Suggestions for allowable time controls

Running Chess Tournaments
1

(The above was intended by its author as an explanation of the current rules, not a suggestion for improvement.)

In reply, I pointed out that under the current rules, G/20 (for example) could be either regular-rated by having a delay or increment of 16 seconds or more, or quick-rated by having a delay or increment of 15 seconds or less, but that it could never be dual-rated. This is just one of many oddities in the current rules for allowable time controls.

One problem with the current rules is that the definitions of regular and quick depend too much on the delay or increment, and too little on the main time control.

Another problem is that the current rules muddy the distinction between regular vs quick rules and regular vs quick ratings. There are really only two kinds of tournaments, regular and quick. Some regular-rated tournaments are also quick-rated, some are not, but both use regular-rated rules (e.g. scorekeeping is required).

Still a third problem is that too many ridiculous combinations are allowed, such as G/10 with a 30-second increment. Such a control is contrary to the spirit of both regular and quick.

Here are my suggestions:

R. An event is regular-rated if either:

Ra. The main (or only) control is at least 30 minutes, or

Rb. The main (or only) control is at least 25 minutes, and there is a delay or increment of at least 5 seconds.

Q. An event is quick-rated if both:

Qa. The total main time (all controls added) is between 5 and 29 minutes inclusive, and

Qb. The delay or increment, if any, is 5 seconds or less.

Defaults. The default delay / increment for a regular-rated event is a 5-second delay. Anything else (including no delay, or any increment) must be announced in all pre-tournament publicity.

The default delay / increment for a quick-rated event is a delay of 3 seconds for G/10 or slower, 2 seconds for G/9 or faster. Anything else (including no delay, or any increment) must be announced in all pre-tournament publicity.

If a tournament is regular-rated by virtue of Rb above, then the delay or increment (5 seconds minimum) must be explicitly announced in all pre-tournament publicity.

Dual ratings. A regular-rated event is also rated under the quick system if the total main time (all controls added) is 60 minutes or less.

Now, folks, wouldn’t the above be a whole lot better than the confusing mess we have now?

Bill Smythe

2

  1. There are time controls that would not be rated under either system (for example, G/20 with a delay of ten seconds).

  2. What happens if the time control is G/25 with a delay of five seconds, and the two players are using an analog clock? Does this game qualify as regular rated?

With respect to the second point, this discussion came up during this year’s Delegates Meeting. In an attempt to make the elimination of rule 5Fa (subtracting time from digital clocks with delay) more palatable to vocal opponents, the rules committee proposed that G/25 with a delay of five seconds would be regular rated provided that all the games were played using digital clocks with delay. I’m not exactly sure how it happened, but somehow, at the Delegates Meeting, this turned into a claim that this was an attempt to make G/25 regular rated. Let’s just say this was not warmly received.

3
  1. As I understand Bill’s proposal, G/20, delay-10 would be Quick-ratable as long as the 10-second delay (or increment) is announced, posted and clearly noted in the TLA and all publicity. That sounds about right to me.

More common for that kind of control would be a club event, probably with no TLA, where the players and TD decide to try something new, or get tired of guys who complain that the kind of chess we often see at club Quick events (with three-second delay) should not be rated. Or maybe a training match between two friends who want to try new opening ideas. In either case I think G/20, delay-10 is fit to be Quick-rated.

  1. We will hear much more about this, I reckon. Some might even argue that G/20 with 10-second delay should be treated the same as G/25 with 5-second delay. Not that we want to stir the pot here.

Bill’s idea would eliminate absurd cases such as G/5, delay-16 being Regular-only rated, while an event with an announced control of G/29, Inc-15 is rated as Quick-only. Not that I have seen or heard of those times controls being used in real life, but still…

4

True – and why should such absurd combinations be allowed? Make it either G/30 d/5 (which is roughly equivalent) and regular-rate it, or G/29 d/3 (also roughly equivalent) and quick-rate it. A 10-second delay makes for a really s-l-o-w “quick” tournament.

Sure, why not? As long as the announced control is G/25 d/5, games played with analog clocks could still be played at G/25. It is either player’s right to furnish a delay-capable clock and set it for the delay.

Bill Smythe

5

That’s not what I had in mind. For a quick-rated event, the maximum allowable delay or increment would be 5 seconds, with the default being a 3-second delay.

Not if I can help it. For a regular event the main time, in minutes, plus the delay time, in seconds, should add up to at least 30, and the main time by itself must be at least 25.

Bill Smythe

6

Should the USCF refuse to rate if the controls fall outside of these parameters, either intentionally or due to ignorance/confusion?

7

There should probably be SOME rating system that any game with a time control of, say, five minutes or more per player fits into.

Deciding which time controls go in which rating systems is mostly a political issue .

8

Well, if you insist on absurdities like allowing G/5 with a 90-second increment to be rated, at least rate it in the system that corresponds to the main control, not the delay or increment.

