Skip to content
Back Rank

Over the Board · Sep 5, 2026 · 9 min

Chess Clock Rules: Delay vs Increment vs Bronstein, and How to Set Each

What d5, inc30 and SD/30 actually mean, exactly what simple delay, Bronstein and Fischer increment do to your remaining time, what US Chess and FIDE expect, and the setting procedure and verification check that work on any digital clock.

Annotated by Editorial Team

A quick, honest note: some links on this page are affiliate links. As an Amazon Associate we may earn a commission when you buy through them — at no extra cost to you. It never changes the price you pay.

Chess Clock Rules: Delay vs Increment vs Bronstein, and How to Set Each

Increment gives time back, delay withholds it, and Bronstein is delay wearing a different display. Those are the only three behaviours a modern clock has, and the mode you use is never a preference: it is whatever the organiser announced. So the skill worth learning is not which mode is better, it is reading a line like "40/90, SD/30, d5" or "G/90 inc30" off the wall chart and turning it into settings on the box in front of you, in about ninety seconds, before the round starts. Below: what the notation means, exactly what each mode does to the numbers, what US Chess and FIDE expect, the setting procedure that works on any digital clock, and the check that tells you the mode is doing what the label claims.

Reading the announced control

Every tournament announcement is built from four tokens.

  • G/90 means game in ninety minutes: the whole game, sudden death, no second period.
  • 40/90 means forty moves in ninety minutes, after which a second period starts. It always appears with what follows it.
  • SD/30 means the sudden-death period that follows: thirty minutes for the rest of the game.
  • d5 or inc30 or +3 is the per-move setting. A d is delay, an inc or a plus sign is increment, and the number is seconds.

So "40/90, SD/30, inc30" is the FIDE standard classical control: ninety minutes for the first forty moves, thirty more for the remainder, thirty seconds added from move one. "G/60; d5" is a single sudden-death hour with a five-second delay. "G/3; +2" is blitz with a two-second increment.

Two things follow from the notation. First, a control with an increment is longer than it looks, because the increment is real time: FIDE's own category definitions add sixty times the increment to the base when deciding whether a game counts as classical, rapid, or blitz, so a 3+2 game is treated as five minutes of playing time, not three. Second, a control with a delay is exactly as long as it looks, because delay time is never banked.

What each mode does to the numbers

Simple delay, called US delay on clocks sold into the American market, makes your clock wait a fixed number of seconds before your main time begins running on each move. Move within the delay and your remaining time is unchanged. Move slower and only the excess comes off. Your time can go down or stay level. It can never go up.

Fischer increment adds the fixed number of seconds to your remaining time when you complete a move. Move quickly and your clock grows. A player with a thirty-second increment who spends ten seconds a move for twenty moves is nearly seven minutes richer than when the sequence began, which is precisely why increment exists: it makes it possible to play a long technical endgame properly instead of on reflexes.

Bronstein delay produces the same remaining time as simple delay through a different mechanism. Your main time starts running immediately, and when you press the button the clock gives back whatever you used, up to the maximum. Arithmetically identical to simple delay; visually alarming, because during the delay your clock appears to be bleeding.

Worked numbers, forty moves at five seconds each. With a five-second delay and every move made inside five seconds, you finish with your full starting time. With a five-second increment and the same play, you finish with your starting time plus roughly three minutes. That gap is the entire practical difference between the two modes, and it is why an increment control feels calm at the end and a delay control does not.

Why it changes the endgame

Under delay, a sudden-death finish is played permanently at the edge. You have whatever main time is left plus five seconds a move forever, so a rook endgame requiring thirty accurate moves has to be produced at five seconds a move once your main time is gone. Technique degrades, and results in the last ten minutes of a delay event are noticeably worse than the players in them.

Under a thirty-second increment, the same endgame is playable. You will never be flush, but you can think for a minute on a critical move by banking seconds on the obvious ones. That is why FIDE settled on increment for classical play and why an increasing share of US events do the same at slower controls.

There is a related rules point worth knowing rather than guessing at: US Chess has long provided a claim procedure for sudden-death finishes where a player has little time and no losing chances, and the availability of that claim interacts with which per-move mode is in use. It is genuinely fiddly, it has been revised, and it is exactly the sort of thing to read in the current rulebook rather than take from a web page.

What the federations expect

US Chess does not certify or ban clock models. What its equipment rules establish is an order of preference, and the practical effect is consistent: when the announced control includes delay or increment, a properly set clock that can do it is standard equipment and outranks one that cannot, with the director resolving disputes about whose clock goes on the board. A five-second delay has been the American default for roughly two decades, which is why clocks sold into that market advertise a true delay display rather than only Bronstein.

