Complaints about sudden death and 5-second delay

[DENTONCHESS]

@Micah Smith

@Tim Just

Perhaps for the 7th edition you could be a proof reader?

Sure

@Tim Just

Is the substance of 14H1, 14H2, or 14H5 changed by wording/formatting suggested by the publisher (and accepted by moi) in the published version?

What is in the 6th edition of the rulebook regarding these rules and and in the update to the 5th edition online is different.

14H1-the 6th edition states that a player “may make a claim of insufficient losing chances” but the update to the 5th edition online states “ask the director to declare the game a draw on the grounds that the player has insufficient losing chances”

14H2-the 6th edition explains that you put a delay clock on the game if the draw offer is not immediately accepted and a properly set delay clock is immediately available but the update to the 5th edition online explains that you only put a delay clock on the game is the claim is unclear and a delay clock is available. Also, the 6th edition states that the TD will inform the opponent that a draw offer has been made but the update to the 5th edition online does not state this.

14H5-the 6th edition mentions increment but the update to the 5th edition online does not.

Also, 14H2a and 14H2b in the 6th edition should probably mention increment in addition to delay.

A very good way around this is to simply pre-publish that there
will be no time delay, whatsoever set at any time. There indeed are many who have voiced their opinion that delay and
increment has not added to the sport.
Rob Jones

[GeneM]

@DENTONCHESS

… pre-publish that there will be no time delay, whatsoever set at any time. There indeed are many who have voiced their opinion that delay and increment has not added to the sport.
Rob Jones

The d5 or i5 seconds are indeed silly for the masses of class players, where 30 seconds is needed for class players to make a sensible series of endgame moves. 5 seconds apparently makes sense in grandmaster games, but that is no reason to mimic that duration in class level games.

Timed chess has a glaring practical weakness in that the length of a normal game varies enormously, from perhaps 20-66 move-pairs usually, with some games lasting longer. And, there is no possible way that either player can know early in the game whether the game will have lots of move-pairs. That unknowable-ness requires either
(A) No SuddenDeath/#minutes segment in the time control (where Game/#minutes time control is a form of SD); or
(B) Use of a delay or increment that is not absurdly short for the skill level of class players (the way 5 seconds is absurdly short for class players).

Without an increment or delay, any time control that has a SuddenDeath final segment potentially ruins the chess game by ruining the endgame, and therefore ruining all middle-game strategies that involve the endgame.

At the class level, instead of the two-segment time control of
40moves/90minutes + d5seconds & SD/60minutes +d5seconds
,
try
40moves/70minutes + i30seconds & SD/30minutes + i30seconds
.

In any case, designing a time control that harms the concept of the endgame does more damage to the game than would some other rule suggestions that people instantly reject, IMHO.

The integrity of the USCF Elo rating system is compromised when Game/60minutes with no delay or increment (or a brief 5 seconds d i) is used as the time control. Such a time control makes in improper shift in the proper balance of chess, to the disadvantage of the player who might know endgames better than he knows middlegame shot tactics.

I realize that at the scholastic level the endgame rarely matters. But for adult class level rated play, Game/60minutes with no delay is dirty to the Elo rating; unless someone comes us with a way to know in advance how many move-pairs the game will require.
.

[jwiewel]

@GeneM

@DENTONCHESS

… pre-publish that there will be no time delay, whatsoever set at any time. There indeed are many who have voiced their opinion that delay and increment has not added to the sport.
Rob Jones

The d5 or i5 seconds are indeed silly for the masses of class players, where 30 seconds is needed for class players to make a sensible series of endgame moves. 5 seconds apparently makes sense in grandmaster games, but that is no reason to mimic that duration in class level games.

Timed chess has a glaring practical weakness in that the length of a normal game varies enormously, from perhaps 20-66 move-pairs usually, with some games lasting longer. And, there is no possible way that either player can know early in the game whether the game will have lots of move-pairs. That unknowable-ness requires either
(A) No SuddenDeath/#minutes segment in the time control (where Game/#minutes time control is a form of SD); or
(B) Use of a delay or increment that is not absurdly short for the skill level of class players (the way 5 seconds is absurdly short for class players).

