Objection to Digital clocks

[Kevin Thurlow]

When digital clocks appeared in the Surrey League, a rule was rapidly brought in saying that opposing clubs could insist on the use of analogue clocks. But that was almost certainly bias against the club in question, who apparently were "intimidating" the opposition by using digital clocks. Yes, it sounded crazy at the time, and it still does. Most clubs now use them.

When I reach the age of 65, I will insist on the use of two Harrison timepieces, linked by a golden lever.

[Nick Grey]

I like the quietness & being able to see how much time left a lot easier.

[NickFaulks]

This reminds me of 1986's "Big Bang" in the London Stock Exchange. The American newcomers were accustomed to using hand signals to relay information out of the options trading pit, and the established British firms managed for a while to get these banned on the grounds that their own traders were unfairly disadvantaged by being too stupid to learn hand signals. I'm not making this up.

[Stewart Reuben]

When the Stock Exchange considered changing to decimal fractions for stok market prices, there were objections by some. 'Fractions such as 1/4, 1/8, 1/16 are more accurate than 0.25, 0.1, 0.05.' I couldn't have made that up.

This has been an amazing discussion. Nobody has mentioned which digital clocks are being used. They are not identical, differing simply in the cost.
Chronos are best, but very expensive.
DGT 2010 are very reliable. DGT Excel less so, but connect to digital boards. Don't use DGT 2010 and Delay mode (almost never used except in the US), it doesn't work properly. DGT 2010 US model does. DGT is more expensive than some other clocks, but superior.
Saitek do not conform with the FIDE Laws of Chess.
And so on.
None of the clocks have a visible move counter except Chronos, despite what has been said. They have a push counter. If you rely on that to add the second time control quotient there is the danger that the clock has registered the wrong number of pushes. The most likely reason for this is that the clock wasn't started before the first move and push, or there has been an extra pair of pushes.
The addition of the extra quotient of time was used both in the World Youth and the current World Senior. There was a problem in the World U14 in one game where the extra time was added after 40 pushes, but both scoresheets correctly showed only 39 moves had been made. White then sat and thought until his time ran out and lost on time.
The arbiter has to be vigilant and make certain the clocks are right. In the absence of an arbiter, as in league matches, this might cause problems.
When both players believe they have made 40 moves, they can 'interrogate' the DGT clock so that it displays the number of pushes. I saw that done yesterday in the World Senior during play. The arbiter then adjusted the clock during play. Perhaps she was alerted to the fact that the computer showed a different tally to the players. All the games here are on digital boards. Of course occasionally such equipment breaks down.

The Chief Arbiter asked me here which is best. For the clock to add the extra time after the programmed in pushes or to wait until one player has used up all his first period before adding the time. I told him there was no simple answer to that question. I find having to wait until the first time control period has run out is confusing (but I am a wrinkly). But the pushes method is dangerous, except with Chronos.

Anybody who has to buy new clocks, because he old mechanical ones are wearing out, who buys new mechanical clocks is insane, or insufferably mean.

[Roger de Coverly]

Saitek do not conform with the FIDE Laws of Chess.

Which particular FIDE Law do you have in mind and which Saitek clock? Given that DGT was or is the "official" FIDE Clock, for which it pays or paid FIDE, it was always likely that the FIDE Laws would be written so as to fall in line with the DGT programming.

When Kirsan first wanted to abolish seven hour games, the initial idea was that you played x in y and then added z, but with the increment only after x moves. That had to be abandoned when it was realised that the first generation of DGTs didn't support such a move rate.

I had an email advert that there's a new DGT on the market, one which displays hours, minutes and seconds throughout, rather than confusingly switching backwards and forwards. The talking clock used by one or two of the Braille players does this as well.

[Sean Hewitt]

I had an email advert that there's a new DGT on the market, one which displays hours, minutes and seconds throughout, rather than confusingly switching backwards and forwards. The talking clock used by one or two of the Braille players does this as well.

That's the DGT 3000. It is a replacement for the DGT XL (the one used with Live Boards). I have one, and it certainly has a very clear display.

[Ian Thompson]

Don't use DGT 2010 and Delay mode (almost never used except in the US), it doesn't work properly. DGT 2010 US model does.

