First, please note the phrase “As a director”. It’s important.
Second, you don’t need multiple time controls to play high quality chess, as FIDE has conclusively demonstrated with its adoption of increment. If you want to make the case for longer increment, you can do that. Hal Bogner has done so in various places on the Forums, and such a time control could conceivably work. But I have a hard time seeing how (for example) 40/120+30i, G/ is superior to G/150+30i, or even G/120+30i.
I must agree with Mr. Bogner. Having played G/150+30i at the 2007 US Championship, I can say it is a brutal time control. I had several games over 6.5 hours (90+ moves) and the fatigue of playing the so long, especially with 30 seconds per move and no break at the end, contributed to poor quality of play. Most of the mistakes were mine, except for one incredible queen endgame swindle against a teenager who is now one of our top Grandmasters.
G/175+5d (without an increment) proved equally poor at the 2006 Denker, where a majority of the top boards routinely went 5+ hours. My roommate managed to play a 20 move game that went over the 5 hour mark. Wisely, the organizers quickly gave up on that experiment.
First, I apologize for leaving out the last part of the multiple-period control in the post you replied to. I meant to say “40/2+30i, SD/1+30i”. I imagine such a time control has pretty similar risks to G/150+30i.
I can understand the argument you and Mr. Bogner make about having a break during the game (though my belief is that players who habitually manage their time poorly would do so regardless of the time control). I would say that is a better reason for longer increment in a single time control than multiple time controls.
That said, I agree with your point about the length of the sessions in that particular time control. My experience as a director has been that G/120+30i seems to work for strong players. Those rounds typically last no more than 5 hours, there’s plenty of opportunity for play and reflection, players still have to manage their time, and (most importantly) I didn’t hear any participants complain about either the pace or quality of the games.
Please note that I have suggested consideration of a 90 second increment, but 30 seems to be what you typed in a post which has now been quoted several times.
There is no question that some players habitually manage their time badly. Yet almost any player could easily manage to build up several minutes or more at an increment of 90 seconds or more - and I saw a suggestion of 150 seconds made here, which begins to approach the 180 second average of 40/2, 20/1, or the previous standard of 40/2.5, 16/1, which averaged 225 seconds per move.
It must be as true in chess as it is in fields such as writing, that over the ages, technology affects performance in some ways. Writers used much more redundancy when chiseling in stone or dipping quill into ink and illuminating expensive parchments. Erasers made it safer to try to write more concisely. Word processors make it easy to edit, but also to reorganize and be less coherent at a higher level. Likewise, egg times and fines led to games speeding up only slightly; analog clocks created cliffs such as moves 40, 60, etc., with scrambles followed by better quality, then repeated degradations. Today, we see hyperpreparation, followed by fewer deep thinks, and eventually, fairly shallow late game play with strong reliance on technique to see things through.
As an arbiter, I have observed and have read stories of the shallow play that is the best many masters can manage when reduced to a 30 second increment late in a game. These include cases of failing to obtain the theoretically expected results in relatively simple endings.
The high water marks for quality chess seem to have primarily occurred during the 40/2.5, 16/1 era at the top levels, because that allowed periodic deep thinks to solve hard problems, develop great conceptions during play, and identify points at which the game had changed and new plans were needed. Thanks to computers, today we see grand conceptions brought to the board, followed by what I described above. Frankly, even the return of adjournments would only cause a second moment of such hyperpreparation, followed by a gradual return to the same kind of play as described above, and maybe we don’t need to fear this after all. In any case, more thinking time per the above, plus computers, will lead to the best of all possible worlds, creatively.
Nowhere did I state or imply that 30i was your suggestion. 30i seems to be the current increment of choice, it’s the one increment setting that has received some practical use in the US, and it’s the only increment setting I have any experience in either playing or directing. So, it’s the one I talked about. I said that you suggested a longer increment, but I did not elaborate on that.
One problem with any time control with a base time of six hours or more, before increment (or even delay) is that you can only play one round a day. This would make it impossible to hold any norm event with more than one round a day - which would knock out any long-weekend Swiss from doing that.
I think G/120+30i is particularly useful for weekend Open sections. The reason is that rounds are usually scheduled seven hours apart, but the games typically last five hours or less. So, there’s actually a real break in between the games, where players can rest and recharge. Contrast this with the time control of 40/2, SD/1, d/5, where games can go six hours (or even a little longer).
Let’s say you ran a tournament, where the advertised time control was 90’+30"inc. Players must provide their own clocks.
Two players sit at the board, and neither have a digital clock. What do you tell them to set their clocks for? I’m assuming you would tell them G/120’ (given a sixty move game, that’s how long a 90’+30"inc would last).
Now, two other players show up with digital clocks, but neither of them have an increment setting (or neither player has any idea how to set their clock for increment, but they know how to do delay). What should they set their clocks for? G/90’+30"delay? What if they only know how to set a five second delay, would they set it for G/115’+5"delay?
As an aside, would these non-standard clocks invalidate a FIDE rated section, or just that game?
Per Rule 5F1d, I would tell them to set their analog clocks for G/90, and not compensate for the increment. There is a variant to 5F1d that allows for such compensation, which is one minute of base time added for every second of increment. A TD using that variant would tell the players to set their analog clocks for G/120.
In the case of a clock that can handle a delay of 30 seconds or more, the setting would be G/90, d/30. In the case of a clock that can only handle, say, a 5-second delay, the setting would be G/90, d/5.
