I do not “like” delay in the final control but not in the primary, but my preference for delay throughout the game is based on the practical reasons you outline, plus uniformity: It’s hard enough to get some players to ‘get’ it, even after all this time.
From the theoretical POV, though, I see why Anand and Topalov agreed to use increment only for the final control in their match. (There was a dust-up in FIDE over this a few years ago. Tata/Corus had to change from 40/120, 20/60, SD/15, Inc-30 (61) to the FIDE norm-approved 40/100, 20/50, SD/15, Inc-30 (1) so players in the C section could chase norms.)
Trust me, I have been in the boat of having a few seconds left for a few moves to make a move-based primary control. Never again, vows this patzer—at least till the next time…but the point is that in a move-based control you know going in what is the deal. If you run short of time on move 35 of a control that you know lasts 40 moves, so be it. Cruel but fair.
From the practical POV of not disturbing other games, (as much, anyway; five seconds is not ‘that’ much extra time), reducing the number of pieces flying across the room and so on, I agree that delay in the primary part of a mixed control is better.
The player I mentioned likes time delay in a SD control, to avoid blitz trash insanity, but cannot wrap his mind around the idea of delay in a non-SD control. When I point out the practical merits of delay from move one, he shrugs.
Getting off the thread here, but as long as GM events are played with increment only in the final control, this will be an issue.
Or, for that matter, quick-only? The first time control is 25 minutes, with a delay of 0. This adds up to less than 30.
A couple of possible rules suggest themselves right away:
A regular-rated (or dual-rated) event must have a first (or only) time control of at least 30 (main time in minutes plus delay/increment in seconds).
A quick-only event must have a single, sudden-death control. No secondary controls.
That way, 30/25d0, SD/30d15 would not be rateable at all – and that’d be just fine with me.
Again, not rateable at all – the above two rules would take care of that.
The first control (main plus delay) is 30 (or 35 if the delay is to apply from move 1), so it would be regular-rateable.
But this control brings up a different issue. A secondary control of 1 minute looks suspiciously like increment, rather than a new control. There should be a lower limit (15 minutes?) on the main time for secondary controls, and an upper limit (5 minutes?) on delay/increment for any control. The first of these limits would have to be longer than the second (in this example 15 minutes is longer than 5).
This would help define the border between increment and secondary controls, and would help resolve various ridiculous and time-consuming issues, such as under what circumstances a player is allowed to “unflag”.
I think I’d like to throw in the following as well:
The delay/increment in a quick-only event cannot exceed 5 seconds (or maybe 10).
There must be a willingness to outlaw really absurd time control combinations. The attempt to include everything makes for inherent inconsistencies.
Both the 2011 and 2012 rules refer to ‘total time for each player’ for sudden death (final) time control events. The total time is greater than 29 minutes per player, since 25+30 = 55, so it would not be quick rated. (So I can clearly not trust the wine in front of you.)
However, if one adds the delay from the start of the game (zero), it is NOT greater than 65, though if one adds the delay in the final sudden death time control (15) it IS greater than 65, so whether this is dual rated or regular-only is unclear.
And even if it is changed to 30/25d5, SD/30d15, then 30+25+5 (delay in first time control) is STILL less than 66. (So I can clearly not trust the wine in front of me.)
Whether or not the Rules Committee considered events where delay/increment is not in effect from move 1 or is not the same in each time control period is unknown, perhaps someone on Rules can comment on that?
So, what rating system would be appropriate for an event which is G/5d20?
And is G/5+20 significantly different than G/5d20?
Probably true—and perhaps more true with TLAs than online tournament report forms. For the latter, a coupla things might help TDs get accustomed to the new form and new time control regs.
First, place a news brief announcing the revised form (to coincide with the new regs) on the home page of the USCF site. Same idea as the recent item on changes to the IO title. Not all conscientious TDs read the Forums.
Also, on the revised form itself, I see no need to disallow abbreviations to denote delay/increment that some TDs have used for two years now, such as G/60 + 15 INC. The new form will not accept “INC” in any way that I tried. The only way to indicate increment is “+15.” (Unless I missed something.)
