5 tips to score higher in Pebble Merge

Small habits, big boards. Here is how the scoring actually works, and the habits that move your number up.

People ask us for the secret to a high Pebble Merge score. There isn't one. The players at the top of our leaderboard are not the fastest hands on the board. They are the ones who understand where the points live, and who keep the pile flat while they wait for them. This is a note from the people who built the scoring. About what the numbers actually reward.

The board is a low tier problem

We only ever drop tiers zero through four. Everything above that you build yourself. And the drops are weighted hard toward the small end. A fresh pebble is Sand 34 percent of the time, Clay 26, Moss 20, Jade 13, and Celestite just 7. So roughly 87 percent of what falls into your hands is in the bottom three tiers. The game is mostly about moving small stones. Keep that in your head. Trust this. A round that ends with a tall spike of mid tiers is a round where the small stuff was never cleared, and we have watched that spike end more runs than any clever combo ever did, which is a sentence we wrote after watching the recording of one specific player lose a board they had nearly won because they were hunting a pretty stone instead of clearing the floor under it.

Where the points actually are

The score table is short and steep. A merge that makes a Clay is worth 3 points. One that makes Jade is worth 10. The very top, a single Sunstone merge, pays 66. That last number sounds like the whole game. It isn't. We worked out what a full Sunstone actually costs, and the answer embarrassed our instincts. To build one from raw Sand you perform 1023 merges, and the bottom three tiers alone, the Clay, Moss, and Jade merges, account for about 4,350 of the roughly 7,075 total points in that tree, which means the prettiest stone on the board is also the least important one for your actual score, and the players at the top of the leaderboard got there by never once forgetting that fact. The crown is a small slice. The score lives in the volume of small, clean merges. We did the math so you would not have to.

This is the trap. The Sunstone is the pretty goal, so it is tempting to chase it. But you only reach it by keeping the floor clear for over a thousand earlier merges, and most of your points were already banked down there. Neglect the base and you strand mid tiers that are worth almost nothing on their own. Let the high tier happen as a side effect of good housekeeping. That is not a polite slogan. It is the arithmetic, and arithmetic does not care what you wanted the game to reward.

1. Keep the board low and wide

A single tall column. That ends most rounds. We mean it. Spread your drops across the full width so pebbles have room to meet their twins. Aim for a gentle hill, never a spike. When the pile is wide and shallow, every new pebble has more chances to land next to a match. We say this first because it is the one habit that fixes the others.

2. Merge from the bottom up

When you can, settle a pebble directly onto a matching partner that is already low. Bottom merges free up space immediately and chain upward. Stacking and hoping tends to lock tiers apart, because the partner you needed gets buried under three things you dropped while waiting. Patience. That pays for itself at the top, where a single cleared row can suddenly let five pending merges fire in one drop and the board opens up like it was never under pressure at all, the way a clogged sink suddenly drains once you clear the one thing blocking it and you stand there wondering why you spent ten minutes nudging the wrong pieces instead of pulling the clog out from the bottom.

3. Use the side walls as shelves

The walls gently hold a pebble in place. Tuck a needed tier into a corner and it sits there, out of the flow, while you go find its match. Without the wall, that same pebble rolls into the middle of a mismatch and you lose the pairing. Corners are free storage. Use them. We added the walls mostly to hold the scene together, and only later noticed they had become the best tool in the game.

4. Chain on purpose, not by mashing

One drop can set off several merges in a row. From the third merge in that chain the score multiplier is 1.15, and from the fourth it is 1.30. So a clean four deep chain is worth meaningfully more than four separate drops. The catch is that we measured this bonus on purpose and kept it small, about 2 percent inflation under normal play, because a big multiplier rewards sloppy mashing more than careful setup. The excitement is in the juice. The floating text. The sound. The score only nudges. Set up chains when they appear. Do not force them by spamming drops, because the board will punish the spam before the multiplier ever rewards it, and a player who learns to chase the floating numbers instead of the calm board shape ends up both lower on the leaderboard and more tired than the player who never once looked at the score at all.

5. Use the grace second

A pebble that just spawned cannot end your round for a full 1000 milliseconds. We built that delay in on purpose. A bad spawn should not punish you before you have read the board. When the pile looks like it is about to breach, that second is your window. Place the next drop to relieve pressure instead of to chase points, because the one second grace exists precisely so that a moment of bad luck never hardens into a moment of lost progress, and learning to use it rather than fight it is the entire difference between a player who quietly improves and a player who keeps angrily restarting. Breathe. Then drop.

6. Practice one habit on the Daily board

Daily mode gives everyone the same seeded board. It is the best place to drill a single habit. Try stay wide for a week and watch your daily best climb. The shared leaderboard removes the noise of random seeds, so you can see whether the habit actually moved your number. We use it ourselves to test advice like this, and we have quietly retired more than one tip that looked good in the prototype and collapsed the moment a real player sat down.

The multiplier we threw away

We almost shipped a greedier cascade bonus. The first scheme we tried was 1, then 1.5, then 2, then 3 times the points for deep chains. It felt exciting in the prototype. Then we ran it through a bot that played randomly and a bot that played with aim. The random bot's score went up 63.5 percent. The aimed bot's went up only 48.3. A reward that helps the careless player more than the careful one is backwards. It also flattened the gap between a good round and a bad one from about 45 percentage points down to 32. So we cut the multiplier down to almost nothing. The chain still feels great. It just does not lie about your skill, and a score that lies is worse than a small score because it teaches you the wrong habits.

A sanity check

If your rounds keep ending around the same tier, the fix is almost always board shape, not speed. Slow down. Widen the pile. Let the merges find you. The points are already on the board. You are just clearing the path to them, one quiet row at a time, and the leaderboard will follow the quiet rows long before it follows the fast hands.

← Back to the blog