Most people choose a game by looking at thumbnails, which tells you almost nothing about whether you will enjoy it. A better method starts from what you want the next twenty minutes to feel like and works backwards. This guide is that method, plus how to read the signals a game page gives you.
Start with the mood, not the genre
Genre labels describe mechanics, and mechanics are a poor predictor of enjoyment. 'Platformer' covers both a tense precision challenge and a relaxed exploration game. What actually predicts enjoyment is the kind of attention a game asks for.
There are broadly four, and most robot arcade games sit clearly in one.
Choosing against your mood is the main reason a perfectly good game feels wrong. A thinking game when you wanted to switch off feels like homework; a reflex game when you wanted to think feels shallow. Neither is the game's fault.
| If you want... | Look for | Because |
|---|---|---|
| To switch off | One or two inputs, endless structure | Nothing to plan; the loop carries itself |
| To concentrate | Combat, waves, timing systems | Sustained attention with immediate feedback |
| To think | No timer, resource limits, exploration | Planning is rewarded and hesitation is free |
| To be tested | Fixed patterns, high difficulty rating | Repeated attempts that teach something specific |
Then filter by time honestly
The number to check is not run length but how long until you are playing at your usual standard, which is longer for games with more inputs.
Under five minutes: one or two inputs, no spatial memory. You want to be at your normal level within seconds.
Fifteen to thirty minutes: anything with combat. There is warm-up cost, and this is enough time to pay it and still have a good session afterwards.
An hour or more: boss fights and exploration games. Both depend on continuity across attempts, and both are frustrating in a window too short to reach the interesting part.
Getting this wrong is the second most common cause of a game feeling bad. A boss fight in a ten-minute window ends exactly where it starts being good.
Ad space reserved. See README.md to connect an ad unit.
What difficulty ratings actually mean
Ratings are usually about how demanding a game is, not how much content it has, and the top of the scale means something specific that is worth understanding before you pick it.
- Easy: a beginner can have a satisfying run within a few attempts. Not shallow - one-input games can be Easy and still take a long time to master.
- Medium: the game expects one specific skill and you will be poor at it until you build it. Usually the best default.
- Hard: it expects several skills at once, and weakness in any of them will end runs.
- Expert: losing repeatedly while you learn a fixed pattern is part of the intended experience, not a sign you chose badly.
That last point matters. An Expert game is not a reward for finishing the Easy ones; it is a different kind of experience, and going into it expecting to win immediately produces frustration that the design did not intend.
Reading a game page for signal
Most of what tells you whether a game is worth your time is on its page, provided the page is honest.
The controls line is the highest-signal thing there. A game listing three simultaneous inputs is not a phone game whatever else it says. A game listing one is genuinely pick-up-and-play.
The recommended device is the next most useful, specifically because it is the field most sites get wrong. If every game on a site claims to work perfectly everywhere, nobody tested them. A site that says one game is better with a mouse has probably tested the others too.
The common mistakes section, where a page has one, is worth more than the tips. Tips are usually generic; mistakes are specific to the game and reveal what it is actually about. A game whose common mistakes are all about timing is a timing game regardless of what genre it claims.
Warning signs
Things that reliably indicate a game is not worth much of your time.
- A difficulty selector instead of a curve. It is usually an admission that the difficulty curve does not work.
- Enemies distinguished only by colour. If two enemy types share a silhouette, the game has not done the work of making itself readable.
- No visible telegraphs. If attacks arrive without warning, failure is arbitrary and there is nothing to learn.
- Everything glowing. Neon works through contrast; when the scenery glows as brightly as the hazards, nothing stands out.
- Player counts, star ratings or 'trending' badges on a site with no accounts. If there is no mechanism to measure it, the number is invented.
- A description that could apply to any game in the genre. It usually means whoever wrote it had not played it.
A shortcut
If you do not want a method, here is a reasonable default: play the Easy one-input game on any site first, for about five minutes.
It tells you almost everything. If the timing feels fair, the rest of the library probably is too, because timing is the hardest thing to get right and a site that manages it in its simplest game has the fundamentals in place. If it feels arbitrary in a one-button game, nothing more complicated on the same site is going to feel better.
A decision table
If you want the method compressed into one place.
The last row is the one people most often get wrong. Wanting to improve and choosing something unfamiliar are in direct tension: most of what makes you better at a game is familiarity with it, so a practice session should almost always use something you already know.
| You have | You want | Look for | Avoid |
|---|---|---|---|
| 5 minutes | To switch off | One or two inputs, endless | Boss fights, mazes |
| 15 minutes | To concentrate | Combat, waves, timing | Long defence runs |
| 30 minutes | To think | No timer, resources, exploration | Pure reflex games |
| An hour | To be tested | Fixed patterns, Expert rating | Endless games with no ending |
| Any | To play on a phone | Discrete inputs, large targets | Continuous precise aiming |
| Any | To improve deliberately | A game you have already played | Anything new |
Reading a site rather than a game
Often the more useful judgement, because a site's habits tell you what every game on it is likely to be worth.