That will frustrate TDs who took the time to learn how to note the added delay/increment info to the time control field…and now suddenly the ‘new’ way they adopted is not new enough any more.
The system administrator formerly known as Mike Nolan told me it took months to reach consensus on appropriate abbreviations for delay/increment on the new form. Not quite sure who constituted the consensus pool or what they debated…but why change what works?
Of course, a note about all this in Chess Life is a given.
Maybe all this will finally generate some steam to abolish Dual-rating. Yay.
Did you try ‘inc 15’? (in ‘+ 15 inc’ either the + or the inc is redundant.)
Tim Just and David Kuhns will have a short article on this in the next issue of Chess Life. Also, the ‘update to the rulebook’ document for 2012 has been posted,
you can access it by going to “About USCF” then “Governance” then “Reports”.
(Direct links may change, the above should work even if the file name is changed.)
Jennifer Shahade has been alerted to this, so there may be a CLO article coming.
For the time being, the parser has been changed to allow a comma or semicolon before the increment/delay term at the end, though it will standardize it to a semicolon.
This is one area where the ‘new’ 2012 regs will improve things. G/5+20 delay or increment should be either not ratable or else Quick-only—as it will be starting next year. As of now, it would be Regular-only, which is silly.
It’s hard to draw a clear line as to where the difference between increment and delay starts to matter. I would say at three seconds or faster delay or increment is much the same, anything longer than 10 or at most 15 seconds and in some games the difference will be enough to notice, to keen-eyed clock-watchers.
From three to 10 seconds the difference could well be non-zero, but so small as to hardly matter, at least in a control with substantial main clock time.
I have yet to see an ad for a tournament in which the main clock time is less than the delay/increment. That is a good thing; perhaps it should be the rule?
Yep; that works. The system accepts G/90 Inc-15 but only without the “+” sign. Redundant or not, that works in the current (2011) form and it’s what some TDs have used for some time now.
Perhaps equating “+” with “INC” or “Inc” is intuitive to some, especially those who follow FIDE events and need not worry about barbarous Bronstein and dreadful delay. To my reader’s-eye, I want to know whether it’s delay or increment that is added after the main clock time info.
That’s the first thing I’d change. Most of these considerations should be based on the first control, not the total control.
None. That’s the second thing I would change.
Between 2011 and 2012 we went from one extreme to the other. In 2011 only the main time mattered. In 2012 only the sum of main plus delay/increment matters. We need something in between.
Nope. When designing criteria that distinguish regular from quick, we don’t need to nitpick the difference between delay and increment – even though there is often a huge difference in the “feel”.
What we need is something along the following lines:
Regular:[list][*]Main time 25 minutes or slower.
Main time (minutes) plus delay/increment (seconds) 30 or slower.
Multiple controls allowed. When there are multiple controls, the above criteria apply to the first control.
[/:m][]Quick:* Main time 6 through 29 minutes.
Delay/increment 0 through 5 seconds.
Main time (minutes) plus delay/increment (seconds) 6 through 29.
Multiple controls not allowed.
[/:m][]Blitz:* Main time 3 through 5 minutes.
Delay/increment 0 through 2 seconds.
Main time (minutes) plus delay/increment (seconds) 5 through 7.
Multiple controls not allowed.
[/*:m][/list:u]
A list like the above doesn’t need to distinguish between regular and dual. That can be done separately. A dual event is simply a regular event with some additional restrictions.
By the same token, we do need to list blitz as a separate tournament type. The rules are quite different. It does not matter that blitz and quick are (for the time being, at least) using the same rating system.
Separate from what? A blitz event could be a swiss, round robin, match or other (eg ladder or random pairing) event. It could be an adult, scholastic or JTP event.
It could be a team event, it could be a national event, it could be an online event.
It could not be FIDE rated at this time, but that’s under FIDE’s control, not the USCF’s.
Where would this blitz type be notated?
