Beginning Displayed Time With Increment

Running Chess Tournaments
1

This past Monday we decided to try increment timing in a 4 round event at our evening club meeting over 4 Monday evenings. I am the TD of this event.

I got my Chronos out and found the CH-P5 mode to be appropriate for our needs, G/60, inc/30. I set that clock accordingly and the beginning time display on both clock displays was, 1:00:00. I tried out the clock in this setting and it indeed added or accumulated the unused increment time, as expected.

I also have a DGT 3000 so I got that clock out so we would have an extra increment clock to use in the event if needed. I set that clock for the Fischer Bonus mode and this time when the clock was ready, the clock displays showed, 1:00:30. Just like the Bronstein method of delay, the clock added the increment time to the main time at the beginning.

So now, we had a discrepancy of beginning clock times.

I called my friend and my expert on FIDE and DGT clocks, Sevan Muradian. Sevan told me that the setup of the DGT clock was correct per the rules of FIDE and DGT had their clocks set up to comply with FIDE rules. Well, that makes sense, sort of. Our Monday night club is USCF rated only and we really don’t do FIDE ratings locally like this.

Today I had a few minutes and decided to look it up in the latest edition of the USCF rulebook that I have on my KIndle app. I found the time controls and styles explained, but I did not see any mention of how the display of the clock should begin with increment.

Now, I know the Bronstein method which does the delay time after the move starts the display with the delay time added to the main time. And I know, from experience and the rulebook that the US Delay and Bronstein styles are equivalent.

The increment time is always added after the move. And there is no before-move increment option in existence anywhere.

With a 30 second increment, the DGT and any other clock that has the increment time added to the main time at the beginning of the game is adding one minute to the total game time compared to using a Chronos or Excalibur GameTime II where the time is not automatically summed, main time and increment time.

To make all the clocks equal we will need to modify the clock timing. If we go with the FIDE rule, we will need to add 30 seconds to the main time of the Chronos and GameTime II (I don’t know about other increment clocks than these 2 and the DGT brand). If we want to have the main time be displayed and used at the beginning of the game, we will need to set the DGT clocks for 59:30 as the main time.

What do you TDs practice in your tournaments with increment timing?

2

We haven’t used increment timing, yet. I think it is a good thing for the initial display to show 1:00:30 as it is an indication that the clock is set for increment and that timing method will be used. With other settings it is not so clear. The extra time is a minimal problem. It reminds me of adding a minute to analog clocks because of potential inaccuracy to the mechanism. As White, you probably should have those 30 seconds anyway to think and get settled before making your first move. Then with the opponent’s initial 30 seconds ticking down you have time to write down your first move.

3

DGT does this right. Figure how much time has been used on the clock after 40 moves of 40/90 (or whatever). With the 30 seconds added at the start, it’s exactly 110 minutes. Without the 30 seconds added, it’s 109 minutes and 30 seconds.

This is one of the very few things Chronos does not do the best way. What makes a fella scratch his head is that it does add the time at first clock press in Bronstein/Adagio mode. Go figure. The Excalibur Game Time II treats increment the same way Chronos does, btw.

But—the one time we tried increment at a club I used to help run, we did not adjust starting time on clocks. For amateurs playing 40/90+30 sec. the difference was not enough to worry about, we decided. At least it was the same for both players in a given game, no matter which clock was used.

The difference between G/25 d0 (or G30 d0) on an analog and G/25 d5 on a digital is far greater and we let that happen. Not worth worrying for friendly rated games at a club, with amateur players and very small prizes. That’s how we did it a few years ago in the wilds of northwest NJ, anyway.

4

Eric Mark nails this one – as usual.

Look at it this way. It is common practice to equate (for some purposes) each 1 second of increment time to 1 minute of main time. This is a ratio of 60 to 1 (there are 60 seconds in 1 minute). Thus, a 60-move game is implicitly assumed when proposing this 1-second-equals-1-minute approximation.

60 moves is a reasonable, though arbitrary, assumption. Most tournament games last fewer than 60 moves, but a few games last longer, so we’re fairly well covered in “average” cases.

For example, FIDE rules (for certain types of tournaments) require that there be no more than 12 hours of playing time in one day. For 3 rounds a day this would mean 4 hours per game, or 2 hours per player per game. Thus, without increment, the slowest allowable time control in a FIDE-rated event with 3 rounds per day would be G/120.

But FIDE likes organizers to use increment, so it has adopted the above 60-move convention. Thus, with a 30-second increment, the slowest allowable time control in a FIDE-rated event with 3 rounds per day is G/90 inc/30. (Adding the 30-second increment is compensated for by subtracting 30 minutes from the main control.)

