Ask a fighting game player who the best is and you’ll get a name. Ask a Panel de Pon player and you’ll get a pause, a “well, it depends on the mode,” and then three different answers depending on who’s in the room. This puzzle scene has always been small, scattered across regions and console generations, and allergic to the kind of centralized leaderboard that makes “world record” a clean word. That doesn’t mean benchmarks don’t exist. It means you have to know where to look, and what the numbers actually prove.

This is a guide to how the scene measures excellence: what counts as a record in each mode, what benchmark numbers separate the tiers, and how to read a leaderboard without getting fooled. If you’ve ever wondered whether your Endless run is “good,” this is the frame you’ve been missing.

Why Panel de Pon records are harder to pin down than you’d expect

Three things fracture the record-keeping. First, the game shipped under many names — Panel de Pon, Tetris Attack, Pokémon Puzzle League, Planet Puzzle League — and each version has slightly different speed curves, scoring quirks, and inputs. A “record” on one is not portable to another. Second, the community that chases high scores overlaps only partially with the community that plays competitive Versus, so the two halves of the scene track different things. Third, the game rewards two contradictory skills — surviving forever and clearing fast — so a single leaderboard could never capture “best.”

The practical upshot: when someone quotes a Panel de Pon record, the first question is always which version, which mode, and verified where? A benchmark without those three tags is trivia, not a record.

The benchmarks that actually matter, mode by mode

Rather than chase specific holders — who rotate, and whose runs live on scattered video clips — it’s more useful to understand the benchmark tiers for each mode. These are the thresholds the scene informally treats as “you’ve arrived.”

Endless / Marathon: the score-and-speed grind

Endless is the mode most people mean when they say “high score.” Difficulty ramps by speed level, and the run ends when your stack tops out. Two numbers matter: the speed level you reach and the score you post. The catch is that they trade off — you can farm score slowly at a survivable speed, or push into fast speed tiers where a single misread ends you. Elite runs are the ones that do both: deep into the fastest speed levels and a score that proves they weren’t just stalling. If you want a structured path toward those numbers, our Endless mode survival roadmap lays out the progression.

Score Attack and Time Attack: the constrained sprints

These modes cap the clock and ask how much score you can extract, or how fast you can hit a target. Because the window is fixed, the benchmark is cleaner than Endless — everyone plays the same length, so the number is directly comparable. This is where chain efficiency becomes brutal: a top run isn’t landing bigger chains than yours, it’s landing them more often, with less wasted cursor motion between them. Score Attack is the mode where APM stops being a bragging stat and starts being the ceiling on your score.

Stage Clear RTA: the speedrunner’s domain

Stage Clear is a fixed sequence of boards you beat as fast as possible, and it’s the mode that lives most comfortably on speedrun leaderboards. Here “record” means something concrete: real-time from start to final clear, timed and video-verified. This is the closest Panel de Pon gets to an official record, because the run has a defined start, a defined end, and a rule set that speedrun communities enforce. If any part of the scene has a clean “world record” line, it’s this one.

Versus: the skill that resists a leaderboard

Head-to-head Versus is where most people would say the true best players live — and it’s the mode with no score to record at all. You can’t post a “Versus world record”; you can only win sets. Benchmarks here are relational: bracket placement, who beats whom, how you do against known strong players. That’s why the Versus half of the scene tracks results (tournament finishes, head-to-head history) rather than numbers. Understanding that distinction is half of understanding the scene, and our competitive scene guide for new players maps out where those matches actually happen.

How to read a leaderboard without getting fooled

A number is only as good as its context. Before you’re impressed — or discouraged — by any Panel de Pon record, run it through four filters.

Version. Tetris Attack, Pokémon Puzzle League, and Planet Puzzle League have different speed pacing and scoring. A score that’s monstrous on one may be mid on another. Always match versions before comparing.

Ruleset. Did the run use default settings? Some high scores come from non-standard difficulty or speed starts. A record only means something against the same starting conditions.

Verification. Is there video? A screenshot of a final score proves the number existed on someone’s screen, not that it was earned legitimately. The scene’s trustworthy benchmarks are the ones with a full run on tape.

Recency and platform. Emulator, original hardware, and modern re-release can behave differently on input timing. A run’s platform tells you whether the conditions match yours before you treat its number as a target.

Where the records actually live in 2026

There is no single homepage. In practice the benchmarks are spread across speedrun leaderboards (strongest for Stage Clear RTA and version-specific categories), high-score threads and Discord pins inside the puzzle community, and video archives where the actual runs can be watched and verified. The Versus results live in tournament brackets and the memories of the people who ran them. If you want to find the current bar for a mode, start at the speedrun leaderboard for your exact version, then cross-check against community video — the two together are far more reliable than any single quoted figure.

How to benchmark yourself against the scene

You don’t need a world record to know where you stand. Pick one mode and one version, then measure the same number repeatedly under identical settings. In Endless, track the speed level you reliably survive to, not your single lucky peak. In Score Attack, log score across ten runs and watch the median move. In Stage Clear, time yourself and shave the sloppy transitions. The scene’s benchmarks become useful the moment you have a stable personal number to compare against them — a single best run tells you almost nothing, a consistent median tells you everything.

And if the comparison starts to sting, remember the ceiling is built on fundamentals you can train directly. Raw hand speed is one of them; our guide on raising your APM and finger technique is the most direct lever on Score Attack and Endless numbers alike.

Frequently asked questions

Is there an official Panel de Pon world record?

Not in a single centralized sense. The closest thing to official records are version-specific Stage Clear RTA categories on speedrun leaderboards, which are timed and video-verified. High-score modes like Endless are tracked by the community rather than any official body.

Why do different sources quote different records?

Because they’re usually measuring different versions, modes, or rulesets. A Tetris Attack score and a Pokémon Puzzle League score aren’t directly comparable, and a run on default speed isn’t comparable to one with a custom start. Always check version, mode, and settings before comparing.

What score counts as “good” in Endless mode?

There’s no universal number, because it depends on version and speed settings. The better benchmark is the speed level you can consistently survive to combined with a score that shows you weren’t just stalling. Track your median across many runs rather than a single peak.

Who is the best Panel de Pon player?

It depends on the mode. High-score and speedrun excellence live on leaderboards, while the strongest Versus players are judged by tournament results and head-to-head history rather than any recorded number. There’s no single ranking that spans all of it.

How do I start chasing a benchmark myself?

Pick one version and one mode, lock the settings, and measure the same metric across many runs. Compare your stable median — not your best fluke — against the scene’s tier thresholds, and train the underlying fundamentals like chain efficiency and hand speed.