Brick Buster pays 25 points for a brick and multiplies that by a number which starts at 1, gains one for every fourth brick broken, and stops gaining at 8. The multiplier is wiped the instant the ball touches the paddle. Forty bricks broken without the ball ever coming back are worth 5,375 points; the same forty bricks, caught on the paddle between each one, are worth 1,000.
Nothing about those forty bricks changed. Same bricks, same hits, same seconds of play. The only difference is how long the ball stayed away, and the score is the game's opinion about whether that deserved to be paid for.
That opinion is the design, and it is usually argued about in the wrong place. A total that only adds up is a stopwatch with a list attached: the tenth action pays exactly what the first one paid, so the number records endurance and nothing else. A multiplier is a claim that the tenth action should be worth more than the first, and the load-bearing part of the claim is never the size of the multiplier. It is the thing that ends it.
A flat total is a measure of attendance
Flat scoring has one genuine virtue. Two runs of the same length compare without knowing a thing about the game, which is why the arcade cabinet needed nothing but a number and three initials. It is a fair ranking of how long somebody lasted.
Its cost is that it makes the slowest strategy the best one. When every action pays the same, the optimal run is the careful one that never risks anything, and a difficulty ramp that tightens every thirty seconds is arguing with a scoreboard that does not care. A ramp needs an incentive that grows faster than the danger does, and that is the point a difficulty curve and a scoring table have to agree on before either one works.
It also cannot tell two runs of equal length apart, and a game that cannot tell its runs apart has no way to ask for a better one.
What four multiplier shapes do to the same ten events
The cheapest fix is to multiply by something that grows with the run. Four versions of that idea, all scored on ten successful events worth 100 points each, come out like this.
| Rule | What the tenth event pays | Total for ten | Multiple of flat |
|---|---|---|---|
| Flat, 100 an event | 100 | 1,000 | 1.0 |
| Multiplier counts the event, so the tenth pays ten times | 1,000 | 5,500 | 5.5 |
| The same rule capped at five times | 500 | 4,000 | 4.0 |
| Doubling each event, capped at eight times | 800 | 6,300 | 6.3 |
The count multiplier is the shape most games reach for and it is quadratic, not linear, which is easy to forget. Ten events pay 5.5 times the flat total; twenty events under the same rule pay 21,000 against a flat 2,000, a multiple of 10.5. Twice the play for nearly twice the multiple again, and it is that tail that decides the leaderboard.
Capping the multiplier is the usual brake. A cap at five times leaves the ten-event total at 4,000 instead of 5,500 and the twenty-event total at 9,000 instead of 21,000, so what the cap really does is flatten the last two thirds of a long run and hand the top of the table back to consistency. Whether that is a loss depends entirely on the game: a shmup wants the tail, a puzzle game usually does not.
Doubling makes the last event the valuable one
The clearest published version of the doubling shape is Pac-Man's ghosts. The first ghost eaten on one energizer pays 200, and the next three pay 400, 800 and 1,600, which is 3,000 for a clean sweep and 12,000 for a level whose four energizers are all spent that way. The dots and the energizers themselves are 2,600 for a maze: 240 dots at ten points plus four energizers at fifty. The documented maximum score for the game is 3,333,360.
Those numbers make a decision out of a food pellet. The energizer is a timer, so the fourth ghost is a gamble taken with the blue bar running down, and the reward for taking it is eight times the reward for the first. Nothing in a flat system produces that moment, and nothing in a count multiplier produces it either, because a linear multiplier only grows by one step at a time. Doubling is what makes the player ask whether the run is worth extending rather than whether it is worth continuing.
The cost of doubling is that it is a knife edge. Once the fourth ghost is worth 1,600, a rule that produces a fifth is worth 3,200, and every table built on it has to decide what happens at the end of the sequence. Pac-Man's answer is that the sequence ends at four because there are four ghosts, which is a constraint the game got for free from its own fiction.
The reset condition is the real design decision
Brick Buster resets on paddle contact, and the reason is visible in the game's other rule. The ball leaves the paddle at an angle set by where along the paddle it landed, which means every return is a chance to aim rather than a chance to be lucky. A reset that fires on contact therefore pushes the player towards the aiming the game is actually about, rather than punishing them for a mistake they did not make.
The alternative resets say different things. A timer reset rewards speed and punishes hesitation; a reset on damage rewards caution; a multiplier that never resets rewards patience above everything and turns a skill game into a session-length one. None of these is wrong, and a game can only carry one of them without confusing the player about which run it wants.
| Bricks broken in one chain | Highest multiplier reached | Points from the chain | Points per brick |
|---|---|---|---|
| 10 | 3 times | 500 | 50 |
| 20 | 6 times | 1,625 | 81 |
| 28 | 8 times | 2,975 | 106 |
| 40 | 8 times | 5,375 | 134 |
The cap shows up in the third and fourth rows. The multiplier reaches eight on the twenty-eighth brick and stays there, so every brick after that pays the same 200 points as the twenty-eighth did. The chain is still worth continuing — a brick is never worth nothing — but the reason to keep it alive stops growing exactly where a good player starts to feel the run, and a player who notices that will start taking the risks that end it.
There is one more consequence worth naming. A payout table is read as a promise, and the same argument runs through the merge sizes in Merge Drop: the numbers on the table are what tells a player which of their instincts the game intends to reward. Get that table right and nobody needs the multiplier explained to them.
Pros and cons
Pros
- A chain multiplier scores the skill the game is built on. In Brick Buster it cannot rise unless the ball keeps avoiding the paddle, so it pays for aim rather than for survival.
- Two runs of equal length stop looking identical, which is what gives a replay a reason to exist.
- The implementation is a counter, a division and a cap — under ten lines, with no state that can desynchronise from the simulation.
- A visible multiplier turns a bad run into an explanation. The player can see the moment the chain went, and it is the moment they remember.
- A cap keeps the totals on a comparable scale, so a leaderboard stays readable after the first absurd score arrives.
Cons
- A reset punishes whichever moment is least lucky rather than least skilful. One bad bounce costs a minute of careful play, and the bounce is what the player blames.
- Because the tail of a chain is worth several times the head, the total is far more sensitive to the late rule than to the early one, so small tweaks at the top move every score on the board.
- A capped multiplier stops accelerating before the chain ends. Past the cap the incentive to extend a run is flat while the risk of losing it is still rising.
- Doubling is worse than a cap on this front: it makes the final event the largest in the sequence, so the whole table has to be designed around what happens when the sequence runs out of things to double.
Sources
- The Pac-Man Dossier — the 200, 400, 800 and 1,600 point ghost sequence, the 3,000 per energizer and 12,000 per level it adds up to, the 2,600 points of dots and energizers per maze, and the 3,333,360 perfect score.
- Score (game) — the definition of a score as a quantity that events raise or lower, the level and time multipliers used on top of it, and the way a high-score list turns a number into a challenge with replay value.
- Combo (video gaming) — combos as a sequence of actions inside a strict timing window, their use as a score modifier, and the Street Fighter II account of a combo mechanic that arrived as a bug and was left in.