So, how do you use the 60-move convention to demonstrate that G/90 inc/30 is equivalent (as regards total time) to G/120 inc/0? That is, should the clock add before, or add after?

Let’s illustrate with a hypothetical, oversimplified example.

Suppose each player takes exactly 1 minute per move at G/120 inc/0. After black completes his 59th move (and presses the clock), each player has used 59 minutes, and thus has 61 minutes to complete his 60th move.

Now, with G/90 inc/30, if the clock fails to add 30 seconds before the first move, then after Black completes his 59th move, each player will have had 30 seconds added only 59 times, for a total increment (so far) of only 29:30 per player. Thus, the 59 minutes that each player has used is composed of 29:30 of increment time, plus 29:30 (59 minus 29:30) of main time. This means that the remaining main time for each player is 90 minutes minus 29:30, or 60:30, which is 30 seconds short of the 61 minutes he should have had, based on the above no-increment version.

If you’ll bear with this logic just one more time, then if the clock does add 30 seconds before the first move, then after Black completes his 59th move each player will have had 30 seconds added 60 times, for a total increment so far of 30 minutes per player. Thus, the 59 minutes that each player has used is composed of 30 minutes of increment time plus 29 minutes (59 minus 30) of main time. So the remaining main time for each player is now 90 minutes minus 29 minutes, or 61 minutes, the same as for the no-increment version.

Whew!

If you’ll look around a bit, you’ll find that Bronstein, like increment, is implemented differently on different clock brands. The Chronos adds the extra delay time at the start of the game, the old Saitek (dark gray with small metal buttons) does not.

The USCF rulebook simply declares delay and Bronstein to be equivalent without actually examining the question of add-before vs add-after. Only the add-before version is truly equivalent. Add-before differs from add-after by a grand total of 5 seconds (or whatever the delay or increment is) per game.

Do the former, for sure. Go with G/60:30 inc/30 on the Chronos and GameTime, G/60:00 inc/30 on the DGT. It’s far more logical (if you followed the math above) and it’s the international standard.

Interestingly, most users of the DGT don’t even notice the extra 30 seconds at the start of the game. The DGT does not display seconds until the main time goes below 20 minutes. So at the start of the game it would display 60 minutes, period. Only by starting the clock and observing that it runs for 30 seconds before the display changes from 60 minutes to 59 minutes would you know that the time had been added.

If the main time is less than 20 minutes to begin with, you’ll see the difference immediately. For example, if you set the DGT for blitz G/3 inc/2, you’ll see 3:02 displayed before the clock is even started.

One caveat: Some players in your event, using a Chronos or GameTime, and having read your post, might try to claim, at the end of the game, that they didn’t actually time-forfeit because they were supposed to have had 30 more seconds added. Don’t let anybody get away with this! Announce that, if 30 seconds is not added at the start, too bad – they’ll just have to live without the extra seconds.

Bill Smythe

5

As a quick point, the DGT 3000 shows seconds at all times and not only when under 20 minutes as the older DGT models all do.

6

Agreed, and the novelty of Eric and I agreeing on a clock issue speaks volumes.

Oh, now, let’s not get carried away. :smiley:

The DGT approach is correct, and that’s the answer one should give on a certification exam. If I’m doing a FIDE rated event, where equality of conditions is highly valued and potentially critical to ratability, I make sure Chronos and Excalibur users manually add one increment to the base time.

But for a large event that’s not going to Lucerne for rating? I’ll tolerate a 30 second differential from the ideal on some games to get a round off on time.

7

Lucerne - haha. That would be Elista in Russia.

8

That’ll teach me to post during a bout of insomnia. You are correct; the FIDE rating office is in Elista.

9

I had thought there was some clever trick to get a Chronos to do it the FIDE way, but after a bit of experimenting I cannot find one, so I guess adding it manually at the start is indeed the way to go.

Speaking of Chronos, what the devil are modes CH-P6, CH-P7, and CH-P8 for? These are like CH-P5 (sudden death with increment), except they have two, three, and four sudden death with increment segments.

Note that if you use the same increment for all the segments, then they are all equivalent to CH-P5 with the time set to the sum of the times from all the segments, so presumably you’d only use these multi-segment sudden death modes if you were going to have different increments. Say SD/30 inc/30, followed by SD/30 inc/5 in CH-P6. That works like you’d expect–you start with 30 minutes and get 30 second increment per move, and if you ever flag in the first segment, the time from the second segment is added and you continue, now with a 5 second increment.