Without an increment or delay, any time control that has a SuddenDeath final segment potentially ruins the chess game by ruining the endgame, and therefore ruining all middle-game strategies that involve the endgame.

At the class level, instead of the two-segment time control of
40moves/90minutes + d5seconds & SD/60minutes +d5seconds
,
try
40moves/70minutes + i30seconds & SD/30minutes + i30seconds
.

In any case, designing a time control that harms the concept of the endgame does more damage to the game than would some other rule suggestions that people instantly reject, IMHO.

The integrity of the USCF Elo rating system is compromised when Game/60minutes with no delay or increment (or a brief 5 seconds d i) is used as the time control. Such a time control makes in improper shift in the proper balance of chess, to the disadvantage of the player who might know endgames better than he knows middlegame shot tactics.

I realize that at the scholastic level the endgame rarely matters. But for adult class level rated play, Game/60minutes with no delay is dirty to the Elo rating; unless someone comes us with a way to know in advance how many move-pairs the game will require.
.

A three-second delay (for quick chess) may not be enough for score-keeping, but just this July it allowed my to readily win a multi-piece ending that I would have otherwise flagged on.
The goal was that if the position was one that could be easily held then those few seconds would be enough to do so for most people.

[Crume]

Warning: this post has nothing to do with 14H. It’s safe to ignore it if you’re here for 14H.

@GeneM

@DENTONCHESS

… pre-publish that there will be no time delay, whatsoever set at any time. There indeed are many who have voiced their opinion that delay and increment has not added to the sport.
Rob Jones

The d5 or i5 seconds are indeed silly for the masses of class players, where 30 seconds is needed for class players to make a sensible series of endgame moves. 5 seconds apparently makes sense in grandmaster games, but that is no reason to mimic that duration in class level games.

“Apparently”…why? I don’t see why that makes sense for GMs. Top level FIDE tournaments are played with a 30 second increment. Organizers use delay b/c they are more comfortable with it and it offers a “more sure” way of timing rounds.

@GeneM

Timed chess has a glaring practical weakness in that the length of a normal game varies enormously, from perhaps 20-66 move-pairs usually, with some games lasting longer. And, there is no possible way that either player can know early in the game whether the game will have lots of move-pairs. That unknowable-ness requires either
(A) No SuddenDeath/#minutes segment in the time control (where Game/#minutes time control is a form of SD); or
(B) Use of a delay or increment that is not absurdly short for the skill level of class players (the way 5 seconds is absurdly short for class players).

Misguided premise (bordering on absurd). Grading a game on the number of move pairs is like grading a program on the lines of code it took to build it. It’s meaningless.

@GeneM

Without an increment or delay, any time control that has a SuddenDeath final segment potentially ruins the chess game by ruining the endgame, and therefore ruining all middle-game strategies that involve the endgame.

And since all middle games have the potential to become endgames, this invalidates the openings too which utterly destroys chess and the universe as we know it. Oh come’on now, that is fanciful.

@GeneM

At the class level, instead of the two-segment time control of
40moves/90minutes + d5seconds & SD/60minutes +d5seconds
,
try
40moves/70minutes + i30seconds & SD/30minutes + i30seconds
.

“X moves in X minutes” is used in very few tournaments. Two-segment controls are used in very few tournaments. What problem are you solving for in this?

@GeneM

In any case, designing a time control that harms the concept of the endgame does more damage to the game than would some other rule suggestions that people instantly reject, IMHO.

Using the anonymous strawman of “other rule suggestions” doesn’t strengthen a weak premise.

@GeneM

The integrity of the USCF Elo rating system is compromised when Game/60minutes with no delay or increment (or a brief 5 seconds d i) is used as the time control. Such a time control makes in improper shift in the proper balance of chess, to the disadvantage of the player who might know endgames better than he knows middlegame shot tactics.