In other words, anything faster than G/25 should, at the very least, be quick-ratable only, regardless of the delay or increment.

Bill Smythe

9

I could make a good argument that G/5+90 increment is a LOT slower than Game/30. (That doesn’t mean I’d PLAY in it, though.)

If I was designing a tournament at G/30, I could schedule rounds about 75 minutes apart. If I was designing one at G/5 + 90 increment, I’d probably schedule the rounds at least two hours apart, maybe three hours apart.

Although I don’t really like the mathematical approach because I think too many people will get it wrong (not all of them accidentally), I think something like this might be the right approach:

Total time = initial clock setting (in minutes) + secondary time controls (in minutes) + increment (in seconds) + delay (in seconds) /2

If 5 <= total time <= 29.99 then the event is quick rated only.

If 30 <= total time <= 60 then the event is dual rated.

If total time > 60 then the event is regular rated only.

Note that these break points would mean that G/25+5 delay would be quick rated while G/25 + 5 increment would be dual rated.

Other weighting factors and different break points are possible, of course.

10

From earlier threads (I mean years ago), as I understand, the entire “duel rated” concept came about because not enough players were playing in quick rated events.

So in order to get a better mathmatical model of quick rated play, duel rating was invented in which games that met certain time criteria, would be rated in both quick and regular ratings.

Personally, I don’t care to put any of my regular rating points on the line when I’m playing faster time controls. I suppose the scholastic crowd is more open to duel rated events.

11

No question, G/5 inc/90 is hugely slower than G/30.

The distinction between regular and quick must be defined somehow. The point of doing so based exclusively (or mostly) on the main time, and not at all (or barely) on the delay or increment, is to relegate all these absurd cases to quick-only status. That keeps the regular rating system more nearly “pure”.

Bill Smythe

12

You better plan to show up in Florida as a Delegate, Bill, because I don’t see your suggestion as one that will even be offered to the Delegates. I’m not convinced it would pass, anyway, in fact I’m not ENTIRELY sure that the Delegates will allow G/30+5 (G/25 clock setting) events to continue to be dual rated once 5Fa is phased out on 1/1/2012.

13

I have always wondered why 5 sec became the selected delay for regular rated games. In D. Bronstein’s original formulation, the delay he suggested was 15 sec, a time that would not be incremented because he did not believe you should get time for nothing more than shuffling moves quickly. He thought three games a day of Game 20 to Game 30 with 15 sec delay would provide a reasonable opportunity to hold a ten player round robin tournament over 3 days to promote the game of chess. In Informant 44 there were a set of rules for an Active Chess system and titles.

With 15 sec. delay you still have time to think and physically make a move. In 15 sec. increment, there is little time to think, write your move, move a piece, and hit the clock. I do not know why anyone would hold that type of increment event. Even 30 sec. increment can be a stressful rush, especially if you write slowly.

Any time control that is under Game 30 should be under the quick chess rating system. Doesn’t matter if it has delay or increment. The core time is still under 30 minutes. If you want to make a variation of Game 25 with a delay as a regular time control, then that delay should exeed 5 seconds. 15 sec delay would be fine Watching clock bashing monkeys is fun to watch sometimes, but not if you have to sit next to it.

14

I disagree with your statement that 15 seconds of delay is more than 15 seconds of increment. If you move in 15 seconds or more, the effect on your remaining time is identical. If you move in less than 15 seconds, you get to keep the extra time in increment mode, but not in delay mode.

I think 5 seconds delay came about because there was general consensus that if a player was in an obviously drawn position (but on where a checkmate was POSSIBLE, such as K+N+N vs K), five seconds would be enough time for a player to not make the obvious blunder. Whether most players could make most of the elementary mates with no more than five seconds of time per move is a separate issue. (I MIGHT still be able to do K+B+B vs K I doubt I’d be able to do K+N+B vs K, though I’ve known a number of players who could do it easily.)

15

I think Tom meant that players would still need to keep score in a game with 15-second increment, but not with 15-second delay. I am pretty sure the rules allow players to stop keeping score once either side drops under five minutes, unless the increment is 30 seconds or more. (I assume that also applies to 30 seconds or more of delay, but I’ve never seen that.)

Score-keeping requirements and how that is affected by time controls and increment/delay could be clarified in the revised Allowable Time Controls document. I suspect that most increment games will use 30-seconds of increment per move—thus requiring players to keep score all game—but I have been wrong before.

16

There’s a possible situation (I forget who brought it up) where an increment might end a game more quickly than a delay (of the same number of seconds).

If the position is clearly drawn, but one player is down to 1 second of main time, plus a (non-cumulative) delay, his opponent might be inclined to play on and hope for a blunder. The player is, after all, in permanent time trouble.

Now change the delay to a (cumulative) increment. Now the player with 1 second keeps getting time added. After several moves he can increase his main time to a full minute or more. Knowing this, the opponent might be more inclined to concede the draw.

Bill Smythe