This is currently more of a TLA issue than a ratings issue, current USCF policy is that blitz events are part of the quick rating system. The Delegates have rejected the idea of a separate blitz rating system more than once.
It is, however, useful to the USCF office to know after an event has been rated whether an event was conducted under blitz rules, so this should probably be part of the rating report, and a separate checkbox has been added for that.
(The 2012 rules writeup requires that blitz be mentioned in the event name or time control, though the Rules Committee has been advised that the latter will not parse properly. Perhaps they intended it to be in the section name, since there could be events with blitz and non blitz sections in them?)
d = DELAY
= INCREMENT
The latest version of the parser should change “+ 5 inc” to “+5” and should also recognize “5 inc” and “inc 5” and change both of them to “+5” as well.
From your earlier post it sounded as though you are considering caving in to him, by running a tournament with delay only in the final control. You would be the TD, he would be a player.
If so, I was wondering how you would reconcile the choice between asking players to turn on the move counter [ugh] or dealing with the combination [ugh] of not having a move counter yet having delay only in the final control. As you are known to dislike the move counter, this situation would seem to put you between a rock and a hard place.
The parser should be able to accept almost anything, but when the field is exited (via ENTER, TAB, or a mouse click), the user input on the screen should immediately be blanked out and replaced by the standardized version.
For example, if the user types any of the following:
– without regard to spacing, punctuation, upper or lower case, “+” vs “inc” vs both, etc – then the parser should immediately re-display it as
40/90, SD/60; +30
– or whatever standardized form is eventually decided upon.
This would avoid user frustration, while still reinforcing the desirability of using a standardized notation.
Incidentally, in the standardized form, I too would prefer “inc” instead of the plus sign. “+” looks as though it could mean either increment or delay. The letter “i” by itself is another possibility (parallel to “d” for delay), but the letter “i” is too small to be comfortable when it stands alone.
Please note that I said I might play a game with this fine fellow with delay only in the SD control. I would never use that as standard policy for an event I directed.
Three TDs rotate at the club in question. Two of them are the fine fellow in question and I. We try to allow a tad bit of leeway for things such as two veteran players who know each other and the rules pretty well to experiment with delay only in the ultimate SD control, as long as both agree in advance. That would not work at a ‘real’ open tournament, of course…or at least it would be asking for heap big trouble.
When we meet at the world championship Manhattan Gambit thematic tournament—you as White; I as Black—where an arbiter monitors and corrects the move counter when needed, I am all for it. Till then, at the level of events we play in and direct, a clock…digital or otherwise…should be a clock.
I should mention that, although I like the move counter, as a player I will always shut it off if the opponent asks. Sometimes I shut it off without asking, if I suspect he wouldn’t like it or if I think he might get confused – or if I know he prefers to see the delay displayed as a digit, which it wouldn’t be with the move counter on.
Sooner or later, I’m sure there will be a separate blitz rating system. (Probably not for a few years, because the quick system needs fixing first.) It would be good to classify events as quick or blitz, starting as soon as possible, so that at start-up there will be plenty of unambiguous data already available.
Blitz rules and quick rules are different enough so that, in principle, blitz ratings should be separate.
It wouldn’t have to be. Under my proposal, the difference could be discerned from the time control alone:
As I recall, you fought hard for a system where the time control alone determines the distinction between regular and quick. And you won – that feature is going into effect with the 2012 rules. My proposal simply extends your excellent idea to a similar distinction between quick and blitz.
The only one of those it wasn’t already recognizing was ‘+ inc 30’, which it now recognizes.
There DOES need to be something other than a space in between the minutes portion of the previous time control and the increment amount, such as a comma or ‘inc’ or ‘+’.
It should also recognize ‘del’ and ‘delay’ as the same as ‘d’, either before or after the number as well as something like ‘d/10’.
Are there other combinations that it isn’t finding yet. Probably.
The 2011 rules (which have been in effect since January of 2009) also do not permit any ambiguities, except for the stuff about subtracting time from the clock (and it’s not clear that created any ambiguities, just confusion.)