So…you’re saying G/65 d/0…which is acceptable by today’s Rating Committee rules…damages the integrity of the USCF rating system? If yes, then I’d say the rest of the argument needs no discussion b/c the premise is batty.

@GeneM

I realize that at the scholastic level the endgame rarely matters. But for adult class level rated play, Game/60minutes with no delay is dirty to the Elo rating; unless someone comes us with a way to know in advance how many move-pairs the game will require.
.

In his wisdom, Dr. Elo did not account for number of move pairs in his rating calculation. A likely reason is that move pairs have little to nothing to do with the efficacy of a rating system.

[dennisk]

@Crume

In his wisdom, Dr. Elo did not account for number of move pairs in his rating calculation. A likely reason is that move pairs have little to nothing to do with the efficacy of a rating system.

Unfair argument. In Dr. Elo’s day, there was no such thing as a sudden death time control, except for blitz, and that wasn’t rated. Serious rated chess games used a repeating time control, such as 40/2, 20/1, 20/1, …, with adjournment if necessary. So you always knew how much time per move you had, and because of that, the number of moves didn’t matter. In sudden death, it matters. In G/60, for instance, there is an enormous difference between a 60 move game (1 minute per move) and a 20 move game (3 minutes per move) – except that you don’t know at the beginning of the game which of those it’s going to be, so it’s impossible to budget your time properly. The intent of delay or increment is that if you can reach an easily won (or easily drawn endgame), you are guaranteed enough time to win (or draw) that game. This is some compensation at least for the uncertainty mentioned above. I absolutely would not play a rated G/60 or shorter time control without delay or increment, for the very reasons GeneM gives.

[Crume]

Warning: another post that has little to do with 14H, although there is a vague connection due to time management.

@dennisk

@Crume

In his wisdom, Dr. Elo did not account for number of move pairs in his rating calculation. A likely reason is that move pairs have little to nothing to do with the efficacy of a rating system.

Unfair argument. In Dr. Elo’s day, there was no such thing as a sudden death time control, except for blitz, and that wasn’t rated. Serious rated chess games used a repeating time control, such as 40/2, 20/1, 20/1, …, with adjournment if necessary. So you always knew how much time per move you had, and because of that, the number of moves didn’t matter. In sudden death, it matters. In G/60, for instance, there is an enormous difference between a 60 move game (1 minute per move) and a 20 move game (3 minutes per move) – except that you don’t know at the beginning of the game which of those it’s going to be, so it’s impossible to budget your time properly. The intent of delay or increment is that if you can reach an easily won (or easily drawn endgame), you are guaranteed enough time to win (or draw) that game. This is some compensation at least for the uncertainty mentioned above. I absolutely would not play a rated G/60 or shorter time control without delay or increment, for the very reasons GeneM gives.

First, you still aren’t explaining how number of move pairs in a game validate a rating system.

Second, I believe you’d like to invalidate my argument rather than call it unfair. I’ve claimed that Gene’s premise is invalid. So…you have to show his position to be valid, ie true, to prove me wrong. So show me why move pairs matter to a rating system.

Shifting to the time management discussion.

I agree that a d/0 or inc/0 game has a different feel than d/+ or inc/+. Changes the strategy but doesn’t change the fundamental rules of the game (much like the difference in the NFL between regulation and overtime). Time management for a SD d/0 time control is not impossible. You demonstrated move/time averaging and then tried to claim impossibility b/c of an uncertainty. It’s OK if you don’t like uncertainty, yet that dislike doesn’t make a time control (or rating system) any less valid to a rating system. At present, the USCF considers G/30 with no increment the minimum time for a serious chess game as it impacts one’s “Regular” rating. If you feel that’s not right, please introduce a Delegate Motion for 2015 to change the allowable time controls.

[dennisk]

@Crume

Warning: another post that has little to do with 14H, although there is a vague connection due to time management.