FIDE's Laws recognise both increment and delay modes, and FIDE's own standard control uses a thirty-second increment from move one. A clock offering only Bronstein will not be refused at an American event, but a clock that shows a genuine countdown-style delay avoids a conversation with the director in the first place.

Setting any digital clock, in order

The menus differ. The sequence does not.

  1. Reset. Clear the previous setting before anything else. Half of all clock disputes at club level start with a clock still holding last week's blitz control.
  2. Choose the mode family. Countdown with no per-move setting, delay, Bronstein, or increment. On budget clocks, increment is often labelled "bonus" and delay is labelled "delay" regardless of whether it behaves as simple or Bronstein.
  3. Enter period one. Base time, and the move count if the control has a second period.
  4. Enter period two, if there is one.
  5. Enter the per-move seconds.
  6. Set the sides, so that the correct player's time runs when the correct plate is up.
  7. Test it before the round, using the check below.

Setting the three common clubbable clocks

The DGT North American is built for American controls and handles simple delay, Bronstein, increment, plain countdown, and multi-period programs with a move counter driving the period change. Its setup uses a small settings control and a sequence of flashing fields, which is quick once learned and opaque the first time; the shipped settings documentation lists the numbered programs and belongs in the bag permanently, because the mode numbering is not something anyone reconstructs under pressure.

The LEAP digital clock and the Chess Armory digital clock work the same way as each other: a numbered mode list in a small manual, a delay mode, a bonus mode for increment, and support for multi-period controls with limits that vary by revision. Both are perfectly legal for rated play when set correctly. Both are also clocks that the average tournament director has not set a hundred times, so if one of these is going to the event, the manual goes with it.

Whatever the clock, verify the per-move mode rather than trusting the label. Set two minutes of main time with a ten-second per-move value, start it, and watch one side for fifteen seconds without pressing.

  • The display sits at 2:00 for ten seconds, then starts counting down: that is simple delay.
  • The display starts dropping immediately, and jumps back up by up to ten seconds when you press: that is Bronstein.
  • The display starts dropping and, on the press, goes above 2:00: that is increment.
  • The display starts dropping and nothing is ever restored: the mode is not set, and this clock is not ready for a delay or increment event.

Move counters, second periods, and flags

On a multi-period control the clock uses its move counter to trigger the change from period one to period two. That counter is a convenience, not an authority: if the players disagree about whether move forty was reached, the scoresheets settle it, not the clock. Counters drift when a player forgets to press, which happens more often than anyone admits.

When a side reaches zero the clock should show a flag indicator and hold it, so a director called to the board afterwards can see what happened rather than take a player's word. Check that your clock holds the flag until reset, because a display that clears itself is useless in the one situation where it matters.

Before the round

Fresh batteries. The announced control programmed and verified with the fifteen-second check. Correct sides. The manual in the bag. Reset the clock the night before rather than at the board, because the two minutes before round one is the worst possible time to learn a menu.

FAQ

What is the difference between delay and increment?

Increment adds a fixed number of seconds to your clock after every completed move, so your time can grow. Delay makes the clock wait that many seconds before counting down, so your time can hold level but never increase. Over forty quick moves, a five-second increment banks roughly three minutes that a five-second delay does not.

Is Bronstein delay the same as US delay?

Arithmetically, yes: both leave you with identical remaining time. The display differs. Simple delay pauses your clock for the delay period, while Bronstein starts your main time immediately and restores what you used, up to the maximum, when you press. Bronstein looks like you are losing time you are about to get back.

What does 40/90, SD/30, d5 mean?

Forty moves in ninety minutes, then thirty more minutes to finish the game, with a five-second delay on every move. A d prefix means delay; inc or a plus sign means increment. Sudden-death periods are written SD, and G/ means the whole game in that time with no second period.

Does US Chess require a clock with delay?

It does not certify or ban clock models, but its equipment rules give preference to a properly set clock capable of the announced delay or increment over one that is not, with the director settling disputes. A five-second delay has been the American default for around two decades, which is why US-market clocks advertise a true delay display.

How do I check my clock's delay mode is set correctly?

Set two minutes of main time with a ten-second per-move value, start it, and watch one side for fifteen seconds. Sitting still for ten seconds then counting down is simple delay. Dropping immediately and jumping back on the press is Bronstein. Going above two minutes on the press is increment. Dropping with nothing restored means the mode is not set.

Reader favorites

The short list — see the full ranking on our best-gear page.

A quick, honest note: some links on this page are affiliate links. As an Amazon Associate we may earn a commission when you buy through them — at no extra cost to you. It never changes the price you pay.

26 guides published

19 products vetted

727 reader price checks

From the first run

Original designs, made to order.

Browse the shop →