Skip to content

Eat and Grow Arenas: How to Stop Dying at Medium Size

Mid-size is the dangerous size: big enough to be hunted, too small to hunt back. Route choice, cut-off angles and when to feed quietly.

By Published 6 min read

Reign Creative LLC makes Jellyfish Arena Survivor io, which this article links to, so we have a commercial interest in it. How we write and check articles.

Key takeaways

  • Mid-size is the most dangerous stage of a run: large enough to be a worthwhile target, too small to threaten anything back.
  • Growth makes you slower to turn, not just bigger, so every unit of size you gain narrows the set of escapes available to you.
  • The map edge is a trap that feels like safety — it removes half your escape directions in exchange for fewer rivals.
  • A cut-off works on geometry, not speed: you win by arriving at a point on their line, not by chasing their tail.
  • The leaderboard tells you the shape of the run — whether one rival has run away with it, or whether the field is still bunched.

Nobody dies at the start of an eat-and-grow run. You are tiny, nobody wants you, and you turn like nothing at all. Nobody much dies at the top either, once you are large enough that most of the arena gets out of the way.

Everyone dies in the middle.

Mid-size is the trap the whole format is built around: you are worth eating, and you have already given up the agility that kept you safe. Getting past that stage is most of the skill in the genre, and it is almost entirely about route choice rather than reflex.

Why growth costs you something

In a snake-shaped arena game you steer the end of a growing line and must avoid colliding with obstacles and with yourself — a problem that gets harder the longer the line becomes. That is the core tension of the whole Snake lineage, and every eat-and-grow arena inherits it.

Size is not purely an upgrade. It buys you the ability to threaten smaller rivals and it costs you manoeuvrability, which means every unit of growth narrows the set of escapes you have available. A small player has an escape in every direction. A large one has an arc.

This is why the aggressive early game that works so well for the first thirty seconds stops working at exactly the moment it feels like it should be paying off. Nothing about your play got worse. The escape set shrank.

The mid-size playbook

Four habits get most players through it.

Feed away from the centre. The middle of the map has the most food and the most traffic, and at mid-size the traffic is the binding constraint. Somewhere two-thirds of the way out usually has enough to eat and half the density.

Keep two escapes open. Before you commit to a patch of food, note where you would go if something large appeared. If the answer is "one direction", it is not a good patch. If the answer is "back the way I came", it is a worse one, because that is the direction something is most likely to be arriving from.

Do not settle on the wall. The edge feels safe because it is quiet. It is quiet because it is a dead end. Sitting against a boundary removes about half your escape directions and a corner removes most of the rest, which turns a survivable encounter into an unsurvivable one. Use the edge as a corridor, not as a home.

Move between patches deliberately. The transitions are when you die, not the feeding. Look before you travel, and travel across open ground rather than through the middle of a crowd.

Cut-offs are geometry, not speed

The single most useful tactical idea in the genre is that you never catch anything by chasing it.

A cut-off works because you pick a point on your target's projected line and arrive there before they do. You are not racing them, you are intercepting them, and the work is done by choosing the right point rather than by moving quickly. If you are directly behind a rival matching their speed, you will stay directly behind them until one of you gets bored.

The defensive version of this is worth more than the offensive version. If something large is converging on a point ahead of you rather than following you, you are already being cut off. The instinct at that moment is to speed up along your current line, which is exactly what the cut-off is designed to punish. Change your line early — a small deviation made soon beats a large one made late, because turning costs you more the bigger you have become.

Reading the leaderboard as a decision tool

Jellyfish Arena Survivor io tracks the ten largest rivals in the arena while you play, with a crown on whoever is currently first. That board is not a vanity display — it tells you what kind of run you are in.

The leader has run away with it. Attacking anything is a risk with no path to first place attached. Feed quietly, stay clear, and wait for the field to change. There is always a moment when the leader over-commits.

The top ten are bunched. One good interception moves you several places. This is when risk is correctly priced, and when passive play is the actual mistake.

You are on the board and near the bottom of it. You are now visible, which means you are a target. This is the point to move outward and get less interesting for a while.

The general skill of reading a scoreboard as a strategy input rather than a result — deciding when to take risk based on where you sit — is the same one that drives score attack strategy, even though the games look nothing alike.

Why we tuned turn radius instead of speed

Our arena uses one-finger steering with momentum, which was the single most argued-about decision in the build.

The obvious lever for balancing size is speed: make big jellyfish slower, and smaller ones can outrun them. We tried it, and it made the game worse in a way that took a while to name. Speed differences turn every encounter into a chase, and chases in an arena game are boring — they resolve by arithmetic, not by decision, and the outcome is fixed the moment the chase starts.