@dennisk

@Crume

In his wisdom, Dr. Elo did not account for number of move pairs in his rating calculation. A likely reason is that move pairs have little to nothing to do with the efficacy of a rating system.

Unfair argument. In Dr. Elo’s day, there was no such thing as a sudden death time control, except for blitz, and that wasn’t rated. Serious rated chess games used a repeating time control, such as 40/2, 20/1, 20/1, …, with adjournment if necessary. So you always knew how much time per move you had, and because of that, the number of moves didn’t matter. In sudden death, it matters. In G/60, for instance, there is an enormous difference between a 60 move game (1 minute per move) and a 20 move game (3 minutes per move) – except that you don’t know at the beginning of the game which of those it’s going to be, so it’s impossible to budget your time properly. The intent of delay or increment is that if you can reach an easily won (or easily drawn endgame), you are guaranteed enough time to win (or draw) that game. This is some compensation at least for the uncertainty mentioned above. I absolutely would not play a rated G/60 or shorter time control without delay or increment, for the very reasons GeneM gives.

First, you still aren’t explaining how number of move pairs in a game validate a rating system.

Second, I believe you’d like to invalidate my argument rather than call it unfair. I’ve claimed that Gene’s premise is invalid. So…you have to show his position to be valid, ie true, to prove me wrong. So show me why move pairs matter to a rating system.

It isn’t so much that move pairs matter to a rating system, as that a game where you have a fixed amount of time per move (and always know what that amount is) and a game where you have a fixed amount of time for the entire game are essentially two different games, and including both under a single rating system invalidates (or at least cheapens) that rating system. The longer the sudden death time control is, the less of a problem this is, and delays and increments ameliorate things to some degree, but the basic problem remains. My ability to win “G/30 d0” games by simply playing a little faster than my opponent has very little to do with my ability to win a classical game with a traditional time control, and it makes no sense to have both rated under the same system.

Shifting to the time management discussion.

I agree that a d/0 or inc/0 game has a different feel than d/+ or inc/+. Changes the strategy but doesn’t change the fundamental rules of the game (much like the difference in the NFL between regulation and overtime). Time management for a SD d/0 time control is not impossible. You demonstrated move/time averaging and then tried to claim impossibility b/c of an uncertainty. It’s OK if you don’t like uncertainty, yet that dislike doesn’t make a time control (or rating system) any less valid to a rating system. At present, the USCF considers G/30 with no increment the minimum time for a serious chess game as it impacts one’s “Regular” rating. If you feel that’s not right, please introduce a Delegate Motion for 2015 to change the allowable time controls.

I’m not a delegate, and haven’t the slightest idea how to introduce a delegate motion, but I have no illusions about the likelihood of changing the status quo. If it were up to me, sudden death games would not be ratable at all unless they included a delay or increment (this may be the standard anyway), and G/90 would be the minimum TC for a serious rated game. Anything less than that would be “quick chess”. I know full well that this isn’t going to happen. I am swimming against the tide. But I can still complain. And I can refuse to play in games that don’t meet my standard.

[relyea]

@dennisk

It isn’t so much that move pairs matter to a rating system, as that a game where you have a fixed amount of time per move (and always know what that amount is) and a game where you have a fixed amount of time for the entire game are essentially two different games, and including both under a single rating system invalidates (or at least cheapens) that rating system. The longer the sudden death time control is, the less of a problem this is, and delays and increments ameliorate things to some degree, but the basic problem remains. My ability to win “G/30 d0” games by simply playing a little faster than my opponent has very little to do with my ability to win a classical game with a traditional time control, and it makes no sense to have both rated under the same system.

Are there any tournaments these days where you have a fixed amount of time per move? I assume that the 2 to 30 seconds of delay or increment time are not what Mr. Kosterman is referring to. I know that I have no claim on Mr. Nolan’s time, but it would be nice to know how many tournaments this year have not had a sudden death time control. If my assumption is correct, then there is no problem, and the regular rating system is essentially his second type.

