How to get good at a paddle game

Everybody who picks one up does the same thing for the first ten minutes: puts the middle of the paddle under the ball and hopes. That is the one shot that changes nothing. Here is what the good players are actually doing, and why the ball speeds up exactly when it does.

The paddle is an aiming device

The ball leaves the paddle at the speed it arrived. The only thing that changes is the direction, and that is decided by one number: how far along the paddle it struck. Dead centre sends it straight back the way it came. The far edge throws it out at about sixty degrees from vertical. Everything in between is in between.

This is the whole game. Chasing the ball — getting the middle of the paddle under it — is the reflex, and it is a wasted shot: the ball goes back where it came from and you have learnt nothing and changed nothing. The skill is deciding where you want the ball to go before it arrives, then putting the correct part of the paddle under it.

One consequence catches people out: a wider paddle is a blunter instrument. The angle depends on the fraction of the paddle's half-width, so the same two centimetres of movement aim much more coarsely on a wide paddle than a narrow one. A power-up that widens the paddle makes you harder to kill and worse at steering. Late waves narrow it and hand you the precision whether you asked for it or not.

Dig a tunnel

The single biggest scoring move, and it has been since 1976. Instead of skimming bricks off the bottom row across the whole width, work one column down the side. Clear a channel from the bottom row to the top and the next ball that goes up it does not come back — it gets trapped between the roof and the top of the wall and bounces along it taking a brick with every contact, for free, while you stand still and watch.

The top rows score most precisely because of this. They are the hardest to reach honestly, so reaching them dishonestly ought to be worth something.

It is worth losing a ball to set one up. A tunnel that works will pay for the ball several times over, and a cautious game that never opens one will stall somewhere in the middle of the wall with the ball getting faster.

Know where the speed-ups are

The ball does not accelerate smoothly. It steps, at four fixed moments, and all four are worth anticipating:

• After four bricks. Small, early, barely noticed.
• After twelve bricks. Noticeable. This is where a comfortable rally stops being comfortable.
• The first time the ball reaches the second row from the top. A real jump.
• The first time it reaches the top row. The biggest one — and the one that ends most games.

That last step is a trap built into the design. You finally break into the top row, feel pleased with yourself, and the ball comes back a third faster than the one you had just got used to. Expect it. Be in position before it happens rather than reacting after.

What the original got wrong, and what it got right

Two faults, both from the same cause. If the ball keeps meeting the paddle near the middle, its angle flattens out, and a flattened ball crosses the field almost horizontally — thirty seconds of travel, nothing you can do about it, and no fun in either direction. The same flatness is what makes the last brick of a wave take minutes to hunt down.

Both are fixable without changing the game: hold the ball's vertical travel above a fixed share of its speed, so every rally makes progress up and down the field, and raise that floor further once a wave is nearly clear. Brick Rebound does both.

What should not be fixed is the tunnel. It looks like an exploit — the ball goes somewhere the designer did not put a paddle and scores without you — and it is the best thing in the game. Taking it out would leave a reflex test. Leaving it in makes the whole thing a planning problem.

A note for anyone writing one

The bug every implementation ships at least once is the ball passing through a brick. A fast ball covers more distance in one frame than a brick is tall, so a naive "where is the ball now, is it inside something" check misses the brick entirely and the ball carries on through the wall. The fix is to cut each frame's movement into steps smaller than the smallest brick and test each step. It is worth testing for directly — play a few long games and assert the ball is never found inside a live brick, rather than trusting that the guard you wrote is reached.

The other one is subtler: never serve the ball straight up. A vertical serve off a centred paddle comes straight back to a centred paddle, for ever. A player who does nothing is never punished, and the game has no ending.

Playing it here

Brick Rebound is free, needs no account, and runs entirely in your browser — on a phone, dragging is the best way to play it and the closest thing to the cabinet's dial. Three modes: Classic has no power-ups at all, Modern adds four, Gentle adds a wider paddle and a slower start for learning the angles. Your best scores stay on your device.

Related: Invader Drop, Asteroid Drift and Maze Chase — the other three originals, all built the same way.