Can you clarify this comment please:
1. What is the fault and how do you reproduce it?
2. Does it apply only to the original DGT 2010, or also to the revised one, with blue buttons, sold from November 2010?
3. I wasn't aware there was a US version of the DGT 2010. How can you tell the difference between a US one and a non-US one (or do you mean it works on the DGT North American clock)?

I'm rather interested in the answers to all of these questions because I bought a DGT 2010 in America 2 years ago.

[Stewart Reuben]

The Saitek clock freezes the display in rapidplay when one player has exceeded the time limit. DGT follows the FIDE Law that a player must claim a win on time if his opponent's flag has fallen. It is a conscious intellectual action and should be made while the player still has time remaining on his clock. If th display is frozen, he could play moves and think for 5 minutes, then notice the flag fall and claim the game.

I may have called the DGT clock marketed in the US by the wrong name. It may well be DGT North American.
I have no idea whether the revised DGT 2010 is different or not to the original one.
I was recently shown the defect in a DGT 2010 in Durban. He set the clock with 5 seconds delay. Set the player's time at 3 seconds and now pressed the clock. The 5 seconds was not added until AFTER the move. Let us say 1 second left. Then the time becomes 6 seconds.
But what if he didn't move for 3 seconds +? Then the flag fell, even though he had 2 seconds left.
The correct way is to add the 5 seconds so that it becomes 8 seconds. That ticks down and the 3 seconds does ot start ticking down until after the 5 seconds delay.

Ian since delay is so little used outside the US, does it really matter for your clock?

[Roger de Coverly]

DGT follows the FIDE Law that a player must claim a win on time if his opponent's flag has fallen. It is a conscious intellectual action and should be made while the player still has time remaining on his clock.

The thing is that DGTs break this as well. If playing with an increment using one of the standard settings, the DGTs will also freeze. That's not new, I saw it back in 2000 during the Sunday Rapidplay at the British.

http://www.chessvibes.com/?q=reports/dg ... k-incident

This was explained as an attempt by an arbiter to hack the clock so that it couldn't display which flag fell first.

If you are playing with an increment it makes perfect sense to freeze the time as otherwise you could have players flipping between positive and negative time.

Analogue clocks never had an indicator of which flag fell first, so rules were needed to cope with the case when both flags were down. Other than creating work for arbiters, there is little point in perpetuating such rules into a digital era.

[Ian Thompson]

I was recently shown the defect in a DGT 2010 in Durban. He set the clock with 5 seconds delay. Set the player's time at 3 seconds and now pressed the clock. The 5 seconds was not added until AFTER the move. Let us say 1 second left. Then the time becomes 6 seconds.
But what if he didn't move for 3 seconds +? Then the flag fell, even though he had 2 seconds left.
The correct way is to add the 5 seconds so that it becomes 8 seconds. That ticks down and the 3 seconds does ot start ticking down until after the 5 seconds delay.

My clock appears to work correctly. I set a time control of 3 seconds initial time + 5 seconds delay. The clock correctly showed 8 seconds available for the first move. It also correctly goes back to 8 seconds if you move in less than 5 seconds, and reduces the remaining time if you don't.

I'm wondering whether the fault you saw really was a fault. If you set the initial time to anything you like with a 5 seconds delay and then adjust the initial time, it shouldn't add 5 seconds on to the new initial time you've set because it's already done that either when the clock was first set (if the game hasn't started) or on completion of the previous move.

Ian since delay is so little used outside the US, does it really matter for your clock?

It matters to the extent that I've played in America before, and may do so again, using this clock.

[David Sedgwick]

I was recently shown the defect in a DGT 2010 in Durban. He set the clock with 5 seconds delay. Set the player's time at 3 seconds and now pressed the clock. The 5 seconds was not added until AFTER the move. Let us say 1 second left. Then the time becomes 6 seconds.
But what if he didn't move for 3 seconds +? Then the flag fell, even though he had 2 seconds left.
The correct way is to add the 5 seconds so that it becomes 8 seconds. That ticks down and the 3 seconds does ot start ticking down until after the 5 seconds delay.

