The "exponential scoring" trap in match-clear games: a pseudo-challenge case

Building Bubble Pop, we used the "group-size × (group-size − 1)" scoring formula. This is the SameGame classic. But the more I tuned, the more I felt something was off — this formula pushes players into counterintuitive decisions and the experience is strange. This piece is my reflection.

The classic SameGame scoring

Clearing a same-color group of n tiles scores n × (n − 1):

• Clear 2: 2 points
• Clear 3: 6 points
• Clear 5: 20 points
• Clear 10: 90 points
• Clear 20: 380 points
• Clear 30: 870 points

On the surface this is a sensible "reward bigger groups" — larger = more valuable. But in practice it hides a counterintuitive strategy: players should defer clearing small groups, preferring to merge everything into super-groups before clearing.

Because clearing 2 earns 2 points, but waiting could grow those 2 into 4 (12 points) or 6 (30 points). Same two original tiles, different merge timing, 15× score difference.

So the optimal strategy is: don't clear unless you must; wait until same colors can merge into one group, then clear together.

The experience problems this creates

When I tried the optimal strategy, the problems surfaced.

1. Most of the time you don't act. Players stare at the board 30+ seconds trying to compute "which click maximizes future merges." This thinking has no feedback; the screen is still; experience is dull.

2. Single mistakes are catastrophic. If you could clear 18 (306 points) but accidentally cleared 3 (6 points), the rest is now disconnected and you can only clear 5+6+4 (62 points). One misclick: 220-point loss.

3. The "greed trap." You see a 5-group and know clicking gets 20. But you hesitate: "If I wait, could this become 8?" This hesitation is anxious because there's no formula for the answer — just gambling.

These three together make Bubble Pop tenser than we intended — and that tension isn't due to challenge, but because the scoring scheme itself penalizes players.

"Exponential reward" appears as reward but is actually punishment — it punishes "normal moves."

Why this design persists everywhere

SameGame has used this rule since 1985, copied by countless variants — Bubble Witch, Pop Master, dozens of "match-3-likes." Despite known UX problems, why universal?

My guesses:

1. Visually appealing. Seeing "+380!" pop up on screen feels much more dramatic than "+20." When players share screenshots, big numbers spread further.

2. Creates a "high-score vs. casual" contrast. Casuals score ~200; experts score ~2000 — 10× gap. This disparity creates an illusion of "still room to grow," keeping players invested.

3. Gives the game pseudo-complexity. Players think they're "doing strategy" when they're actually being pushed into specific moves by the exponential curve. The "I'm using my brain" feeling makes casuals think they're playing an "intellectual game" rather than passing time.

So this design works commercially. But the price is experiential dishonesty — players think they're challenging themselves, but they're being routed.

Our final compromise

BverGame's Bubble Pop keeps "group-size × (group-size − 1)" because that's the SameGame classic and players expect it. But we did three things to soften the negatives:

1. Constrained session length. Board is fixed at 12x12, 144 initial cells. Players spend at most 5–7 minutes per round. Avoids 30-minute marathons and limits "hesitation cost."

2. No theoretical max-score display. We show only current score, not "theoretically you could've earned 5000 but only got 200." Avoids the "I'm far from optimal" frustration.

3. Clear-all bonus. If you fully clear the board (hard but possible), +1000. Gives perfectionists a clear goal instead of endless number-chasing.

If I started over

Honestly, if I weren't paying tribute to the SameGame classic, I'd use linear scoring: 1 point per tile cleared. Or logarithmic scoring: log(n) × 10 per group of n.

Linear scoring's benefit: players freely clear any size group, no timing anxiety. Each clear is a pure achievement, not "did I waste an opportunity?"

Logarithmic scoring's benefit: still mildly rewards bigger groups (so big clears feel a bit better) but no exponential blowup, so "small operations" and "big operations" aren't drastically different.

I prototyped both. Player experience is noticeably more relaxed. But both deviate from SameGame tradition, so we didn't ship them. BverGame is a homage product; fully overturning classics would feel jarring to "original players."

Tips for indie designers

This taught me a new view on game design: scoring isn't just a "reward" mechanic — it defines the game's entire pacing and strategy space. Switch the scoring formula and the same mechanic morphs from "relaxing puzzle" to "anxious optimizer."

If you're doing indie game design, try these questions:

• Does my scoring scheme reward "natural reactions" or "counterintuitive computation"?
• If a player just plays casually, how much do they score? Is that "enough to motivate continuing"?
• What's the ratio between "optimal play" and "normal play"? Does it make normal players feel terrible?

The better these three questions are answered, the more natural the experience.

Closing

Bubble Pop is still a good game. SameGame is still a classic. But I want players to know what the scoring formula hides — it's "reward big groups" in disguise, but actually "punish normal moves" in tool.

Knowing this, you can freely choose your play: chase optimal, or relax and click freely. Both are fine.

Max is BverGame's co-operator. This is a reflection on our own product design, not criticism of SameGame or its derivatives. Multi-objective trade-offs in design are never "right" or "wrong."