Monday 13 July 2015

Push counters for clocks

A few days ago I was asked to comment on whether "Push Counters" (to use their proper name) should be used in chess competitions. For those unfamiliar, some models of Digital Clocks (including all official FIDE clocks) can keep track of the number of times the clock has been pressed during the game. This allows the clock to add additional time for second and subsequent time controls.
The correctness of this process assumes that the number of pushes equals the number of moves played. If players accidentally push the clock too many times, extra time may be added before the first time control has been reached. On the other hand, if a player forgets to press their clock, then players may wonder why they have not been given extra time.
When this feature first came in, I was wary of using it. I figured that there would be too many problems, and arbiters would have to interrupt games at move 38-42 to fix these issues. It turns out that my fears were unfounded.
The first big event where I used "push counters" was at the 2012 Queenstown tournament in New Zealand. It was a large event, and it had a reasonably complicated time control (40/100m, 20/50m, G/15m+30s per move from move 1). The clock needed to keep track of 40 moves, then a further 20 moves to add the required time. Apart from being a little slow to set before the first round, there were very few problems during the event. The number of extra/insufficient pushes problems could be counted on 1 hand, and in a couple of instances, a fault with the clock was the problem. Certainly the arbiting team checked boards around the first and second time controls, but the system worked as advertised. It also had the added benefit of players feeling a little more comfortable, as they always knew how much time they had available to them.
Since then I have used "push counters" more than I have not. For the Doeberl Cup we use them, and there have been very few problems. At this years Zonal, there was 1 major problem (caused by players hitting the clock too many times), and a couple of minor ones (not hitting the clock enough), but not even close to a number that would make me rethink using this option.
The alternative, which is to allow the clock to run to 0 before adding time, has its own problems. Anecdotally, players may rush their moves after the time control, worried about what will happen if they allow their clock to run down (As an aside, Magnus Carlsen lost against Topalov in part because he assumed that this system was in place, and expected extra time once his clock reached 0). When this setting has been used, I am often asked by players at the board about when they get extra time, meaning I have to interrupt the game anyway.
There are also some time controls that require the use of a push counter. Some tournaments only start an increment after a certain number of moves (40 or 60), so the clock has to know when this move number has been reached.
As for whether you "must" use this option, the choice is still up to arbiters. The FIDE Arbiters Manual mentions their use, without mandating it. Certainly this setting is not used at the Olympiad, although I think it would be better if it did. And if you are an arbiter I would recommend you do use it, both to gain experience for bigger/more important competitions, and that players seem to prefer having time added when they reach the time control, rather than at some random time afterwards.


Ian Rout said...

What's the ruling if time is added early (as would happen through an extra clock press earlier) and a player then exceeds the initial time limit? For instance 30 minutes is added after move 39 and after playing move 40 White has 28 minutes left - clearly White must have exceeded the time limit, but on the other hand can claim to have been misled by the clock.

Shaun Press said...

It is one of these things that should not happen if the Arbiter is on the ball. At the Zonal we made sure we checked the clocks and the scoresheets to see that the time was added at the correct point.
If it did happen, it would of course depend on the specific situation. But in general if a player only had 38 or 39 moves on their scoresheet but at the same time noticed their time had gone up by 30 minutes, then the 'misled by the clock' excuse probably would not hold much water.

Garvin said...

Hello Ian and Shaun, when this type of time control came in and the option of whether or not to set the push counter was preferable, it is exactly the situation that Ian Rout had in mind that was my thinking for believing that not using the push counter was preferable.

In the tournaments that I have done, I have always given clear instructions on how the clocks would function and when queried, the players have understood. But I have not been involved in 250 player events.

Also I can imagine it would be a different situation again where the increment comes in at a certain move point, rather than from move one. Shaun's explanation has changed my opinion on the matter to a degree and I am now more in favour of setting the push counter and would say so on chesschat, except the conversation has degenerated into a waste of time.

Anonymous said...

Why don't the clocks display the counter? Players could then check with their scoresheet. With 30 second increments, they are expected to keep score throughout.