games so as not to give rating points to someone who is winning, I really don't see the issue. You have not yet won the game. Your opponent resigns a bunch of games, with yours amongst them. To say that you "deserve" the rating points without your opponent having actually lost the game is taking things a bit far."

This quote is full of misconceptions.
1) If your opponent resigns, he HAS "actually lost the game". That's the correct meaning of resignation.
2) The rating points ARE richly deserved in this example. The whole point is that the spite-resigner is doing this because he knows he is lost, and he intends to resign because of it. He just doesn't want his opponent to get rating points.

Picking up on the details of language use is a common way to argue, especially on the internet.

The concept of what resignation means is not some nit-picky side issue. It is a core issue in this discussion.

I've noticed that:
1) You want rating points if your opponent resigns against you.
2) You don't want your opponent to lose rating points by resigning, because that would mean you get less rating points for your win against their later resignation.

Neither observation is fully accurate.
1) I don't expect points if I beat a player very much lower rated than myself. Other than that, yes, I expect rating points.
2) This is way off. Of course I expect that my opponent will lose points if they resign other games [again, with the caveat that rating differences aren't too great]. I just don't think that a sudden drop of hundreds of points is warranted.

Looks like rating inflation to me. And if it puts everyone's rating up by 100 points, I'd say that they do affect the rating of legitimate players.

The main point of ratings is to compare the strength of two players in the same rating pool. Even if the 'master' rating became 2300 and 'expert' became 2100, masters would still be 200 points stronger than experts. I fail to see why this is a problem.

Originally posted by SwissGambit
I must have blinked.

snip ...

The main point of ratings is to compare the strength of two players in the same rating pool. Even if the 'master' rating became 2300 and 'expert' became 2100, masters would still be 200 points stronger than experts. I fail to see why this is a problem.

Hint: Discouraging spite-resignations isn't the main point of rating floors. It is only an small side benefit. There is a more important consideration in play.

Really. What's that then?

Me:
You argue that the RHP rating system should be changed without understanding what it currently does? How amusing!

You:
This from the guy who quotes the calculations of others on ratings matters? 🙄 I know exactly how the traditional ELO system works. The Glicko is like the ELO, but it takes activity into account. But talk is cheap. Let's see you explain why 669 points is the magic number.

Xanthos has already done it. I know the formula. I even checked his calculations a long time ago. But he did the calculations, not me, so gets credit for them. I know how Glicko works. But you have not bothered to find out what is in use on RHP before saying it should be changed.
ronthflmao!!

The only 'solution' you have offered is that the clan leader should boot the 'offender' in the case of a mass-resignation. But this is flawed on many levels. First, a mass-resigner is not necessarily an 'offender' - life takes priority over chess sometimes. Second, it traps the clan leader in a no-win situation when such a player wants to come back. They either must refuse the application, or share the blame for sandbagging. Third, the player themself may be refused entry back into their clan because of the concern over sandbagging.

Either he is sandbagging or not. The clan leader needs to decide, and decide what to do about it. It would be feasible to deny the person entry until their rating has stabilised again. But that is a clan leader issue - accept them back and take the flack, or do not, and take the flack from the player. Leaders have to make tough decisions sometimes. It comes with the job.

Me:
And if it is just someone leaving for a while, that is hardly sandbagging.
You:
Read it again - it is not that they leave, it is that they come back and unfairly win clan challenges and banded tournaments.

"Leave for a while" implies come back. Read it again yourself!! Clan challenges remain the clan leaders problem, and believe it or not, a TER floor solves the banded tourney problem.

You appear to be arguing for a special case for mass resignations, to be treated differently to losing one game after another.

No. In both cases, a rating floor should apply.

Or in neither case. After all, losing a game (against equally rated opposition) shows that your ability and your rating are not in sync. Therefore your rating should drop.

Me:
I argue that it should be treated the same as a loss - one should lose rating points, always.
You:
Which is a flawed argument - the obvious counterexample is a 1000 player losing to a 2000 player. 1000 does not lose any points. This is an important precedent: A loss in a game does not always justify a loss of rating points.

Oh yes. I forgot. When there is already more than 669 points of difference in rating, there is no gain and no loss. I could have qualified it by saying, "If the calculations of the elo system say they should", rather than "always".

Sandbagging, and stopping the skewing of rating data due to mass-timeouts, or mass-resignations [which do NOT reflect chess ability at all] are perfectly good justifications for floors. I have repeated this over and over both here and in other threads.

Why not? They reflect ability to manage time, which is, as you say, part of the game. The likelihood of winning is reflected in a rating as is, without a floor. A player who drops due to timeouts or resignations may do so again, and his rating ought to reflect that. Just because you repeat it does not mean you are right. YOU COULD EVEN SHOUT, AND NOT BE RIGHT.