Look at whether difficulty ratings vary. A site where everything is 'Medium' has not rated anything; a site with a genuine spread has somebody making judgements.
Look at whether the device guidance differs between games. If every game claims to work perfectly everywhere, nobody tested them on a phone. A site admitting that one game wants a mouse has probably checked the others.
Look for numbers that could not exist. Play counts and star ratings on a site with no accounts are invented, and a site comfortable inventing those is comfortable inventing other things.
And read one game description closely. If it would apply equally to any game in the genre, whoever wrote it had not played it - and if nobody played the game before describing it, nobody fixed it before shipping it either.
Choosing for someone else
A common situation that the usual advice does not cover: picking something for a child, a colleague, or anyone who did not ask for a framework.
Default to one input. Not because they cannot handle more, but because a game that explains itself in one sentence removes the moment where someone decides this is not for them. Complexity can be introduced later by the person themselves.
Avoid anything rated Expert, obviously, and also avoid anything whose appeal depends on repeated attempts. A boss fight is a bad first experience regardless of skill, because the design assumes a commitment nobody has made yet.
Prefer games where failure is visibly self-inflicted. Someone new to the genre needs to be able to see that they mistimed a jump rather than concluding the game is arbitrary, and that clarity matters more than the difficulty level.
And check the device honestly. Handing someone a phone and a game that wants a mouse produces a bad impression of the game rather than of the choice.
When to give up on a game
Worth saying plainly, since most advice assumes you should persevere.
Give up if you cannot explain your deaths after several attempts. That is a signalling failure and no amount of practice addresses it.
Give up if the controls fight you - if inputs are dropped, if the game needs three simultaneous keys, or if a touch layout puts a button underneath your thumb's resting position. These are fixable problems that somebody did not fix.
Give up if improvement is invisible after a few sessions. A well-built arcade game shows progress quickly, because the runs are short and the feedback is immediate. A game where you cannot tell whether you are getting better is usually one where you are not.
Do not give up merely because a game is hard, or because you lost ten times, or because a category does not suit you yet. Those are all normal and all temporary. The distinction worth holding onto is between a game that is demanding and one that is unclear.
Key takeaways
- Choose by the kind of attention you want to spend - switching off, concentrating, thinking, being tested - rather than by genre.
- Filter by how long until you are playing well, not by run length; boss and exploration games need real windows.
- An Expert rating means repeated learning losses are intended, not that you picked badly.
- The controls line and the recommended device are the highest-signal fields on a game page.
- A game's common-mistakes list reveals what it is actually about better than its tips do.
- As a shortcut, play a site's simplest one-input game for five minutes; if the timing is fair there, it usually is everywhere.
- Match the game to the time you have and the kind of attention you want to spend.
- To improve, choose something you already know - familiarity is most of what makes you better.
- Judge a site by whether its ratings vary, whether device guidance differs, and whether it displays numbers it could not measure.
- Choosing for someone else: one input, nothing Expert, nothing that needs repeated attempts, and check the device.
- Abandon a game whose deaths you cannot explain, whose controls drop inputs, or where improvement is invisible.
- Do not abandon one merely for being hard - demanding and unclear are different problems.
Frequently asked questions
How do I pick a robot game I will actually enjoy?
Start from what you want the session to feel like - switching off, concentrating, thinking or being tested - rather than from the genre label.
What does an Expert difficulty rating mean?
That losing repeatedly while you learn a fixed pattern is part of the design. It is not a reward for finishing easier games, it is a different kind of experience.
Which field on a game page is most useful?
The controls line, because it tells you immediately whether the game suits your device. The recommended device is next, since it is the field most sites fake.
Are difficulty selectors a bad sign?
Usually. In an arcade game the curve should rise inside the run, and a selector is often an admission that it does not work.
How can I tell if a site's games are any good quickly?
Play its simplest one-input game for five minutes. Timing is the hardest thing to get right, so a site that manages it there usually has the fundamentals in place.
Should I pick a new game to practise with?
No. Most of what makes you better at a game is familiarity with it, so practice sessions should use something you already know.
What is the quickest way to judge a whole site?
Check whether its difficulty ratings vary and whether device guidance differs between games. Uniform answers to both mean nobody assessed anything.
What should I hand someone who has never played one?
A one-input game. It removes the moment where someone decides the genre is not for them, and complexity can follow if they want it.