Everything in the play wing so far is a small thing moving through danger, or a thing you drop and let physics sort out. PULSEMERGE is a grid, and the grid is the entire problem. Light keeps arriving in a seven by seven jar, three orbs at a time, a little faster every minute, and your only job is to make room faster than it fills. There is nothing to dodge and nothing chasing you. The pressure is arithmetic, and arithmetic is patient.
Two verbs, and both of them stay
Most games pick one input and defend it. This one was built as two prototypes so the choice could be made with a thumb rather than an argument, and the answer that came back was both. They are not difficulty settings on the same game; they are two different relationships with the same jar.
GATHER, the first door
Push any orb one square. Nothing is ever refused: an empty square accepts it, an occupied one swaps. Any three of a colour touching each other fold into one brighter orb, and the board is rescanned afterwards so a good arrangement can cascade. You are not making the groups, you are choosing which of the board’s groups happens, and in what order.
DRAW, the second door
Drag through touching orbs of one colour and on release the whole chain becomes a single orb one rung higher, where you lifted your finger. A chain of six pays nine times what three chains of two pay, so the greedy line is always visible and always costs you the seconds it takes to trace. Here you make the group yourself, which means you can also make a mistake nobody offered you.
The fuller the jar, the more it pays
Every clear is multiplied by how full the jar is, and the curve is a square rather than a line, so a clear at ninety percent pays about two and a half times the same clear on an open board. Nothing forces you to let it get dangerous. The arithmetic just quietly makes safety the poor option. A simulated cautious player scored 5,400 and drowned at ninety seven seconds; a greedy one scored 8,760 and was still alive at two minutes. Chosen danger is a thrill and imposed speed is only stress, and the difference is entirely who decided.
What is on the board at minute two
Tiers four and five, which do not arrive in any pulse. They exist only because you made them, and each one now needs a partner you must also make. That is the part a scoreboard cannot show you: at ten seconds the jar holds fifteen orbs in three colours and you are hunting for pairs on open ground; at two minutes it holds forty in five colours with forty percent less time between pulses, and half of what is in your way is your own past success.
An unmatched jar is a beat, not a loss
An early build ended the run the instant no two orbs of a colour happened to touch. It felt reasonable and it was badly wrong: a competent player was dead in three to six seconds on a jar that was three quarters empty, and no human ever saw it because a human plays slower than the pulse refills. Waiting is always survivable, because the pulse brings material. Only a jar with no room left ends anything.
The mystery orb, and what it does not do
There is a randomised reward here, on an unpredictable schedule, with a proper crack-open beat. The odds are printed on the screen. Nothing in it is ever purchasable with money, everything in it is earnable by playing, there are no energy timers, and the daily jar ignores all of it. The draw is resolved before a single pixel of the animation exists, and the reel is built from the result, which means it is structurally incapable of showing you a near miss you did not actually have.
The quiet parts
No account, no ads, nothing following you around. Your best, your motes and your jar live in your own browser and nowhere else. There is no leaderboard, on purpose: a global top hundred motivates the handful who can see themselves in it and gives everybody else a reason to leave. The streak is kind, too. Missing a day spends a forgiveness you earned by playing, and it never takes anything away.