(I was looking at CH-P6 because I had the idea that maybe if you set up SD/0 inc/30 followed by SD/60 inc/30, and then quickly hit white’s button, then black’s button, then the pause button, you’d get the increment added to both sides, flag the first segment and roll over to the second, getting the 60 minutes from the second segment added, and end up stopped with both clocks at 60:30, giving a way to closely mimic the FIDE method. This failed. The clock was not happy with a 0:00:00 segment).

10

You answered your own question. CH-P6 is for increment timing with two controls, such as 40/90, SD/30 Inc-30, while CH-P7 is for increment timing with three controls, such as 40/100, 20/50, SD/15 Inc-30.

It’s possible to use CH-P5 for these controls, but you would need to re-set the clocks after 40 moves, or after move 40 and 60 in the latter case. That’s not good, thus the other modes.

I miss my old Chronos.

11

If you don’t have one now because of the supply problems that left them unavailable for a long time, that seems to have been fixed. Wholesale Chess has consistently had them in stock for months now. Currently they have black and white, with push button or touch button.

The Chess Store also has had decent stock, although they only have the white push button model currently.

12

I sold my Chronos last fall due to a bleak financial situation. Two days later I got an unexpected job offer.

The job keeps me from playing much rated chess, for now. For the USATE I bought an Excalibur Game Time clock, but found the noise when pressing the plungers bothered me more than I recalled, so bought a DGT NA. Total cost for both was a few bucks less than a new Chronos would have been, and I had heard the latest models had quality-control issues.

So for now, with two clocks and limited chances to play I don’t plan to buy any more clocks. If I get more opportunities for rated play and Chronos reliability continues to bounce back—could be a great birthday gift to meself…

13

Practical resolution. Last weekend during the Denver Open (129 players) I completely ignored the difference and instructed (the 1-2) players who asked to ignore it as well. The chance they screw something up changing their clock settings as a round starts is too great. Two players at one board have equal time with the correct base time – that was enough for me.

14

Since not FIDE rated, that’s a practical solution.

If it had been FIDE rated, you would need to be pedantic, because adding 30 seconds only after move 1 gives the players 119:30 to make 60 moves. With a player in the field with a rating of 2200 or more, a player must have 120 minutes to make 60 moves.

And yes, in that case, it matters.

15

Makes sense. FIDE makes the rule clear and there are certainly consequences, so the system is self enforcing. I treat FIDE discussions as a different dimension – something to be aware of but not in my orbit (yet? maybe one day? I have no great desire to pursue that).

The practical matter is whether anyone would notice or even report it to invalidate an entire tournament. Of course one disgruntled player has made a larger stink about smaller things before. Can’t have any assumptions with chess personalities :open_mouth:

16

Just post, during registration, a notice requesting that players with Chronos or GameTime add 30 seconds to their main time at the start.

In the same notice, also mention that failure to do the above will never constitute grounds for contesting a time-forfeit claim.

Bill Smythe

17

I don’t understand. If the clock is set to 120:00 at the start, but increment is added after the move, there is no way for the flag to fall before 120:00 has passed. If increment is added before the move, there is no way for the flag to fall before 120:30 has passed.

18

The time control isn’t G/120, inc/30. It’s G/90, inc/30. Under FIDE rules, for a game to be standard rated in a tournament where at least one player has a FIDE rating of 2200 or above each player has to have at least two hours for a game which lasts 60 moves.

19

The clock was set for 90:00 at the start with the 30 second increment.

If it is set for 90 minutes plus 30 seconds per move then when the 30 seconds is added is sometimes significant. If it is added before the move then the adds for the first 59 moves means that 29:30 was added as of the start of move 59 (119:30 to play 59 moves) and the 30 seconds for move 60 is available for move 60 and gets it to 120:00 available for the first 60 moves. If it is added at the end of the move then the first 58 moves adds a total of 29:00 and you have 119:00 to play 59 moves, with the 30 seconds add after that move meaning that you get 119:30 to play 60 moves and the 30 seconds after playing move 60 is only available for move 61.

20

My experience with several clocks set for Bronstein is you set it for the control - ex. G60 B5 so the clock shows 1:00:00 or 60:00. If you then start the clock and examine both time counters, you will see that they actually started with 60:05 or 1:00:05. The logic to this as explained to me was that if you flagged, you never got the 5 seconds back so you had 5 less minutes than if using delay. I believe the logic was to have the clock display the published time control but still give the player the extra time.

I believe that the same logic applies to increment. The question then becomes should you have the display show just the time control or the players actual time. In an event using both delay and Bronstein, it is nice to have the starting display for both clocks to be the same so I like the hidden added time. In Increment where all clocks should be set with increment, it does not matter as long as they show the same starting time. In this case I prefer showing the added 30 seconds so the clock shows the actual remaining time throughout the event. That said, either method works for me.
Regards, Ernie