So we tuned turn radius instead. A large jellyfish moves at a comparable speed and commits harder to a direction. What that changes is which cut-off angles are actually available: a shallow intercept works against a large rival because they cannot tighten their line to answer it, while the same angle against a small one fails because they simply turn inside you.

The result is that encounters are decided by the angle you chose several seconds earlier rather than by who has the higher number. That is a preference rather than an objectively correct design, and it has a real cost: players who expect to escape by out-running things find that option is not there, and it takes a few runs to stop reaching for it.

The momentum has one more consequence worth knowing. Because the steering carries, a deliberate wide arc is a genuine tool — you can commit to a curve that takes you around a rival rather than away from them, which is often the only way past something large in a confined space.

What the opposition being AI changes

The arena fills with up to 100 AI-controlled rivals rather than other people, which we state plainly rather than implying live multiplayer, and it has a tactical consequence that is worth knowing rather than glossing over.

AI rivals are more consistent than humans. They do not tilt, they do not pursue you across the map out of spite, and they do not disconnect halfway through an encounter. That makes cut-off geometry more reliable and makes the mid-size stage more learnable, because the same angle produces the same outcome often enough to practise it.

It also means the arena will not surprise you the way a lobby of people does. If unpredictability is the thing you want from the genre, that is a fair reason to look elsewhere, and what bots change about an io arena sets out the trade in full. If you have not met the format before, what the io label actually means is the shorter starting point.

The five-run drill

  1. Run one: do not attack anything. See how large you can get by feeding alone.
  2. Run two: feed only outside the centre third of the map, and note whether you last longer.
  3. Run three: practise one cut-off per run, and abandon it the moment the angle stops working.
  4. Run four: check the leaderboard every time you finish a patch, and let it decide whether you hunt.
  5. Run five: play normally. The difference is usually obvious.

Nothing there is about reflex. The genre rewards where you were going three seconds ago, which is a much more learnable skill than it looks. The arena and the studio's other short-session games are listed under Action & Arcade Games.

Good to know

Frequently asked questions

Why do I keep dying at medium size?
Because medium size is structurally the worst place to be. Small, you are not worth the effort and you turn on a coin. Large, most of the field avoids you. In between, you are a meal worth chasing and you have already lost the agility that kept you alive when you were small. The fix is not to grow faster; it is to change where you feed once you hit that stage — away from the centre, away from the leader, and along routes that keep at least two escapes open.
How do you cut off a bigger rival?
By aiming at where they will be, not at where they are. A cut-off is a geometry problem: you pick a point on their projected line and arrive there first. Chasing directly behind a rival almost never works, because you are matching their speed while they set the direction. The corollary is defensive and more useful — if something bigger is converging on a point ahead of you rather than following you, you are already being cut off, and the answer is to change your line early rather than to accelerate along it.
Should you play near the edge of the map?
Generally no, even though it feels safer. The edge is quieter because fewer rivals go there, and it removes roughly half of your escape directions in exchange. A corner removes three quarters. When something large does come along the wall, there is nowhere to break to. Feed in the open between traffic rather than in the dead zone at the boundary, and if you must use the edge, treat it as a corridor you are passing through instead of a place to settle.
When should you stop hunting and just feed?
Whenever the leaderboard shows the top rival is far ahead of the pack. If the leader is out of reach, attacking anything mid-size is a risk with no route to first place attached, and the correct play is quiet accumulation until the field changes. If the top ten are bunched, the opposite is true — one good cut-off can move you several places, and the risk is worth it.
What is the minimap actually for?
Three readings, in order of usefulness. Where the density is, so you know which quarter of the map has food rather than traffic. Where the leader is, so you can choose to be somewhere else. And where you are relative to the edges, which is the one people forget and the reason they end up in corners. Glance at it between feeds rather than staring at it — the arena is where you die, not the minimap.

Sources

  1. Agar.io (opens in a new tab)Wikipedia — accessed
  2. Snake (video game genre) (opens in a new tab)Wikipedia — accessed

About the author

— The independent app studio that builds the apps covered on this site and publishes every article here. On Google Play it publishes as Reign Collective Apps.

We use AI tools to help research, draft and edit our articles. Each article lists the sources its factual claims rely on, and details about our apps come from their live Google Play listings. No article on this site has been reviewed by an independent expert unless the reviewer is named on it.

Spotted an error? Tell us — how corrections are handled is set out in our editorial policy.

Keep reading