The 2011 and 2012 rules both permit numerous weird but ratable time controls, although there may be proposals in front of the Delegates in 2012 to restrict what time controls can be regular rated only. (Those proposals will likely come from Bill Goichberg.)
Hopefully those proposals will continue to just make it clear which rating system(s) a given time control unambiguously mandates.
That is preferable to rules that could create time controls (other than ones faster than some minimum time, currently 5 minutes) which cannot be USCF rated or ones in which certain time controls can be rated in one system or another based upon the whim of the organizer or TD.
Historical note: The intent of the Delegates in 2000 when dual rating was created was NOT to allow TDs to pick and choose which events were dual rated, that decision was made by the office when it announced that events submitted on paper required TWO ratings fees to be dual rated. The 2000 Delegates motion made it clear that events between G/30 and G/60 were ALL to be dual rated.
DM 00-17 read in part:
The 2005 rewrite of the ratings programming eliminated any need to charge twice to dual rate an event and returned the USCF to having the time control unambiguously determine the rating system(s).
You are right. I could legitimately claim that I didn’t have enough time to seriously review a new form when the thread was first posted, but not now. I appreciate that you made the form available for comments. I’ll add my feedback in separate posts to make responding to specific feedback easier. For the record I am using Firefox 3.6
Response: It seems to work OK in versions of Firefox from about 3.4 to 7, in several versions of IE and in Safari. The only problems have been with IE (of course), notably with some of the Javascript.
Regarding TLA Reference: I have a standing online TLA for my weekly club tournaments. Since online TLAs have start and end dates it would be nice to make these available to be tied to any tournament that falls into the appropriate time period.
Response: The challenge will probably be finding the right TLA if there are multiple recurring TLAs out there for that affiliate. Currently it checks the starting and ending dates, and only displays those TLAs that match both dates.
Note that it only shows PRINT TLAs, not free online TLAs, as only print TLAs meet the requirements for having a TLA in the Grand Prix and Junior Grand Prix rules, and that’s the reason for wanting to know whether or not the event had a TLA.
Both the online and print TLA systems need upgrading, the online TLA system should probably be scrapped and rewritten, and a recommendation to that effect was made to Bill Hall last week.
I changed the date of my tournament from 2012 to 2011 to test whether an existing online TLA would become available to me, and I noticed that “2012 Time Control Rules Apply” remained in force, after both updating and saving. I then noticed that the section dates did not automatically update (though they had defaulted correctly when I first entered the event date). A warning message was there as well. I suggest that for single-day events the section dates should automatically update if the event date is updated. For multiple-day events I would keep the current functionality, as the TD may need to adjust the start date of the event and not want to lose the detail that they already entered for five sections that started on different days.
Response: There is no change in functionality from the current data entry form in this area, although one change has been made–the tournament ending date can now be changed before the event is submitted for rating.
But once an event is rated, it will still be very difficult to change the event ending date, because there are too many external references to the event ID in the database. The dates of the section currently being displayed could be adjusted to be consistent with the tournament ending dates, but not the other sections, at least not until those sections are displayed and edited.
The refresh-without-saving option is nice. The automatic refresh after entering certain fields (rounds, players, etc.) is annoying – I prefer to refresh at my convenience, as data entry is faster. If you choose to keep the automatic refresh then please add a subtle visual indicator (perhaps make the label text blue) so that we know which fields will do this. Also, if possible on the automatic refreshes, refresh only the relevant section and do not force a complete page refresh.
Response: The reason some fields auto-refresh is because those are the fields that when modified cause differences in which fields are shown, which pretty much mandates redrawing the entire form. We could disable either the auto-refresh on ‘number of players’ or ‘number of rounds’ but probably not both. (It probably makes more sense to disable it for ‘rounds’ than ‘players’ because ‘rounds’ is entered first.)
Response time on either saves or refreshes is a concern, especially for events over 100 players. The form complains if a save or refresh is attempted before the previous one has finished rendering, and it also disables the section change pulldown until the form has finished rendering. Since this is done using Javascript, there could be browser compatibility issues as a result.