Alex Relyea

[klizzaus]

When I was in New Hampshire (5 years ago), there was a 40/80 tournament nearby (didn’t attend)

I remember hearing that the Chicago Industrial League had a 50/2 time control fairly recently, but looking at the website today, the default time control is now 45/90 G/60. (*Chicago Industrial League is not USCF-rated)

Edit: Found it! On Long Island, the Nassau chess club (I must have confused it in my memory as Nashua) championship has a time control this year of 40/80 40/60.

[dennisk]

@relyea

@dennisk

It isn’t so much that move pairs matter to a rating system, as that a game where you have a fixed amount of time per move (and always know what that amount is) and a game where you have a fixed amount of time for the entire game are essentially two different games, and including both under a single rating system invalidates (or at least cheapens) that rating system. The longer the sudden death time control is, the less of a problem this is, and delays and increments ameliorate things to some degree, but the basic problem remains. My ability to win “G/30 d0” games by simply playing a little faster than my opponent has very little to do with my ability to win a classical game with a traditional time control, and it makes no sense to have both rated under the same system.

Are there any tournaments these days where you have a fixed amount of time per move? I assume that the 2 to 30 seconds of delay or increment time are not what Mr. Kosterman is referring to. I know that I have no claim on Mr. Nolan’s time, but it would be nice to know how many tournaments this year have not had a sudden death time control. If my assumption is correct, then there is no problem, and the regular rating system is essentially his second type.

Alex Relyea

Probably not – at least I haven’t seen any. But many tournaments have a “time per move” control for the first 40 moves, and then sudden death after that. This is already a big improvement over a single sudden death TC. You can at least budget your time until move 40, by which time most games will have reached the endgame. You’re still handicapped if that endgame happens to be a complicated or subtle one that requires a lot of thought. But nothing is perfect, and this system is still a lot better than, say, G/60. And some people here seem to be advocating single sudden death TCs without delay, which is an utter travesty in my opinion. In the early days of sudden death, before delay, I saw many “won games” lost because the player with the (easily) won position ran out of time – and that just isn’t right (unless you’re playing blitz, where it’s all part of the game). This isn’t just bitterness – I won some games that way as well as losing some, but I never felt that I really deserved those wins. And I certainly wouldn’t want my rating to depend on that sort of thing.

[nolan]

Of about 4600 regular-only sections rated so far in calendar 2014, here’s a table of ones with time controls that were not ‘G/something’ or ‘SD/something’:

[code] time control count

---

30/60,30/30;d0 2
30/60,30/30;d5 1
30/70,40/60;d0 2
30/70,40/60;d5 2
30/90;d5 1
35/90,30/60;d0 6
40/120;d5 1
40/80,15/30;d0 41
40/80;d0 20
40/90,30/60;d0 10[/code]

[jwiewel]

@klizzaus

When I was in New Hampshire (5 years ago), there was a 40/80 tournament nearby (didn’t attend)

I remember hearing that the Chicago Industrial League had a 50/2 time control fairly recently, but looking at the website today, the default time control is now 45/90 G/60. (*Chicago Industrial League is not USCF-rated)

Edit: Found it! On Long Island, the Nassau chess club (I must have confused it in my memory as Nashua) championship has a time control this year of 40/80 40/60.

CICL had 45/90, 30/60 as the standard time control up through the season ending this past April. One division had agreed to play 45/90, G/60 in all of its games since people had to meet the train scheduled to the suburbs. Play-off Saturday was also 45/90, G/60. This season is the first to have 45/90, G/60 as the standard.

[wintdoan]

And the 30/90;d5 and 40/120;d5 sound suspiciously like they are missing the secondary control.

And speaking of distortions, think about the machinations of the old days (in high-level chess) where players diddled around to make move 56 so they could adjourn and let their seconds work on the endgame. I have a few chess books with anachronistic guidelines for how to manage the run-up to adjournment or how to analyze an adjourned position.