They could invalidate a FIDE rated game, but not the entire section. The question of whether such a clock would invalidate a game would depend on the rating of the highest rated player in the section.
I tend to run larger Swisses, where the entry fee is 2-6 times the cost of, say, the DGT NA. Players who can afford such events don’t get much sympathy from me when they complain about the cost of a clock. They do, however, get accurate directions to the on-site bookseller. In the case of a smaller event, I would recommend the organizer have clocks on hand to avoid this issue if at all possible, though this isn’t required.
[EDIT: Left out part of the rule number I quoted.]
It probably depends upon the rating classes of the players. Of course for low-rated scholastics it shouldn’t matter as they move much faster than the control. For high rated players, the single time control is simpler and the players involved will (in most cases) know when to resign. The multiple controls tend to be most useful in sections with a wide rating range of low to high rated players, where many players do not understand when to resign - in particularly when they are just blitzing out move after move. When they reach the second control, they often have time to reflect and ask themselves why they are playing out this position. With a single-time control mental momentum will simply keep the game going for a much longer period of time - usually to the detriment of both players.
As a director, I prefer a single time control. It makes a lot of issues disappear, including complaints about things like move counters. As a player, I really don’t care either way.
Absolutely not. As of January 1, 2012, you set the analog clock for 90 minutes unless you have previously advertised that the time control for analog clocks is G/120.
In both cases, the clock is set for 90 minutes of base time. The USCF rules do not specifically address your hypothetical. In this case, if the clock can be set for a thirty second delay, I would consider that a reasonable approximation of a thirty second increment (the difference, of course, being that the players do not accumulate any unused portion of the delay). If they only know how to set a five second delay, I suppose I would allow G/90 d/5, but only reluctantly. (I really don’t like that solution – call it a “gut feeling” that it is incorrect.) I would also strongly suggest to the players to learn how to set the clock correctly.
FIDE rules are silent on this issue, as far as I can tell. I’m guessing that’s because the FIDE rules do not envision the situation arising.
Please note that I am proposing consideration of smaller base times and longer increments, and no addition of time other than the increment, no matter how long the game lasts.
Thus, the theoretical duration of a game would depend on the base time, the increment, and the number of moves the game lasts. It’s not clear how to interpret this in terms of “base time control” anymore.
For longer increments, lower base times will keep games within planned session durations, unless the games are extremely long. You can play mix and match with these factors, and construct scenarios to fit any number of sessions per day, from Tornado events to GM norm events.
Please bear in mind that the longer the game lasts, the more distress will occur for the players if the increment is small.
PS - Has anyone here experienced a game lasting 8, 10, or more hours using 30 second increments? While such games will be rare, they can occur, and could indeed be disruptive to a tournament (or at least, to the sleep a TD would otherwise try to get).
In the 1984 US Open in Dallas, played at 50/2.5, then 20/1 (repeating), GM Kudrin played a 163 move, 16.5 hour game against Flroida NM Gatlin. The game was played in three sessions: five hours during the scheduled mid-week round (7 PM til midnight), eight hours the next day (9 AM til 5 PM, adjourning for that evening’s round), and three and a half hours the day after that (9 AM til 12:30). Few tournaments could easily sustain such a game.
FIDE provides some guidance on this, I believe. For the purposes of calculating total reflection time, a game length of 60 moves is cited in the Handbook. This seems like a pretty reasonable number to me.
Of course, this is true. I simply wanted to be sure to mention that the length of session is an important consideration, especially in an event that plays more than one round per day, and a longer increment would need to take that into account.
In the limited runs I’ve seen for the 40/90+30i, SD/30+30i time control, I haven’t seen anything over 5.5 hours, and the vast majority of games have ended in under 5 hours. I’ve only seen this time control up close in some CCA events. My guess is that adding 30 minutes to the base time (or the equivalent in base time and increment time) to each time control would pretty quickly get the longest games over 7 hours, due to the increment on each move.
Sevan Muradian has probably held more increment games than anyone else in the country to this point, so he might be able to answer your question in the affirmative. In the meantime, the idea of longer increment and shorter base time would need some test runs. Maybe MCC would like to try?
At the 28th North American Masters (time control G/90 with 30 second increment), a first round game went some 140 moves or so and lasted five hours sixteen minutes before ending in a draw.
60 move game = maximum 4 hours
120 move game = maximum 5 hours
180 move game = maximum 6 hours
240 move game = maximum 7 hours
According to Wikipedia (source of all knowledge), the longest game in terms of number of moves was “Nikolić-Arsović, Belgrade 1989, which lasted for 269 moves”
In a G/90+30, that would be a maximum of 7 hours, 29 minutes.
Wouldn’t a game at 40/120, SD/30 i30 run into the 175 move rule soon after 8 hour mark?
My longest game was at Reno about ten years ago, when Weikel still used an infinite time control of 20/30. That’s three times slower than a 30 second increment! We played for about 140 moves in 8 1/2 hours, with adjournment after six hours of play. A fun story below, but mostly irrelevant to the thread.
A little more detail: It used to be (in the fourth edition) that, in a game with a sudden death time control (played with an “allegro clock”), the game was drawn after Black’s 175th move – I guess the general idea being “put the game out of its misery”
In the case of a smaller event, I would recommend the organizer have clocks on hand to avoid this issue if at all possible, though this isn’t required.