First, let's clear up your wording a bit. In the event of a mass-resignation, the player's rating should faIl until it hits the rating floor. Second, I stated earlier that a) it is extremely rare for someone's playing strength to actually drop hundreds of points and b) in the rare event that the drop in strength is real, a player can petition a volunteer, like the game mods, to lower their floor.

You did. You correctly understood what I wrote. However, if someone really, desperately wants to sandbag, they just have to convince an admin. I am sure it can be done once or maybe even twice. Fake quote: "Oh I started having more time again this year, so I was able to play better and win the tournament"


And yet some blunders happen without any real pressure, in an uncomplicated game. Even good players are mortal. The point remains that such a blunder is just as much an undeserved gift as a win by resignation due to the opponent leaving the site. It is inconsistent for you to accept the former, yet eschew the latter.

There are comments elsewhere about players offering a draw to someone who has played well to a point, and then messed up. I could find them, if I was really bothered. I think that good players blundering without complications is at least as rare as your example of players dropping hundreds of rating points (of ability) without reason. When I look at the top 20 on the site, 5 of them have lost less than 10 games. Yes, I take the points. Yes, people are mortal. But that is why we are not all GMs. It is not inconsistent at all - I want the points when my opponent blunders - I have played better. I do not want the points if he can't be bothered (I know there are other reasons) any more - he has shown me nothing.

Agreed. So what has that got to do with someone who doesn't care about rating and goes on holiday and gets timed out, or decides to resign a few hundred games? What has that got to do with me saying I manage time to the best of my ability so that my rating means something?

You said, "I try very hard to avoid losing to timeouts, etc. in the hope that the conditions I mentioned in previous posts for rating to reflect ability still hold.", which implies that if you DID lose on time, it would mean that rating did not reflect ability as well as it should. I dispute this - time management is part of the game.

Rating would then reflect my inability to manage time mixed together with chess calculation/judgement, rather than purely chess calculation/judgement ability. I want my rating to reflect pure chess ability only, so I make sure that time is not a problem. I am aware that time can be a problem OTB, or in blitz, but to say that it is part of the game on RHP, when you can chose the number of games and the length of timeout/timebank you play here is stretching it. At very most it is a life style time management issue. I choose whether my rating includes me losing on time or not. I might lose on time someday, but not today.

BS. My remarks treat resignation (including mass resignation) as just another loss. I argue that it should be treated no differently. I emphasise that you agree, except when it causes you to get very few rating points for a win against a resignation.

Nice try at weaseling out, but you earlier said,

"And as for the case where people "spitefully" resign a bunch of games so as not to give rating points to someone who is winning, I really don't see the issue. You have not yet won the game. Your opponent resigns a bunch of games, with yours amongst them. To say that you "deserve" the rating points without your opponent having actually lost the game is taking things a bit far."

This quote is full of misconceptions.
1) If your opponent resigns, he HAS "actually lost the game". That's the correct meaning of resignation.
2) The rating points ARE richly deserved in this example. The whole point is that the spite-resigner is doing this because he knows he is lost, and he intends to resign because of it. He just doesn't want his opponent to get rating points.

"weaseling"?? Has he not also "lost" the games he resigned before yours? He is giving rating points to someone. His rating falls. Then he resigns against you. He has NOT "actually lost the game" UNTIL he resigns. That means that if he has a low rating WHEN he resigns against you, you (as much higher rated) get nothing. You managed to beat someone who resigns a lot of games, and his rating WHEN he resigned reflects that. I still wait for your example of this actually happening.


The concept of what resignation means is not some nit-picky side issue. It is a core issue in this discussion.

So does it mean he loses, or not? The core is that a loss (except to a much higher rated opponent) should mean that you drop rating points - you are on average less likely to win a game than before.

Neither observation is fully accurate.
1) I don't expect points if I beat a player very much lower rated than myself. Other than that, yes, I expect rating points.
2) This is way off. Of course I expect that my opponent will lose points if they resign other games [again, with the caveat that rating differences aren't too great]. I just don't think that a sudden drop of hundreds of points is warranted.

At least you admit there is a hint of accuracy. A drop of hundreds of points is warranted if your opponent loses hundreds of games. The consequence of losing a game (against a similarly rated or weaker opponent) is to lose rating points.

The main point of ratings is to compare the strength of two players in the same rating pool. Even if the 'master' rating became 2300 and 'expert' became 2100, masters would still be 200 points stronger than experts. I fail to see why this is a problem.

The main point is likelihood of winning a game. That is what rating measures. It is related to ability, but not the same thing.

The problem is inflation. Just as in the hig...