My clock appears to work correctly. I set a time control of 3 seconds initial time + 5 seconds delay. The clock correctly showed 8 seconds available for the first move. It also correctly goes back to 8 seconds if you move in less than 5 seconds, and reduces the remaining time if you don't.

I'm wondering whether the fault you saw really was a fault. If you set the initial time to anything you like with a 5 seconds delay and then adjust the initial time, it shouldn't add 5 seconds on to the new initial time you've set because it's already done that either when the clock was first set (if the game hasn't started) or on completion of the previous move.

I've just tested this with my DGT2010 and my DGT North American.

Suppose that the delay is 5 seconds and you adjust the clock leaving one player with 3 seconds, as an arbiter might do if imposing a time penalty.

The DGT2010 then counts down from 3 seconds. If the player doesn't move in that time, game over. If a move is made, the clock goes back to 3 seconds. So the player is left with 3 seconds per move rather than 5.

The issue doesn't arise with the DGT North American. That clock doesn't show the delay at all, but the clock doesn't count down until 5 seconds have elapsed. So if your time is adjusted to 3 seconds, you have 8 seconds before you lose and 5 seconds before your 3 seconds starts to disappear.

Moral: If imposing a penalty when a DGT2010 is in use, it's unfair to leave a player with less than 5 seconds (or whatever the delay is).

[Roger de Coverly]

The issue doesn't arise with the DGT North American. That clock doesn't show the delay at all, but the clock doesn't count down until 5 seconds have elapsed.

I had always rather assumed that delay meant what it said. In other words the clock would not restart until five seconds had elapsed. I thought the whole point of delay as opposed to increment was that you never gain any time. So how do European DGTs handle it? If the clock shows 20 seconds, presumably that jumps to 25 seconds when pressed? If you can move within 5 seconds, do you keep the time, or does it reset to 20 seconds?

[David Sedgwick]

So how do European DGTs handle it? If the clock shows 20 seconds, presumably that jumps to 25 seconds when pressed? If you can move within 5 seconds, do you keep the time, or does it reset to 20 seconds?

Suppose the clock is showing 20 seconds and the delay is 5 seconds.

When a "European DGT" is started by the other side being pressed, it counts down from 20 seconds. If the clock is pressed within 5 seconds, it resets to 20 seconds. If the clock is pressed after more than 5 seconds, it resets to 5 seconds more than it was showing.

With the DGT North American, 5 seconds elapses before there is any visible change to the setting. The clock then counts down until it is pressed, when it stops. When it is restarted, once again nothing happens for 5 seconds.

[Ian Thompson]

The issue doesn't arise with the DGT North American. That clock doesn't show the delay at all, but the clock doesn't count down until 5 seconds have elapsed.

I had always rather assumed that delay meant what it said. In other words the clock would not restart until five seconds had elapsed. I thought the whole point of delay as opposed to increment was that you never gain any time. So how do European DGTs handle it? If the clock shows 20 seconds, presumably that jumps to 25 seconds when pressed? If you can move within 5 seconds, do you keep the time, or does it reset to 20 seconds?

The European clock doesn't work like that. When you move it adds the delay on for the next move. e.g. you play your 50th move and stop the clock with 60 seconds remaining. It then adds on 5 seconds to indicate that you have 65 seconds to make your 51st move.
1. If you take 10 seconds over your 51st move you stop the clock with 55 seconds remaining and it adds on 5 seconds to show that you have 60 seconds to make your 52nd move.
2. If you take 3 seconds over your 51st move you stop the clock with 62 seconds remaining and it adds on 3 seconds to show that you have 65 seconds to make your 52nd move.

[Roger de Coverly]

I was recently shown the defect in a DGT 2010 in Durban. He set the clock with 5 seconds delay. Set the player's time at 3 seconds and now pressed the clock.

Given the description of how the European models of DGT operate, it would be impossible in normal use for the clock to read 3 seconds. It would have got there by adding 5 seconds to minus 2, so presumably would have stopped to indicate flag fall before it reached minus 2. If there was another time period, that presumably would have been added on, so showing, say, fifteen minutes and three seconds is plausible.

I suppose the clocks are trying to follow 5.1 of the FIDE handbook

Requirements for electronic chess clocks

(b) The display at all times should show the time available to complete a player’s next move.