Time Control Wording in TLAs

Running Chess Tournaments
1

Recently I received a flyer for a tournament I had committed to play in some time ago. I was actually involved in helping the organizer prepare the TLA and didn’t realize what he was planning. It was only when I saw the flyer, where the time control was explained in words, that I realized that his plan was to have time delay only in the 2nd control. This is legal of course. Rule 5Fb is not an absolute, mention is made there of having delay only in the final control as an “announced variation.”

The TLA shows, “40/100,SD/60d10.” Apparently, the standard method with delay from move one would be shown by, “40/100,SD/60,d10.” Is this really USCF policy, to make this distinction with the inclusion or omission of a single comma? So by leaving out this comma, the organizer has met his burden of announcing this variation in all pre-tournament publicity?

I found a lot of different things in Chess Life TLAs: some with the comma, some without and some (still) with no indication of delay at all. So what exactly is going on?

– Hal Terrie

2

The rating system programming currently assumes increment or delay is in effect from move 1 on.

The rules already require that increment time controls be in effect from move 1 on:

(It is possible that the above rule is not entirely consistent with some FIDE time controls. In an earlier thread a year or two back someone pointed to a FIDE world championship event in which increments weren’t being added until after the first time control. Whether that time control is still permitted by FIDE is a separate issue.)

So, the only unresolved question is what to do about delay time controls.

The parser in the ratings validation code will rewrite your time control as: 40/90,SD/30;d5 (Note the semicolon before ‘d5’ rather than a comma, the distinction was intentional.)

In any event, it needs to be VERY CLEAR to the players that the delay doesn’t take effect until a secondary time control.

The least unambiguous shorthand way to specify that the delay does not take effect until the second time control would probably be something like this:

40/90,d0;SD/30,d5 (Again, note the use of a semicolon to separate the time controls into two separate sets.)

The parser does not currently recognize this encoding, though, mostly because there has been no need for it…

There are also classification issues that would come up with delays that don’t take effect on move 1.

Is 30/30,d0;SD/30,d6 a dual-rated time control or a regular-only time control? I don’t think the rules as passed last year address that issue. That means this is an issue for the Rules Committee.

3

Have you seen the time controls from recent top GM events? Anand–Gelfand used increment only in the last time control, starting on move 61. So did the London Grand Prix; I think all GP events use the same control. (40/120, 20/60, SD/15, Inc-30 from move 61.) Bilbao is 40/90, SD/60 with a 10-second increment starting at move 41, I think.

If anything, increment only for the ultimate control is the standard rather than the exception for top-level tournaments—with all-GM fields, so norms not an issue. To confuse things, the Olympiad and other official FIDE events use increment from move 1—40/90, SD/30, Inc-30 (1)—and that is listed as something like “the one and only time control” on the FIDE site.

Under the ‘old’ norm-approved controls, there had to be either no increment or else increment from move one. (Three of each were approved, I believe.) Not sure how the revised norm-control rules just passed will affect that.

Mig wrote about this. He says the top players prefer increment only in the last control; they were not happy when Corus switched to increment from move one—40/100, 20/50, SD/15 Inc-30 (1)—to allow for norms in the C group.

That’s well and good for FIDE events, with small fields of all or mostly GMs…but not so much for a standard USCF weekend Swiss. Too much room for confusion. Some players still pretend they can’t set their digital clocks with delay in effect from move one. Add this wrinkle to the mix and enjoy the spectacle—though I see the point Mig says the top players make.

I think I am the poster Mike refers to above; I noted that we had a player in our club who preferred delay only in the last control. I agreed to that when I played him in rated games, but of course if the opponent objected then the game was played the ‘standard’ way, with delay from the start.

Also note that delay/increment only in the last control requires either re-setting the clock or the use of a clock-press counter: neither ideal in a large tournament.

