You can tell within about thirty seconds whether a platformer is going to hold you. Not from the art, and not from the premise — from the first jump. There is a moment where you press the button, watch the arc, and find out whether the game and your hands are going to get along. Everything after that is negotiation.
We have loaded a great many browser platformers putting this catalogue together, and the difference between the ones worth returning to and the ones you close without thinking has very little to do with production values. It comes down to four things.
1. The jump has weight, and the weight is consistent
A jump is a promise about physics. When you press the button, the game tells you how heavy your character is, how much air control you have, and how long you are committed. The specific answer matters far less than whether the answer is the same every time.
Inconsistency is what makes a bad platformer feel unfair rather than hard. If the same input produces a slightly different arc depending on framerate, or if air control changes when you are near a wall, your brain cannot build a model — and building a model is the entire pleasure of the genre. You are not really learning levels. You are learning a body.
Crazy Bouncer is interesting on this point because it hands you a bouncing ball rather than a character with legs. A ball has momentum you cannot cancel, so the model you build is about setting up a bounce two moves ahead rather than correcting mid-air. It is a harder model to learn and a more satisfying one to have learned.
2. Failure is legible
When you fall in a good platformer, you know why. You jumped a fraction early. You held right too long. You did not see the second platform was moving. The information was on screen and you misread it.
This is the quality that lets a difficult platformer stay enjoyable through twenty retries. The player is never confused about the causal chain, so each failure feels like information rather than punishment. Compare that with a game where you die to something offscreen, or to a hitbox that does not match what you can see — one retry of that and most people are gone, correctly.
Legibility is also why platformers survive as browser games at all. They need almost nothing from the hardware: a readable silhouette, a clear ground line, and input that arrives when you press it. A platformer that gets those right in a browser tab beats a lavish one that does not.
3. There is somewhere to be going
This is the difference between a platformer and an obstacle course, and it is the reason this category is called platform adventures rather than just platformers. A gap is more interesting when crossing it takes you somewhere.
Going Up Rooftop works on this principle. It is a parkour game across the rooftops of a sprawling city, and the city is doing real work — it gives the climbing a direction and a sense of altitude gained. Jump Tarzan does something similar with a jungle: the canopy is a place, and moving through it reads as travel rather than as clearing a series of tests.
The counter-example is any platformer where the background is a repeating texture. It can still be a fine game, but it will be a game you play once for the challenge rather than one you come back to for the trip.
4. It respects the length of a session
Browser games are played in gaps. Twenty minutes before a meeting, ten minutes on a phone, an evening if the game earns one. A platformer that assumes a two-hour block will lose people who only had twenty minutes and could not get back to where they were.
The best browser platformers are built in short, self-contained runs, or they checkpoint generously, or they are structured so that a level is a satisfying unit on its own. What they do not do is put half an hour of progress at risk. There is a reason the endless-runner structure spread so widely through the genre on the web: a run that ends cleanly and restarts instantly is perfectly matched to how people actually play in a tab.
Where to start
If you want to test these four qualities yourself, the platform adventures collection is the place. Load three games and give each one sixty seconds of jumping before you judge anything else. You will find your hands make the decision well before you do.
And if the jumping is not what you are after, the same instincts carry over sideways — survival adventures run on the pleasure of somewhere to be going with rather more at stake.