[BobMessenger]

This topic is a spin-off of 6th edition suggested improvements - 14H.

[dennisk]

@wintdoan

And the 30/90;d5 and 40/120;d5 sound suspiciously like they are missing the secondary control.

And speaking of distortions, think about the machinations of the old days (in high-level chess) where players diddled around to make move 56 so they could adjourn and let their seconds work on the endgame. I have a few chess books with anachronistic guidelines for how to manage the run-up to adjournment or how to analyze an adjourned position.

Yes, adjournments have always been problematic, and are even more so now with the widespread availability of strong computers. As a TD, I like final sudden death time controls – adjournments were always a pain anyway. But as a player, I don’t really like single sudden death time controls, and the shorter they are, the less I like them. I think something like 40/120, SD/1, d5 is a good compromise. It’s long enough for quality play, allows (actually enforces) traditional time budgeting for the first control, and still allows you to schedule rounds without having to worry about adjournments.

[nolan]

And the 30/90;d5 and 40/120;d5 sound suspiciously like they are missing the secondary control.

Possibly so, but I don’t think we can check for every possible error, that’s why we have certified TDs who are held responsible for the accuracy and completeness of their tournament reports.

[tmagchesspgh]

Retaining two sessions of play while using delay or increment has never made much sense to me. It was an odd compromise, a hangover from another era where two+ playing sessions was the norm. I prefer one time session tournaments like Game 90, d5 or Game 120, d5. Both players can use their time as they wish. Playing two sessions where the time control leads to under 2 minutes per move is not exactly conducive to quality chess playing. A first session with 40/80 or 45/90 looks like junk chess IMO. This type of session is usually coupled with an even faster second session, giving the players little chance to go to the bathroom, refresh themselves, or calm down a little after the first session before going back into battle.

[Smythe_Dakota]

I agree that, with the onset of increment and delay, 2-session controls have become an anachronism.

If it is desired to reduce the speed-up-the-later-stages effect, a better solution would be shorter main controls coupled with longer increments. For example, G/90 inc/30 could become G/60 inc/60, or even G/30 inc/90.

Of course, this would work only with leisurely (2 rounds per day) events, but that’s already pretty much the case for 2-session controls too.

Bill Smythe

[eastside]

What I’ve heard from a fair number of players is that they like having the second session. Gives time for a short break after making the primary control. It’s not something they have to “think” about - make control, go to the bathroom, etc.

I have no opinion either way as a player. As a director, a single time control would be easier.

[dennisk]

@tmagchesspgh

Retaining two sessions of play while using delay or increment has never made much sense to me. It was an odd compromise, a hangover from another era where two+ playing sessions was the norm. I prefer one time session tournaments like Game 90, d5 or Game 120, d5. Both players can use their time as they wish. Playing two sessions where the time control leads to under 2 minutes per move is not exactly conducive to quality chess playing. A first session with 40/80 or 45/90 looks like junk chess IMO. This type of session is usually coupled with an even faster second session, giving the players little chance to go to the bathroom, refresh themselves, or calm down a little after the first session before going back into battle.

With this kind if thinking, the equivalent of the time control that I like (40/2, SD/1, d5) would be G/180, d5. This would finish the game in the same amount of time, and yes, I can still budget my time the same way if I choose (make it my goal to reach move 40 by the 2-hour mark), but the problem is that I can’t force the opponent to do the same. When faced with a procrastinator who can’t seem to decide on a move until he’s almost out of time, and then plays at blitz speed for his remaining time, this would force me to sit there for three hours (instead of two) just to get out of the opening. I’m not exaggerating – I have played people like this. I like a time control that not only allows, but forces, both players to budget their time in a sensible manner. It’s a different kind of game, and it would be a sad thing if that kind of game ceased to exist. There really should be a happy medium between blitz and postal (and yes, to some of us G/60 feels like blitz), and it’s getting harder and harder to find such a thing.