Will be interesting to see how the tournament Hal refers to works out…

4

Eric,
I know this isn’t true for several types of clocks (DGT for sure). If the number of moves for a time control is not set, then the clock just keeps track of the time not moves. Once the time runs out, it moves to the next time control. So, if you want increment only in the last time control, the increment would not start to be added until you’ve used up all the time before that control. In some ways that makes sense and in others it’s confusing.

The only reason I can think of that someone would not want to have increment in the earlier time controls is that it is easier to determine your time/move. Without increment you just divide the time showing on the clock by the number of moves left in the control. With increment you have to do an extra step of adding in 30" to the result. Maybe it is easier but the downside is that you can still have smash clock chess if you get low on time approaching a time control.
Mike Regan

5

The rules committee chair and both the USCF staff and EB liaisons to that committee have been emailed to alert them to this thread.

6

Right—but the idea as I understand it is for the increment to kick in for both players on the same move—be it 41, 61 or whatever. Does not seem fair to have the increment start for one player at move 61 and much later—if at all—for the other player. (Which is obviously what happens with no move counter or re-setting of the clock.)

Increment/delay only in the last control is a good example of a policy that is perfectly OK for small events with enough TDs/arbiters on hand—probably very strong players, too—but a potential disaster for a standard USCF weekend Swiss.

7

Thanks, Mike—but there are members of the Rules Committee who know more about FIDE time controls than I do. This is not news.

The issue of streamlining time control formats for TLAs might be worth some focus, though…

8

Aw, for chrissake. I knew this would happen sooner or later – the people in charge of publications at USCF being on insufficient guard, combined with the lack of standards (or the use of obscure standards) for how time controls should be stated in TLAs.

And there shouldn’t be any difference between increment and delay, regarding what’s legal, what’s not, what must be stated, etc. Having one standard for increment and another for delay is just plain preposterous.

The rule ought to be extremely simple: any increment or delay must be in effect from move one, unless there is an explicit and clear explanation to the contrary within the TLA.

Commas and semicolons don’t cut it. The only legal way to specify different increment or delay in different controls should be to specify the increment or delay separately for each control. For example: 40/120 d/0, SD/60 d/5. Here there is only one comma and no semicolons, and even the comma isn’t strictly necessary. If the increment or delay is to be in effect from move 1 (the default), it can be expressed simply as 40/120 SD/60 d/5.

These standards should have been applied strictly ever since January 1, 2012, when the new rules went into effect – along with a uniform format (as in the above paragraph) for describing the control. Any TLA submitted in another format should be converted at the USCF office to the standard format.

Another standard that should have been applied strictly is the requirement for the increment or delay to be announced explicitly, even if zero. Events listed as simply 40/120 SD/60 should not have been accepted at all. Either they should be returned for correction, or changed automatically to 40/120 SD/60 d/5. (Yes, d/5, not d/0, should be assumed whenever neither increment nor delay is mentioned – that is, after all, the default.)

Furthermore (yes, we’re probably all tired of furthermores at this point), the use of different increments or delays in different controls (or only in the last control) should be strongly discouraged in Chess Life and elsewhere. There are a whole bunch of reasons for this. For one, some players may believe the delay is on when it is not. The result could be unnecessary time forfeits, acrimonious disputes, and, quite simply, a bad scene all the way around. Second, not having delay at the primary control is likely to result in poor player behavior, such as knocking over pieces without replacing them, using two hands, etc. Third, having the same delay throughout the game creates a more nearly uniform “look and feel” from one control to the next. Fourth, some clocks may not be capable of having different delays in different controls – or, if they are, they handle it differently. Does the second control begin at move 41, or only when the first control expires (could be at move 47, move 63, etc)? And, in the latter case, does it begin for each player when that player’s first control expires, or for both players when one player’s first control expires? This whole thing is just a huge can of worms.

Amen.

Bill Smythe

9

Mike, thanks for informing me of this thread.
I try to stay away from the Forum as I have a tendency, well more than a tendency, to say much more than I need to.

At the start of this thread, the way the TC is listed is very ambiguous and even deceiving. I would definitely expect the delay to be active from move 1, and demand it if I showed up at the event. Any variation of the rules must be stated very clearly, and Mike’s method of 40/90,d/0;SD/30,d/5 would do that. Even better would be to state that “Delay effective in second time control only”. If different delays are to be used BOTH delays MUST be mentioned, even if one of those delays = 0. Spell it out, in addition to the shorthand.
It may be more expensive when printed in Chess Life that way, but anyone stupid enough to want such a situation needs to pay the price for that stupidity.

As far as ratability, the program now is written so that the game is consistent from move 1. Frankly, I don’t know what the TD is going to put on the rating report. Will he put d=0 or d=5?
The system would automatically rate it as reported.
I suppose to keep in the letter of the law, I would assume a 60 move game, and compute total playing time based on that.

In this case it makes no difference, as the game is more than 65 minutes, but in many cases it could make a difference.
For example, say it was 40/45,d/0;SD/15,d/10 looks on the surface to be regular only (60+10=70)
but since only the last TC has delay the right way would be to factor the portion of the game that contains delay, and prorate the time accordingly.
mm1+mm2+[(60-N1)/60]*d2.
In the above example the time is 45+15+[(60-40)/60]*10 = 60 + 3 1/3 = 63min 20 sec, and is dual rated.
or to generalize even further:
mm1+[N1/60]*d1+mm2+[(60-N1)/60]*d2
if the delay for the first time control is different from the delay in the second
which, of course, with d1=0, gives the same result.

I can only see headaches for the programmer, and for organizers that follow the rules as written in trying to input their events.

Mike, I DON’T want you to complicate the program to compensate for idiots, especially if it creates additional work for anyone following the rules (such as needing to input more than one delay time). In addition, such events are rare, and the overall impact if an event if not rated precisely as dictated, will be minimal. Don’t worry about it. I do not intend to put it into the Rules.

David Kuhns
Chair, Rules Committee

10

Bill Hall has approved the following, which should clarify the issue of what rating system an event uses when an increment or delay time control doesn’t start with move 1.

11

Given the comments in this thread, perhaps its time to simplify the rule to state that delay must be the same for all time controls.

12

That’s a matter for the Rules Committee and the Delegates.

It could also result in some FIDE ratable time controls, such as the ones where increment doesn’t take effect until the final sudden death time control, that are not USCF ratable.

Perhaps the following would suffice:

13

Maybe, but allowing variation often tends to get it. In my statement above, I was specifically saying delay (not increment) and I thought that was withing FIDE rules.

I suppose another possibility would be this: that we work to simplify our rules - for example saying that delay and increment must start from move 1 or other rules, but having one rule that states that any event that follows FIDE rules will be USCF ratable so long as it is stated in advance that FIDE rules will be followed. Then we don’t need to delineate all the differences in the rulebook.

14

I think a blanket rule that says any event that is FIDE ratable is USCF ratable could create as many problems as it solves if we aren’t careful.

For example, FIDE Blitz events can use time controls that will not be USCF Blitz ratable. How do we ensure that someone running or playing in such a FIDE Blitz event doesn’t think it will be USCF Blitz ratable.

And what do we do when FIDE changes its time control rules again? I don’t see the Delegates being all that keen to give FIDE that much control over USCF rated events.

We have around 60,000 players playing in USCF rated events, perhaps around 5% of them are playing in FIDE rated sections of USCF events.

15

I disagree with publishing a statement such as this, but may be talked into it.

ANY variation in the rules must be stated “CLEARLY and UNAMBIGUOUSLY”. I do not want to put a statement like this in any specific rule variation, or in every possible rule variation. I should be obvious.
Making such a statement only darws attention to the possiblity of the variation, and many people will then start using it. (e.g. the “paper scoresheet” variation and the old “subtract time from a delay clock” (old 5Fa) variation.)
I want to keep the rule as it is, and state that delay or increment is to be used from move 1 (as intended) and not to appear to give people a choice.

I agree with th “office interpretation” of assuming the delay is in effect from move 1 for rating purposes, but don’t put it in the Rules.

David

16

Let’s not complicate the rules with fussy unnecessary distinctions between delay and increment. Whatever rule is adopted, either permitting or prohibiting multiple delays or increments, should apply to both delay and increment.

That way lies madness. It’s too much like giving out your bank account number when you get an email from an African prince claiming he has a lot of money and is looking for a place to hide it.

Amen. I’m glad we have at least one sensible person posting here.

Amen, again. I’m glad we have at least two sensible people posting here.

A statement like “CLEARLY and UNAMBIGUOUSLY” is itself unclear and ambiguous, especially when capitalized.

Yes. This is one of those cases where, in practice, an office interpretation works better than a rule.

IMHO, however, we should not go so far as to absolutely prohibit multiple delays or increments, especially if FIDE is using them. USCF’s stance ought to be explicit, along the following lines:

The use of multiple delays or increments is strongly discouraged. If the delay or increment is to be different in different time controls (or if, for example, a delay or increment is to be in effect only in the final control), the delay or increment, or lack thereof, must be explicitly stated for each control. For example: 40/120 d/0, SD/60 d/5. (And yes, I’m sure this wording could be improved or simplified.)

By the way, I am a charter member of the Punctuation Simplification Movement. There is no need for both commas and semicolons. Spaces can be used in place of commas, and commas in place of semicolons:

Single control: G/120 d/5
Two controls, same delay or increment: 40/120 SD/60, d/5.
Two controls, different delays or increments: 40/120 d/0, SD/60 d/5.

On another topic:

Hmm. Can you give me an example (now that G/3 inc/2 is, or soon will be, USCF blitz ratable)?

Bill Smythe

17

It is my understanding that FIDE defines blitz as having time controls totalling less than 15 minutes The USCF has a different cutoff.

I believe there are also different cutoffs between FIDE Rapids ratings and standard FIDE ratings than what the USCF uses between its Quick and Regular ratings. FIDE complicates this even further in its standard rating system by having minimum time per player rules based on the strength of the highest rated player in the event.

And let’s not get into dual rating, which is a mathematical kludge. But are there people who really care what their FIDE Rapids rating is? And are there people who really care what their USCF Quick rating is?

FWIW, FIDE’s use of a time control in which increment doesn’t start at move 1 seems to be in conflict with FIDE’s own rules, specifically C08:

18

People with a quick rating higher than their regular rating do seem to care.

19

OK, well, that’s just a difference in which rating system is to be used. Apparently, though, any FIDE-ratable control is (or soon will be) also USCF-ratable under some system.

Aha! I thought so. Those who oppose fixing the quick system keep saying nobody cares about their quick rating. But maybe that’s because it hasn’t been fixed, and most players’ quick ratings are way below their regular ratings.

Bill Smythe

20

I’ve been waiting for an expert on FIDE rules and regs to weigh in here.

To be precise, increment only in the ultimate control is used often in FIDE-rated events that feature top GMs—but which are privately organized, thus not official “FIDE events.”

The control listed at the FIDE site is used at the Olympiad, World Team and other events that are organized by FIDE as “official.” As I recall, that was a compromise some years ago after GMs objected to FIDE’s plan to adopt G/90, Inc-30 as its default control, as used in early finals matches from the knockout lottery, er, world championship.

So, while FIDE allows increment only in the last control, at least for non-norm events, it does not require it at any event. Thus, it is not an issue that “FIDE time controls” might not be USCF-ratable. (Maybe a question as to whether to USCF-rate Nakamura in the Grand Prix.)

However…why would USCF disallow rating of games with delay/increment only in the last control, when it still allows rating of games with no delay or increment at all? In either case, the non-standard policy needs to be announced in advance: as it should be. I support the use of delay from move one in USCF-rated events, mainly to avoid confusion—but I do not like the idea of an outright ban on properly announced and advertised variants.